Arvo Pärt MkIII "Heavy Metal" Testing...

[JonMarsh]

Notice pricing has increased for all the designs in this section of the forum. People should use the BOM's and check with suppliers for current prices

This Saturday I got around to testing the SCC300 Sonic Craft woofers, and the TC2+ woofer in the Arvo MkIII test mule. The Sonic Craft woofers did exhibit the sensitivity gains I was looking for over the Titanic MkII's, but had a bit higher mechanical Q than I was expecting. Their response is pretty smooth out to 400 Hz or so, but gets a bit ragged above that. No surprise all in all, but didn't exactly Knock me off my feet.

The TC2+ had a lot to live up to, given expectations about this driver, and for $15 more apiece than the SCC300, it delivered handily.

Total sensitivity was not as high in the mid bass as the SCC300, but low bass held up quite well in the U baffle down to nearly 50 Hz without EQ. Quite pleasing was the lack of glitches in the impedance curve until well above 1 kHz, and the relatively low amplitude of these glitches- this is a very well behaved large cone.

http://audioworx.virtualave.net/TC2Z.jpg

I also did chirped sine testing, and this was interesting- the TC2+ is clearly very articulate and "fast". Very interesting driver.

Here's the test mule with the "full aluminum load"- TC2+ woofers, HiVi midwoofers, and SS9800 Scanspeak tweeter.


http://audioworx.virtualave.net/APS03.jpg


And from the back,

http://audioworx.virtualave.net/AP2S04.jpg



I went back and tweaked up the passive crossover design after importing the new measurement data. Didn't take a lot of work to adjust for the new driver impedance and response- no topology changes were involved. Note that NO diffraction control is in place, so there's some roughness in the upper end of the tweeter reponse that should be "gone" in the finished speaker. Love that SS9800 tweeter....

This is without any external EQ, folks.... even I wonder if this isn't too good to be true, but we'll just have to build the crossovers and see what happens. At least at the measuring point, about 4 feet from the front of the speakers, this should be the response. It will be interesting to look at the system power response over a 30 degree arc, and see how that turns out.


http://audioworx.virtualave.net/AP2Plot-1.jpg


Hmmmm, got to start getting the coils and caps together! Now, the question is, do I build the crossover starting from the bottom up, or the top down? Probably better do an inventory on parts and see what I've got- I bought quite a few for the last version, so by unwinding a few inductors, I should be set to go soon.

Regards,

Jon

[Dennis H]

Dang! That's impressive unequalized response, Jon! :T

Ya know, that looks about like a natural B2-40 low end rolloff. That means, for HT, you could use most any prepro with a 40 Hz XO. Most of them have a B2 highpass and an LR4 lowpass so that would give a nice LR4 acoustic XO to the sub.

[Dennis H]

Jon, did you check the T/S parameters of the TC drivers. The O-Audio site is confusing. It says they are dual voice coil drivers but the published T/S parameters appear to be for a single coil driven with the other open. The following numbers from the PDF file are internally consistent so that implies Q was calculated with one coil.

Re = 4.1
Bl = 14.2
Qe = .57

But that would give a low Q with both coils driven. Todd's simulated response curves for both coils driven are inconsistent with that. He has a 60 liter, Q = .7 box but that's the correct size for a single coil, not both coils. And he has the same sensitivity for either series or parallel wiring. As well, he has Re = 3.8 on the HTML page but that's not internally consistent with the other numbers.

Help. Confused. :?

[ThomasW]

When Jon and I last spoke, he said when wired in series, the VC's were 5.something ohms. I don't remember if he said 5.6, 5.9 ........ anyway something like that.

I'm pretty sure he didn't measure other T/S parameters. Our fundamental concern was if the stated efficiency was accurate (it is), and where the 1st cone breakup mode was (~1.4kHz) :T

[JonMarsh]

Thanks for the kind words... there is some shaping in the crossover network, obviously, and the way the network design works with the tapered U baffle for the woofer, but no external active EQ. The original Arvo I hoped to get down to 70 Hz flat without EQ; this is better, and as you observe could support a lowish crossover point to a monopole sub. That's exactly what I have in mind... when these are done, next task will be to get my HE-15 out of the closet... :B

The best thing about this is that it's picked up about another 3 dB of sensitivity; I expet the net sensitivity to be 87 dB for 2.83 VRMS; not exactly a SET friendly load, but considering most of the Seas Excel based systems are on the order of 82-83 dB after baffle step compensation, not shabby.

And no stack of amps and active crossovers, as with Linkwitz's Orion and Phoenix.

We'll see if it all pans out as well as hoped, but then there's the MkII current version with DPL-12's at Thomas's, which works quite well... so I have some confidence.


Jon, did you check the T/S parameters of the TC drivers. The O-Audio site is confusing. It says they are dual voice coil drivers but the published T/S parameters appear to be for a single coil driven with the other open. The following numbers from the PDF file are internally consistent so that implies Q was calculated with one coil.

Based on what I'm seeing, the specs are reasonably consistent and correlate tolerably well with the driver as delivered.

This is one of those things I hate about this business; ultimately you can only go by the delivered hardware, and hope they deliver it consistently!

That impedance plot is for two drivers; the VC's for each driver are wired in series, then the driver's are wired in parallel. It's pretty normal for the midbass impedance minima to be only slightly higher than the DCR. Based on that, I should be seeing 4.1 to 4.4 ohms at the minima, and that's pretty much the case. OK, I confess, I haven't had my CLIO system calibrated to an NIST standard, but it's probably pretty close...

I didn't measure the T/S parameters yet- I'd like to burn them in a bit and then check. It was a near miracle just finding the time to do the work yesterday- after I finish this post, it's back to work emails. Yuck. I feel like a victim of our success at Infineon....

Anyway, I'm encorage by what I'm seeing; in a different baffle design, I'm thinking that a set of four per side will work pretty well with my RD50 ribbons. While our current preference for crossing over the RD's is 4th order L-R at 600 Hz, I suspect than an 8th order elliptic LR equivalent at 400 Hz might be OK, and just the ticket with these 12's. They seem well behaved up to 800 Hz, aside from the beaming.... ;) a 400-600Hz crossover may be reasonable- something I wouldn't ever consider with any other 12 than the Eton.

Thomas's comments are based on the LspCAD prediction of the minimum impedance; which is based on crossover reactance and interaction with driver impedance; because of how the driver model is converted to an electrical model which interacts with the crossover impedance, sometimes it gives numbers that are a little low- but also, there's the interaction with the midbass crossover, and LF crossover.

I'm probably going to look at this a few times over the next days, and make sure there's nothing obvious I'm overlooking before beginning construction.

~Jon

[Dennis H]

Hey guys,

I know you are partial to metal domes but I'm wondering how hard it would be to make a Scan 9700 or a Seas Millennium work in the Arvo. Just looking at objective facts and figures, without listening to them, both fabric domes look superior to the 9800. The 9800 has 0.1mm Xmax. The 9700 has 0.4 and the Millennium has 0.5. SL has tested (and used) all of them. He found the 9700 had quite a bit less stored energy than the 9800 and the Mill was better yet but not by all that much. Cost, cheapest to most expensive, Mill, 9700, 9800. Of all those factors, Xmax looks like the biggest deal to me but what do I know?

Just curious, inquiring minds and all that good stuff.... :wink:

[JonMarsh]

Well, you know how it is, Dennis, I have my own wacky reasons.... :W

First, let me say that the Excel T25CF-002 is probably the finest soft dome tweeter made; the only other high value part of tested which comes to mind, and which I ruled out because of it's more limited range is the Hiqaphon series.

The SS9800 is flatter in portions of the spectrum which I care about, though, because of the crossover characteristic. Also, part of how I ended up choosing the SS9800, besides it's basic performance, including low distortion due to the SS SD2 motor, is due to the impedance curve and physical properties- it's very easy to drop it into the same hole (literally) as a Vifa XT25, with only minor tweaks if any to the Zobel.

With the crossover I use, equivalent to an 8th order L-R for the first 48 dB or so of attenuation, the specified Xmax hasn't been a problem.

Some of the energy storage issues for the unmodified version of this driver are due to that horrid little diffusor they put on it; the bane of hard dome tweeters and their manufacturer's attemps to make the measured response/dispersion above 15 kHz plot better (numbers over sound). It's very easily removed....

Without the diffusor, there's a shallow roll off above 15 kHz (and I mean shallow), which I suspect is due to the time difference between the center and edge of the tweeter- I've seen the same thing with similar modified MB Quart tweeters.

Last, due to differences in distribution channels, we pay less for the SS9800 than we would for the Excel, particularly at the quantites we buy.


That's an awful lot of rationalization for this early on a Sunday morning! :B


~Jon

[Al Garay]

Is there an update on the Arvo project?

I lost track when Jon was going to Colorado for his Arvo vacation.

Thanks,

Al

[ThomasW]

Jon's Colo vacation was canceled a couple of weeks ago by his boss.......... :evil:

At this point the only thing to 'update' with regard to the Arvo design is to fine tune the XO. Given Jon's recently revised work/travel schedule for Jun, July and Aug, we're not sure when time will be available for the XO update will be finalized........ :roll:

[Al Garay]

When you have a chance, can you expand on your impression of the SCC300? I realize it did not measure up to the TC2+ for the dipole Arvo project. But what about as a woofer for a 3-way (crossed at 275Hz) or a sealed subwoofer (crossed below 50Hz)?

I have a pair of SCC300 but would not be hard to replace with a pair of TC2+. I would appreciate the lower extension, smaller cabinet size perhaps even a 60 liter instead of 81 liter box (better WAF)and less likely to clip with some movies and not to mention the greater flexibility. So, I'm talking myself into a change. Just need some guidance from the master(s).

Thanks,

Al

[JonMarsh]

I haven't had time to do exhaustive testing on the SCC300, because I've been too exhausted myself! However, compared wtih most "subs", it's a better choice if you want to go up to 250Hz or so.

In a sealed box with low crossover, I'd probably pick the DPL12 over the SCC300. OTOH, the SCC300 is more sensitive, but keep in mind it has a lower XMAX also. The DPL12 is one of the lowest distortion drivers below 100 Hz in it's XMAX class. Though designed for dipoles, its a great sealed box driver.

Now, the SCC300 is a little bit of the jack of all trades, master of none school. With any driver, it's the total optimization and tradeoffs, and what do you need for your application? I may end up putting the pair I have into smallish sealed boxes with PE plate amps for the bedroom. Haven't really made up my mind. They would work in the dipole, they would also be fine in an IB; and do have the higher sensitivity I desired (compared with DPL12's or Titanic MkIIs), but for my application the TC2+ was a closer fit. Just right, instead of very close.

Regards,

Jon

[goskers]

Hi Thomas and Jon,

I am very intrigued by the arvo as I am a previous orion owner. My orion complaints were lack of overall lowend output and complexity with multichannel amps and cables everywhere.

The arvo looks like it may be right up my alley when all is done. Can someone make a guess on if the output with two mids and two tc sounds woofers will be greater than the orion's?

Will there be a passive option for a high pass that will make adding a sealed sub for the sub frequencies an option or should that be active?

Thanks and waiting in anticipation. :spin:

wasser

[ThomasW]

Al,

The TC2+ was chosen because of it's 'fit' into the Arvo design with it's all passive XO. There's certainly nothing wrong with the SCC.....

Hi Joel,

I received your email, and have replied...........

[JonMarsh]

Hi Thomas and Jon,

I am very intrigued by the arvo as I am a previous orion owner. My orion complaints were lack of overall lowend output and complexity with multichannel amps and cables everywhere.

The arvo looks like it may be right up my alley when all is done. Can someone make a guess on if the output with two mids and two tc sounds woofers will be greater than the orion's?

Will there be a passive option for a high pass that will make adding a sealed sub for the sub frequencies an option or should that be active?

Thanks and waiting in anticipation. :spin:

wasser

Hello Wasser,

Having greater output capability than the Orion was a consderation in this project- obviously, comparing the XLS 10" 12 mm Xmax woofers and single W22 midwoofer to the dual TC2+ 12" woofers (18 mm Xmax) and dual HiVi M8a woofers shows where I was headed. Another goal was wanting to have a system that could be driven with a sinlge amplifier channel on each side.

Now, the Orion is a great speaker, and I understand full well the arguements behind doing a three way active crossover and implementing all the necessary filters that way. Heck, the first 3 way active speaker I built was in the early 70's, with a homemade DIY rackmount active crossover (12 dB/octave), so it's not like I don't have a little experience there...

But, my preference in this case was also having something of a test bed for listening to relatively high power amplifier designs I mess around with (150 watts/ch and up), and I have two pairs of Aragon Palladiums as well as an 8008 BB X3 and so forth, plus some projects in progress, so amplifier power wasn't an issue.

The idea was to try to juggle and optimize the acoustic aspects of the design so that minimal or readily implmented filtering in the crossover was mostly all that was required, with some small amount of additional LF EQ.

The results with Thomas's first system using the DPL12's was quite encouraging, but to optimize for passive, we needed a bit different woofer parameters, hence the search for a slightly different woofer with a bit higher sensitivity, and readily parallelable (i.e., roughly 6-8 ohms). The TC2+ seems to fit that bill very well.


Do keep in mind that the first limit on output is the woofer and mid bass excursion; even quadrupling the acoutic power capability results only in a 6 dB output level improvement.

The goal was to produce a speaker that can be comfortable with 90 dB+ average and peaks to 110. I think this will be the case in the finished Arvo, and considering the real "footprint" isn't that much bigger than a pair of the bookshelf M8's on stands, I'm pretty excited.

For more output, there's my next project, which invovles the RD50 BG ribbon at it's core, from 600Hz or so to 5-6 kHz; and possibly four TC2+ in an array on each side from 600 Hz down, and some kind of ribbon tweeter array in the final version. Crossed over to a monopole sub below 50 Hz or so; Stryke HE15.

It's annoying having all these parts sitting around and so little time to work with them lately! I've had the RD50's for two years... :oops:

Anyway, I did order about $500 worth of crossover components today from Parts Express, to finish up what this first set should need (did I mention a high power passive crossover at 200 Hz and 1.25 kHz isn't inexpensive?), so with luck the crossovers will start coming together within a week or two. :T Tons of hardware work to do for my day job, though, and overseas travel the second half of July. I can't believe June is over half over already, either! :E

Realistically, I'm expecting to have these updated version running before the end of the summer. I wish it were sooner, but if you saw what I'm doing in my day job right now, I think you'd understand.

BTW, there's no reason this set couldn't be used with an active crossover to a monopole sub, anywhere from 60 or 75 Hz and down. I'd recommend staying as low as possible in the crossover point to keep the benefits of dipole midbass.



~Jon

[goskers]

Hi Jon,

I know this is not complying to the thread but I am interested in your use of the rd50 as well. How would you do the xo in the rd50, tc2+ and may be fountek* full ranger? Active - passive?
This would seem as it is a no-brainer for cost no object, last speaker I would ever build type project.

Thanks :Z

[ThomasW]

At this time it's almost pure speculation, but I imagine there would be a mix of XO's depending on how the system shapes up.

We'll probably buy one Hi-Vi M-12 to do a shootout between it and the TC2+ .

Pretty much any system can be built all passive, but sometimes it's not the best performance choice. Also there are times when going all passive is virtually as expensive as going active.

The Fountek ribbons are one lower cost option, their performance probably isn't on par with some of the higher end ribbons like those from Aurum Cantus or the newly upgraded Ravens.

Given the current backlog of projects, it's likely going to be years before any significant amount of effort is directed toward this project.

[Hank]

Also there are times when going all passive is virtually as expensive as going active.
That's a surprising lesson I've learned on this forum from the Poohbahs.

Given the current backlog of projects, it's likely going to be years before any significant amount of effort is directed toward this project.
Oh-No! That's the project I got the Poohbahs interested in a year (2 years) ago. :cry: :thud:

[goskers]

For now I will go ahead and build the arvo part. I hope that someday the time is made available for the rd50 project to take shape. An end all system would be great!!

Thanks for all the hard work guys.

:T :agree:

[JonMarsh]

I don't know whether Thomas is being overly realistic, or a little pessemistic!

Hmmmm, let's see. Projects.


1. Arvo Part - in progress, obviously!

2. Aragon Xmod- started PCB layout, after completing 4th spin on electrical design. Completion for 1st boards uncertain- Autumn, perhaps? Most parts on hand.

3. Line Array- untitled - how about, "That RD50 Thingie"? ;)
The only thing that stays the same is the midrange-


I'm thinking about doing a "mule" after I finish these Arvo's, setup just for evaluating 4-12" and the RD50, not worrying about a tweeter line for the time being, but just evaulting getting 75Hz to 15 kHz done as well as possible, without a super tweeter. Expected crossover point is ~500 Hz. I could proto this with an active crossover by modifying my X1 active crossover. It has some shelving EQ capabilities, in the LF which would work just right.

While I'll never say never, (regarding an active crossover for this project), the intent all along was to go mostly passive, with perhaps a simple analog EQ box (one of my balanced Diff gain cells), and use an active crossover just to go to a monople sub below 60 Hz somewhere. I've got a Sumo Delilah II I've been reserving for those duties...

The question is, should I build the RD50 mule sooner, rather than later, and temporarily mount my four TC2+ in it for evaluation before finishing the cosmetic versions of the AP2? Probably would be a good idea... plus it would be a fun, if cruel, tease to Hank... :E

Of course, it might also just distract me more than a little!

~Jon

[David Meek]

Hmmmm, let's see. Projects.


1. Arvo Part - in progress, obviously!

2. Aragon Xmod- started PCB layout, after completing 4th spin on electrical design. Completion for 1st boards uncertain- Autumn, perhaps? Most parts on hand.

3. Line Array- untitled - how about, "That RD50 Thingie"? ;)
The only thing that stays the same is the midrange-

What ever happened to the Eidolon clones?

[ThomasW]

We should probably publish a schedule. But that might get us into even more trouble.......... ;)

Right now the priorities are the Arvo, and the Aragon X-mod project.

The RD-50/75 hybrid line array was shelved because we didn't have a good woofer option, and everyone was coming out with new ribbon tweeters. Now we have a solid woofer and tweeter options, but the money has been redirected to the projects listed above.

So yes eventually we'll do the line array, but it certainly won't be this year....

[goskers]

OK, i give. I am definitely sold on the Arvo so I am itching to have that be first on your plate.
;x(
Can you tell me a little more about the new pcb's for the Aragon stuff? Will this be for sale? What exactly does it replace in the internals? How does this stack up with other stuff in regards to sonics and overall presentation?

Again, thanks to all for my new found favorite hideout.

wasser

[ThomasW]

The X-Mod is Jons project for updating the 800x series of classic black Aragon amps.

The goal is to have a powerful amp that has a sonic signature similar to Ayre, Theda or other high-end nonglobal feedback designs.

The plan is to reuse the chassis, heatsinks, transformers and a few selected components from the original PCB's.

The complexity and expense and scope of the project has grown considerably over the last 8 mos. The component cost and construction time will probably make it unrealistic for us to bring it to the retail market. We've talked about trying to make a DIY version, but we're using some fairly rare and not readily available components.

As the project moves from the planning and procurement phase to the actual building stage we'll create a thread about it.

[JonMarsh]

What ever happened to the Eidolon clones?


Ooops - my bad. They should have been on the list. Actually, David, they're more like a cross between an upgraded Eclipse and Arcus, than an Eidelon, as they're two ways, while the Eidelon is a three way. Upgraded, in the sense of lower IM in the bass and upper mids/lower highs. And a bit more bass extension, too. Finding a week off to do the veneering and build the crossover enclosures is the main "hold up"- the crossover design is pretty much done (it's "just" a glorified M8, ya know, take to it's illogical extreme). You know, if some's good, more's better, and too much is just enough.


Can you tell me a little more about the new pcb's for the Aragon stuff?

That's not really supposed to be a "public" project yet- it's been worked on and off for a couple of years- primarily because Thomas and I have quite a few Aragon's between us... some would say an embarassing number of Aragon's! You know, some folks are Bryston freaks, well, we had an Aragon period back when I was building the X1 clones, which actually continued for sometime there after. Even though they're not "state of the art" sonically, they're a pretty well balanced all around design, and as "conventional" amps go, you have to pony up to something like Krell or big ML's to get something better.

But we're also fans of Ayre products (company who's principle designer is the original founder of Avalon), an old friend of mine, and the non loop feedback thing really does work pretty nicely. We've got 2 Ayre V5's, currently both at ThomasW's, and even before getting the V5, I was playing around with some non loop feedback circuits; they have a certain naturalness which is hard to find in solid state gear; more prevalent in very good tube equipment.

Anyway, the Xmod is a combination of a complete new channel PCB, (basically completely rebuilding each channel; only original component used is the heatsink and speaker relay), and a new power supply board will also be incorporated. It's gone through two major evolutions; the current circuit is what I'm planning to have a number of PCB's built up, so we can retrofit amps. The basic design sort of evolved and "got away from itself"; it's designed to be inherently balanced input, using FETs in parallel for low noise and high class A bias current and linearity in the front end stage. The complementary differential stage is cascoded, then coupled into a complementary VAS stage by both common emitter and folded cascode connections. This uses both sides of the differential current to drive the VAS stage, and keeps linearity high. The front end is true class A in that the signal swing required to clip the amp is less than 1/4 of the idle bias current, even considering the heavy local degeneration and low gain. To allow operation of the amplifier as a DC amp and not worry about preamp offsets, I've added an option for a revised front end gain and using a Jensen bridging transformer on the input. Bandwidth with the transformer is reduced a bit on the top end (-3 dB at 130 kHz open loop, compared with 225 kHz), but the bottom end holds up well, being down -3 dB at 0.2 Hz. Yeah, I said two tenths of a Hertz. No need to worry about LF group delay.

The output stage is a variant on an emitter follower triple, designed for very high open loop linearity (typically 0.004%) and optimized for driving capacitive loads without an output decouping inductor. It uses 230 watt On Semi extended beta transistors (MJL4381 and MJL4302) instead of the 150 watt Toshiba devices. All modded amps will have the full complement of outputs usually used in a Palladium or 8008 BB, not the reduced count used in 8008ST or 8002.

The power supply board replaces all the supply caps (roughly doubling storage capacitance, and adding PP film bypass), fast rectifiers, and a new grounding scheme, plus a DC input filter to keep the toroidal power transformers quiet and working in the most linear flux region even when called to deliver high power levels. One of the weaknesses of most toroidal transformers is their susceptibility to flux imbalnce if there's small amounts of DC on the neutral to hot connection (I mean small, like a hundred or two hundred millivolts). The low impedance of the windings, and small core which gives toroidal transformers low voltage drop and relatively light weight can result in significant DC bias current with a small offset voltage, and cause the transformer to start to saturate prematurely under heavy load. Big EI core tranformers are much less susceptible, because the winding resistance is usually higher, and the core size larger for a given primary magnetizing inductance.

Because of the number of parts involved and the local regulation and stuff, I don't know that I'd call this a DIY project- if there was interest or demand, I could probably sell PCBs. The parts aren't hard to get; between Digikey, MCM, and Newark you could source just about everything or close equivalents. I'm ambivalent about doing that, though, because I don't want to publish schematics; there's some things I learned on some consulting projects which are in this design, and that has to be restricted. But I don't really want to start up a cottage industry building new Aragon modules, though I have a tech friend who would like the assembly work. Basically, it's up in the air, which is why we're aren't talking about this much.

Oooops. :oops:

We'll just have to see. Maybe we won't even like how the finished ones sound... but the brassboard is promising, and measures well for an "open loop" type design.

~Jon

[Hank]

The question is, should I build the RD50 mule sooner, rather than later, and temporarily mount my four TC2+ in it for evaluation before finishing the cosmetic versions of the AP2? Probably would be a good idea... plus it would be a fun, if cruel, tease to Hank...
Yes it would, but us DIY-ers have a touch o' masochism, don't we? :twisted:
Financially, I can't invest in RD's, woofs and tweets this year (due to two IV's running from my wallet to two college students), so I'm not pushing the Poohbah's. You should probably finish what's on your plate and save the RD50 Thingies for next year, and hopefully Jon will either manage his job hours down or find a job requiring fewer hours. :T


We should probably publish a schedule. But that might get us into even more trouble..........
Yes it would - the daily questions would drive you nuts. This is supposed to be a fun hobby, right? :scratchhead:

[JonMarsh]

Yes it would - the daily questions would drive you nuts. This is supposed to be a fun hobby, right? :scratchhead:


Yes indeedy it is!

Today it was a fun day- there was $500 worth of crossover parts waiting for me in a relatively small but heavy box... oh goody!

Jantzen and Solen inductors, Dayton polypropylene caps, mills resistors, and AudioCap Theta caps. The AudioCaps are pretty cool- 10uF film and foil are much heavier and denser than their more reasonably priced metalized brethern. The 10uF AudioCaps, though physically much smaller, weigh MORE than the 100uF Daytons. The Daytons are going to be used in the woofer crossover; the AudioCap Theta caps will be going into the tweeter crossover, of course. True film and foil caps are a little pricey, but I think they're justifiable in the tweeter circuit, and sometimes in midrange low pass.

I'm going to do some physical measurements on these caps, too, just to satisfy my curiousity. Though I don't really think ESR and ESL tells much of the story regarding sound.

Regards,

Jon

[JonMarsh]

OK guys, this is in response to requests for prelimin detailed info on current baffle design, crossover, and LSPCAD modeled behavior.

This is NOT the finished design- this is an evolutionary step, what I'm calling Release Candidate 1 on the V2 design. This is a newer/different design than ThomasW is listening to, and pretty pleased with, but that doesn't mean it's better, it's just NEWER; new baffle design, new crossover, etc.

This is not intended for you to go out and make saw dust and solder with, but it's a reference point in the development, and should allow you to estimate the cost and effort involved, as well as the kind of results we're shooting for. OK? 8) Right click and save to disk.

3/17/05 Note the links below are broken and can't be fixed, since the hosting server no longer exists. See the end of this thread or the other Avro thread for the latest data.


Front Dimensioned Drawing PDF (http://audioworx.virtualave.net/ArvoPartV2Front.pdf)


Side Dimensioned Drawing (PDF) (http://audioworx.virtualave.net/ArvoPartV2side.pdf)


Prelim Crossover Schematic PDF (http://audioworx.virtualave.net/ArvoRC1Schematic.pdf)


LSPCAD Plot Diagrams (http://audioworx.virtualave.net/ArvoTC2RC1Diagrams.pdf)


The cabinet diagrams don't detail the Oak side panels or other trim features I'm considering. It also doesn't detail the rear relief bevel I'm using for the drivers- like this...

http://audioworx.virtualave.net/ArvoUpperBack.jpg



Have a good weekend, folks.

~Jon

[Dennis H]

Bravo, Jon. Take that, all you alligators! :W

A couple of questions.....

Is the tweeter wired with reverse polarity?

The mids have a separate Zobel circuit. Does that just patch in somewhere in parallel with the leads going to the pair of drivers or does each driver get its own Zobel?

You show the 2 woofers in series but that's a series/parallel arrangement of the 4 voice coils, right?

[ThomasW]

Dennis,

Yes the tweeter is electrically out of phase. We reverse the polarity of the tweeter output leads at the zobel input/output connector. That way the tweeter itself is then wired in phase. This may seem confusing and of course it is ... :wink:

The M8a are single VC drivers.

As far as I know there are no zobels for the midwoofers

[JonMarsh]

The net electrical phase of the tweeter is out of phase, becuase the crossover, in the transistion region, is mimicing an 8th order L-R, which requries out of phase connection, as does the 2nd order L-R.

The "free floating" zobel is just connected to the input. It's optional, but it reduces the otherwise drastic impedance swings which would occur going from an 4 ohm tweeter to a series connected 16 ohm midwoofer load. With it, the impedance curve is pretty consistent from 100 Hz to 20 kHz. Not that I think there will be very many SET amps used with these... (see the cruel impedance modelled at LF- not sure if that will be "real"- I expected something closer to 3 ohms with dual 6 ohm drivers). But non feedback solid state amps also have a somewhat higher output impedance than conventional high feedback designs, and this will minimize any tonal/voicing errors wtih different amplifiers. I hope... ;)

~Jon

[JonMarsh]

Well, I finally spent a little time on a weekend on something other than (no, I'm not even going to say it, out of deference to Hank and others in similar pain... 8O )

ANYWAY, crossovers built, crossovers wired up to the test mule, test mule setup. I've never had a speaker sound this good before any tweaking or adjustment of crossover, but then I've got a pretty serious stack of drivers here; not much surprising there!

Only one speaker running. Stere image is a little narrow.... ;)

Great definition, very transparent.

I also hooked it up with one of my M8 MkIV and did some video- marked improvement in soundstage widt and depth over two M8 MkIV; I think that's because the dipole is over on the right, and I'm sitting to the left of the CRT projector, and the dipole has more uniform off axis power response. Just a guess....




Listening to "Women of the World Celtic" collection from Putumayo World Music; Maire Brennan, Fiona Joyce, Karen Matheson, Mary Black, etc. Really sweet.... I think I'm gonna like these speakers....

:B

~Jon

http://www.htguide.com/forum/attachment.php4?attachmentid=751



http://www.htguide.com/forum/attachment.php4?attachmentid=752
LF Crossover

(Don'tcha just love those AWG 12 inductors? ;) )


http://www.htguide.com/forum/attachment.php4?attachmentid=753
MidWoofer Crossover



http://www.htguide.com/forum/attachment.php4?attachmentid=754
Tweeter Crossover

The LPAD will be removed and fixed resistors (Mills) installed when value is finalized. Yeah, those are AudioCaps Theta Film and Foil in the Tweeter crossover. Expensive, but they sure sound good.

Female voice sounds REAAALLLLLYYY good on these speakers....

[JonMarsh]

Today I got most of the materials for building the 2nd V2.5 test Arvo, though it will be with a different front panel construction, partly because of problems getting the materials I used for the last two, and to try out something I may use for the RD40 based ribbon system.

Also, the 24 ft Kimber 4TC for wiring the drivers and crossovers together came in. This is a low inductance braid, teflon insulated, with net AWG of 13 per conductor. It will be used for all the driver interconnect wiring.

After considering some ideas for a home constructed "wood look" box for the crossovers, I'm now leaning towards a black anaodize rack chasis for the crossovers for each speaker, sort of like a monoblock amp.

Gotta go get some more glue now, it's a smoking 95+ in Livermore, so I've setup to do gluing in my kitchen! ;)

~Jon

[Dennis H]

Hey Jon,

For the "wood look," it's hard to beat edge-glued solid wood panels. IMHO they would work fine for baffles (maybe 2 layers) as well as for the extra pieces. You can glue them up yourself but, unless you have a surface planer and a surface sander, or a whole lot of time to spend with a belt sander, it's pretty inexpensive to buy them. Here's one source.

Solid wood panels (http://solidwoodpanels.com/solid_wood_panels_for_woodworking_plans.htm)

[Hank]

My humble suggestion would be a "wood" box at the back of the speaker, at the bottom, "resting" on the base. The base would be extended 4-5" to the rear of the bottom of the cabinet, so that the crossover box is vertical, almost as wide as the cabinet, and maybe 3" thick.

[JonMarsh]

Thanks for the ideas, Dennis and Hank.

That's an interesting source, Dennis, and it made me think more also about some stuff available locally that may be adequate to the job. The construction might be fairly easy, now that I'm thinking about it in solid woods, and it would be less expensive.

Hank, your suggestion is along the lines of what I was first thinking about, except that the crossover boards are so large (12" X 10" - a lot larger than the M8 crossover boards) there's no way to make the box that small. With the stacking space required, I think it will be close to a foot thick. But that's probably OK. Since I still hadn't decided on the side panel trim wood, that leaves the the box open, too; though I'm leaning towards maple or oak. I have some solid teak "in stock", but not enough to do all of that. ;)


Have to give this a little more thought, but I like the direction this is going... 8)

~Jon

[David Meek]

Jon, is that an "adjustable" tweeter X-over in picture #4? :T

[Ten 99]

David,

I think you're referring to the adjustable LPad (the potentiometer looking thing). That so that a person can dial in the appropriate sounding level of the tweeter. Once that level is found, then Jon will take the adjustable LPad out of the circuit, and measure it to make a permanent LPad. At least that's how I understand it to work.


Jon,

When I see those Arvo's, I immediately start thinking about an asian look. I actually made some sketches a while back (pen and paper) that give these a very japanese architectural look. I wish I still had my scanner, but I lent it (more or less permanently) to my dad, as he loves scanning in old photos with it, and I almost never used it. I'll see if I can scan it in later this week and upload the pics. It really takes some of the hard edges off of it, while mainting the look and the function. As for veneers, I think I would use a decently figured maple on the baffle, and probably use some Walnut or other darker figured wood on the Sides and supporting portions. Everyone's tastes are different though.

[JonMarsh]

Chris has got the right idea about the trebl crossover, David. Once I get it where I think it is voiced "right" (which isn't necessarily dead flat on axis; depends on how the overall power response looks, and may need to create a diffraction waveguide around the tweeter before finalizing), I'll remove the LPAD, check the resistor values, then dial in something similar with a Millls resistors array, which has lower inductance, and doesn't have a contact to worry about (deterioration over time).


Interesting ideas about the esthetics, Chris. I'm open to any ideas you'd care to post. The very first Arvo "test mule" I built was more finished, but I'd hardly call them elegent; the front panels ultimately may not have exposed wood, but have a light wire or metal tube fram grille with a very open weave mesh grille cloth. I'd be interseted in looking at anything you have time to post.


The original looked like this, and I have been thinking in terms of similar esthetic, but I'm quite open to ideas....


http://audioworx.virtualave.net/APDrivers.jpg

~Jon

[Ten 99]

Please excuse my lack of artistic drawing capabilities. I wish I knew how to use some of the 3D modeling software, but the learning curve on those things requires far more than I want to invest.

This may not be the taste that everyone would like, but it's what I was seeing in my mind when I first saw the shape of the Arvo Speakers.

http://www.htguide.com/forum/attachment.php4?attachmentid=791&stc=1

[David Meek]

When you said "asian" the first thing I thought of was the water tori at Miyajima, Japan

http://www.bbbn.jp/~kokusai7/kokusai/pics/miyajima.jpg

[Ten 99]

David,

It bares a striking resemblence.

Tori Gate (http://phototravels.net/japan/pcd3860/miyajima-64.html)

I was thinking that the "roof" and sides would be a darker richer color, and that the baffle would be a lighter color, like a figured maple. Tiger stripe would be nice. That rear portion of the cabinet holding the subs could be a flat black or other dark but not attention grasping color. The base or floor could be also be black, or maybe even a granite looking Corian or the like?

If these are the masterpieces that Jon and Thomas are striving for (regards to sound), I would hope that they might eventually get that ever-deserving exterior as well. No Sears vinyl siding please. :rofl:

Oh, I suppose you could also "veneer" that baffle with something like a japanese hanging scroll, but that might be a bit over the top on that theme. If you could do a mirror or see thru baffle, it would really give the illusion of a Tori gate.

[Hank]

You know I'm a wood guy, and veneer all my cabinets, so this is going to be a mild shock: metal, well, part metal-look any way. Since so much good veneer would be wasted with all the driver cutouts on the baffle, I'm thinking about laminating the front baffle with WilsonArt's Metalaminate brand Pearl, which is a titanium-looking laminate: http://www.wilsonart.com/laminate/consumer/products/listcategoriesandproducts.asp (Go to Category and select METALAMINATES)
Then do the sides with either solid hardwood or MDF with rounded-over vertical corners and veneer with your heart's desire exotic veneer.
Now that you've pulled yourselves up off the floor and are revived from shock, think it over. ;)

[JonMarsh]

Interesting... It's hard to get an idea of what it would really look like from the web pictures, even the enlarged ones.

But I'll probably order some samples to check out.

Thanks, Hank.

All inputs appreciated.

~Jon

[Andrew Pratt]

Did you get my email jon?

[Hank]

I've ordered several WilsonArt samples and encourage you to if you're considering the laminate. I'm looking at my Pearl and Champagne samples now. I've been thinking for a long time that the Pearl contrasted with a light hardwood or veneer would be a nice modern classy look.

[Eric S]

Hi Guys! This looks like an interesting speaker project. I've been looking up various parts to ballpark a construction cost of these, but I can't seem to find anyone who sells the TC sounds TC+ driver. Where can I find them?

Thanks!
Eric

[ThomasW]

www.oaudio.com

[Eric S]

Thanks for the link, Thomas!

Looks like the TC driver shares some interesting appearance and performance specs with my Audiomobile Mass 12... Very nice line-up of drivers!

Eric

[sfdoddsy]

Jon,

Do you think ti would be worth swapping out my DPL12s for the TC2s? I assume I'd get a bit more output. I'm in the States on holiday and the price I could sell my DPL12s for back in Australia would make the whole exercise price neutral.

Cheers

Steve

[ThomasW]

Steve,

Jon's in route from Singapore back to Calif at this time...........

IMO If you can break even then yes do it. .

[sfdoddsy]

Hmmm, on second thoughts I might wait a bit. I'll be back in the US around Xmas by which time these fancy new drivers in the other thread should be out. Anyone know if they have embedded leads to stop slap?

Cheers
Steve

[ThomasW]

I have a couple of pics that partially show the leads and they are not woven into the spider. I do not have pics of the spiders themselves

[JonMarsh]

Wooo Hooo! Got my replacement tweeters in from Madisound, so it's CLIO testing and crossover tweaking this coming weekend! :B :B :B


~Jon

[Bing Fung]

Hey Jon, do you have a finalize parts list for this kit yet, or is it still R&D?

Sorry for my ignorance, are the backs of the speakers to be enclosed or is it a dipole design?

It's looking very good, and I would certainly be intrested in a set of these... Just because they look so darn cool :T

[JonMarsh]

Hi Bing!

Well, I wouldn't say "finalized" yet, things are in the crossover tweaking stage, don't know if there will be any big changes, but don't really think so. The driver complement is probably fixed, now, based on what I've been hearing. This is a dipole desgin in bass and midrange, up to 1250 Hz nominal.

The response from about 300 up sounds pretty nice, so I'm not expecting to see any changes in that region; just probably fine tuning the woofer to midwoofer crossover.

Stay tuned...

~Jon

[Bing Fung]

Jon,, waiting with baited breath ;x(

Whats the typical power requirements of this array going to be (min/contin/max)?

So could a person set up the Avro with a Bi-amp Active crossover design?

I suppose the passive crossovers will be the heart and soul of the Avro? ;)

Will a person be able to use those Dayton PE Reference drivers as well, if testing in the M8A goes well?

[JonMarsh]

Hi Bing,

So far we've made two versions of the MkII Arvo- ThomasW's are biamped between the 12" woofers and the 8" midwoofers, and use a Behringer EQ unit to take care of the dipole bass EQ, whereas my goal all along was a pure passive crossover, in combination with a small EQ unit inline- either passive, or passive buffered. This is necessary to provide the 6-8 dB of lift below 80 Hz that the design requires.

If you've ever seen Sigfried Linkwitz's site, and his Phoenix or Orion projects, they use pretty complex full active crossovers with a lot of EQ sections, and require tri-amplification. Since I wanted to mess with these with some amplifier designs I'm playing with, I wanted a pretty much full range speaker using just a passive crossover. But it took a fair amount of thought to come up with a driver/panel configuration and crossover frequencies amenable to a passive approach without having an absolutely hideous number of components, but which still deals with the various issues around a dipole configuration.

Because they're dipole, with some passive EQ in the crossovers, they aren't as sensitive as a box speaker with the same drivers. OTOH, becuase of the drivers chosen, and becuase they're 4 ohms over most of the spectrum, they do pull a little extra juice, which helps. Probably the minimum I'd recommend is 75 watts per channel, (8 0hms), and my testing has been done with 8008 Aragon's. (200W rated per channel, 8 ohms) I think they run pretty well with those, though after I get them dialed in I may haul out one of my sets of Aragon Palladium monoblocks- which I'm pretty sure would be MORE than enough under any forseeable circumstance- 125 watts class A, and clipping at about 660-600 watts at 8 ohms (about 1 kW at 4 ohms). Realistically, they'd be overkill, but then, they'd be cruising pretty easy even putting out 100 to 200 watts.

I don't think these are something I'd recommend for a receiver, unless you were using a 80-100 Hz crossover to a sub- in which case, though, you'd be kind of wasting the potential of the TC2+ 12" woofers. ThomasW is using DPL12's in his biamped versions, but I understand he's going to upgrade to the TC2+ also. It's a very nice driver, with a good balance of sensitivity, power handling, Xmax, very low distortion, low rear pumping noise (a function of vent and basket design, important to open dipoles), and relatively extended top end- to over 600 Hz.

As a full range dipole, they won't do Pod race kind of infrasonic stuff, but they should be OK down to the low 30's or high 20's, which is my goal for music. I may add a sub below 40-50 Hz, using my Sumo Delilah MkII crossover, and my unbalanced input Palladiums. I've got an HE-15 and some BPD1203's around. ;) Just don't have enough time to build all my ideas as quick as I'd like! :roll:

The M8a's are flatter overall in the 200 Hz to 1800 Hz range than the Dayton's. The Daytons have more Xmax, and lower distortion below 250 Hz. But in the Arvo's, they're not doing much below 250 Hz. So, I'm not sure they'd make much sense, considering the upper crossover point may be more of a problem with the Daytons. But, I guess, never say never, until you measure. And listen. Since I'm not building these commercially, I don't have a lot of incentive to try out a lot of variations on the theme. But, it might not be too hard to pop a set of the Dayton's into the next test panel just as an experiment and compare measurements for the intended passband. Who knows? :confused:

~Jon

[Bing Fung]

Hey Jon, thanks! I gleaned most of that when I reread the entire thread.

So will a person be able to make a center channel to match these, or should I be looking at the M8 project instead?

My long term goal would be to have a serious set of speakers geared for music, however pull duty in theater mode as well.

These seem like a whole sytem design with a sub that will be designed to be mated to them.

[JonMarsh]

For now, when I go 5.1, I use a phantom center. Becuase of the room independence of the dipoles (not nearly as much side interaction), the intelligibility factor is higher than for box speakers, and for me this works well.

ThomasW and I have done some quasi-dipole center channel designs, as well as some more conventional W (MT stacked) W wide center channels.

In principle, there's no reason I can see that we couldn't do a center based on this concept, but with 12" woofers, it's wouldn't exactly be "tiny" or slim. :E

I'm thinking a shallow U frame, 12's flanking at the outside, or maybe Dayton Reference series 10" woofers, to save space and money; a Dayton 7" midwoofer on the midrange, and SS9800 for the tweeter chores. This would NOT be a down to 30 Hz dipole, more like a 50 Hz design, if it uses the 10's, but that would work fine in an HT setup with 80 Hz crossover to subs (have to have response in center and mains extending at least 2/3 of octave below the crossover point, in my opinion.

There's two, maybe three possible projects for the Dayton RS270, and it's attractive becasue of the sensitivity (90.5 dB) with good Xmax (almost 7 mm), and very flat response to 1kHz (a dream to cross over at 250 Hz to midwoofers, or 600 Hz to RD ribbons).

I'm also considering these for my "small form factor" upgrade of my old X1 clones; bass bins on those were just impractical in my home- the 7" Eton midwoofers will be replaced by Dayton RS180's (already on order), and I'm considering new bass modules and a new crossover using the Dayton 10's. Will be another big honking passive crossover, I'm afraid. ;) I'd like to retire the active one which I used for the early version of this system. Get rid of all those opamps in the signal path. :B

Gotta do some modeling on this- may try to do it with the new LSPCAD 6 demo, but I have to admit I'm finding the new interface and paradigm to be slow going... makes me feel a little "stoopid". :rofl:



These seem like a whole sytem design with a sub that will be designed to be mated to them.

Yes, that's exactly the idea- but one that wouldn't require multiple sets of tri-amp channels. I've done the tri-amp thing in the past, and maybe my thinking is colored too much by commercial considerations (a friend of mine in the biz thinks a tri-amped dipole is an impractical commercial product, too, unless you integrated all the electronics- which would remove too much of customer choice).

I'll certainly keep you folks updated as this continues to develop.

The interesting question in my mind, at this point, is will I be able to make the next dipole project, based on the RD50's I have on hand, as good as or better than I think the Arvo MkII is turning out? It's going to be a challenge.

~Jon

[Bing Fung]

OK, I almost spewed coffee all over the PC when I read the size of the Centre channel drivers... :rofl: Really , what did I think they would look like or use...?

Sometimes I tell you, I don't see the forest for the trees :loser:

Maybe this is a case of the Avros making my eyes bigger than my planned space and and goals... :scratchhead:

A smaller form factor and Ribbon based speaker sounds awesome... So, ... What are tha chances of having you drop everything and getting to work on something like that ASAP? :W

I can see moving the complexities of a tri-amp system to a passive XO & EQ'd as being a nicer less complex solution. These Avros have more engineering in them than I fully realized.

[JonMarsh]

A smaller form factor and Ribbon based speaker sounds awesome... So, ... What are tha chances of having you drop everything and getting to work on something like that ASAP? :W



Uh, again, I think we have a slight failure to communicate, which as usual is probably due to my tendency to assume-

The Bohlender-Graebner ribbons are kinda big beasties- RD50's are 50" tall, about 6" wide; the RD75's, as used in ThomasW's quad amped living room system, are 75" tall. Hence, the model numbers. ThomasW is a big guy, so he needs the 75's; me, I figure I can get by with the 50's, though there's a reason for that choice, in the overall configuration I planned on having a woofer setup underneath.

Now, my current plan is using the RD50s, which are (optimistically, I think) rated for 150 Hz to 18 kHz only from 600Hz to 6 kHz; we've found they work very, VERY well in that region. (that's how I got Thomas to try them out in his big family room system, shoe-horned in between the Acoustat 1+1 and the ribbon tweeter panels). They're a great midrange driver, and this avoids the area where you get the front cavity resonance that requires a notch filter and limits the lateral HF dispersion.

Above 6 kHz, I plan to use a stack of the Fountek JP2 5" true ribbon tweeters. Somewhere around 8 per side, as they're 6" tall, depending on what I think makes sense for a vertical listening window, between seated and standing, and matching up to the line source of the RD50.

From 600 Hz down to 150 or so, it's up in the air- candidates include the Adire Extremis XBL2 6" driver with 12 mm Xmax (got 8 on order for eval), and possibly larger drivers like the Dayton RS270 (10"), or even the HiVi M12, which is also quite solid up to 1 kHz. If 10's, probably 6 per side. If 12's, 4 per side. Hopefully a passive crossover design between whatever midwoofer choice I settle on and the RD50's and ribbon array.

Sitting underneath all this may be a dual 12" dipole woofer; TC2+l, or possibly the new 12's from Chad's company. That will get the top of the line array at about the 6" level. No need to have tweeters and midrange 6" off the ground, huh? ;)

When I was considering a monopole line soruce, the woofer was going to be a single BPD1203. I expect my 1203's to wind up in Aerial Stryke Klone systems, a clone of the Aerial SW-12.

As you can tell, this ISN'T really a "petite" system. :B And yes, this will be the last speaker system I plan to build for a while.... certainly the last large one, anyway!

I've got a few simulations and measurements to do before I decide which way to take this.

Now, if Arvo MkII's are too big for some folks, we could probably downsize them slightly, with something that looks more like the MkI (shown above; uses the Titanic 10" woofers (really an 11"), but, say, with 7" Dayton RS160's. Might even be able to get by with the NorthCreek D28, which would save a few more bucks. (about $200). It would be a couple of dB less sensitive than the MkII, though, and not that much less expensive to build (maybe $300-$400 less for the pair). Crossover, you see.


Now, if you guys could take up a little collection for me, so I could retire early, just think of all the designs I'd be able to crank out for you guys to try out or use! :rofl:

~Jon

[Bing Fung]

Ahh, when I seen RD50, I just thought Raven Ribbon Tweeter, such as...

http://www.zalytron.com/pictures/raven_2_11.jpg

That ribbon system you have planned is Waay outta my leauge.

Smaller Avros you say?

[ThomasW]

By the time Jon gets that system built he'll be so old his hearing won't go above 6kHz ....... No need for a $pendy line array of high-end leaf tweeters then ..... :B

Also 'downsize' and Arvo are perhaps mutually exclusive terms. Unless one's idea of downsize is a little under 48" tall

[JonMarsh]

By the time Jon gets that system built he'll be so old his hearing won't go above 6kHz ....... No need for a $pendy line array of high-end leaf tweeters then ..... :B

Also 'downsize' and Arvo are perhaps mutually exclusive terms. Unless one's idea of downsize is a little under 48" tall

A little under 48" tall? Reminds me of some of the women I see in Singapore...

Oops, wrong topic.

Nah, I think Thomas has got this all wrong. I have a complete Microsoft Project file which proves I'll get the ribbon array speaker done while I still have hearing to 10.21 kHz! Gauranteed! ;) It's ThomasW that won't have hearing beyond 6 kHz by the time I get them finished! :B

I am his junior, you know, even if I rarely respect my elders! :rofl:

Yeah, that's why we get more and more focussed on the LF performance of these speakers, because even if we can't hear, we'll still be able to feel them.

Now, if only women were that way in our doddering old age.... that is, not being able to hear them, but still being able/having opportunity to feel them. Ah, wouldn't that be bliss!

~Jon

[JonMarsh]

Ahh, when I seen RD50, I just thought Raven Ribbon Tweeter, such as...

http://www.zalytron.com/pictures/raven_2_11.jpg

That ribbon system you have planned is Waay outta my leauge.

Smaller Avros you say?


Those Raven ribbons cost about the same as those RD50's! But they are more petite!

As to smaller Arvo's, good question. Another combo that occured to me would be the HiVi SP10 woofer, combined with the RS160 Daytons. Just don't know how small it's practical to make them without messing up some of the good points. But hey, what's life without a challenge, right?


~Jon

[Ten 99]

Bing,

Have you been to this (http://www.salksound.com/) website? This gentleman builds some very nice looking cabinets, and uses a lot of ribbon tweets.

For example:
http://www.salksound.com/gallery/HT3-bubinga.jpg

http://www.salksound.com/gallery/ht1-cherrywalnut.jpg

He also just came out with a 5 channel setup, where he uses 5 MTM cabinets with ribbons. The center channel is laying on it's side, with the orientation of the tweeter rotated 90 degrees as well. There was talk about making them available as kits for DIY, but don't know if he has done that or not. I've heard that vertical dispersion is a problem with ribbons like this, and that unless your ears stay in a pretty much fixed vertical plane, that you lose a lot of detail when you go above or below the plane. I believe that's why so many folks build LA types of setups with ribbons. Here's a picture of them:

http://www.salksound.com/gallery/HTS-5.jpg

BTW, I'm sorry to divert off of the original topic of the Arvo's, which I am much more interested in reading about, but I just thought I would point Bing to that in case he was interested in reading up on them.

[JonMarsh]

Jim Salk does some very nice cabinet work, and I do like his offering crossover options for the HTS series with different baffle step depending on the intended speaker location- we've done the same thing with versions of the M8 two say.

I'm curious how the combination of the Vifa M18 midwoofers and ribbon tweeters works together; to stay clear of the lower resonance in the ribbon, you usually have to cross true ribbons over fairly high, and that requires the midwoofers to work to a fairly high frequency- one above the pistonic cone range, and one above the ideal point for center to center distance for driver crossover. The latter can be approximated by dividing 13568 by the crossover frequency, the result in inches is the maximum spacing to avoid half wave beaming/interference in the crossover region. This is once reason it's hard to due MTM's well, without some of the forward, hard quality Dennis Murphy complains about in most MTM designs.

I haven't heard the G2si, but I'd be concerned about running it low on crossover, given the distortion test results Mark K got with it. It doesn't compare to a ScanSpeak dome or Millenium Excel. Ribbon or not.

For example, for a crossover frequency of 3 kHz, the max center to center distance calculates as 4.5". Doable with a 5" MT two way, but not an MTM with 7" woofers and a tweeter in between.

That's one reason I think the M8 design works out pretty well, even the MTM variants, because at 1250 Hz you've got about 11 inches to play with for the center to center spacing. That's doable with an 8" MTM, with an offset tweeter like we used in Tibor's system or the Arvo's.

Jim's HTS speakers are priced pretty attractively for pre-built systems; a pair of the MTM's will set you back just $1199 assembled, or 899 in kit form with pre-built cabinets, but another $90 for routed front panels (these appear to be some of the very nice PE cabinets). So, about a G note for a set requiring assembly but no wood working.

Hmmm; I wonder how an MTM with dual RS160's and a D28 Northcreek tweeter would compare? Could cross those over easily at 2 kHz; OTOH, with a Millenium Excel tweeter, you could cross at 1400 Hz, which would allow 9.7" center to center spacing... yeah, that's sounding pretty good. I bet the off axis power response would be pretty flat overall, both horizontally and vertically. Price of materials should be similar. Hehe.

OK, someone take this idea and run with it. I can't design every reasonable idea I can think of! ;)

[Hank]

Jon: Wow, that RD50 project appears to be on the verge of getting a bit rich. Keep in mind those of us without bottomless pockets. ;)
I know it will be quite the interesting design - maybe for Fall '05?

[Dennis H]

I think Jim Salk is crossing at 2500, 4th order so he just barely makes the C-C criterion with the W18EX. In the 3-way, he crosses in the 10" TC woofer at 300. I'll bet it sounds pretty good.

[JonMarsh]

Dennis,

I've heard a lot of good things about his "Veracity" system, and I expect the HT3 version sounds very nice- the design and workmanship are exceptional compared with even a lot of "name" companies like Aerial and Joeseph Audio. Do you know anyone who's heard the full range version?


Well, Hank, my pocket's don't feel very "bottomless", which is why I'm looking carefully at some parts that some would consider "budget" grade, like the Fountek ribbons and the Dayton Reference series woofers.

Of course, it would probably be quite feasible to do a simpler version running the RD50 to it's upper limits with the standard notch filter at 6 kHz- I might try that myself at first. And one could probably get pretty good results with one of the dipole midwoofer solutions I'm considering, just crossed over around 75 Hz or so to a conventional monopole sub. I'm thinking dipole on the low bass partly for sound, partly to cut the off axis (including floor) radiation.

Do you have a vote to register? Extremis midwoofers? Dayton RS270 10" woofers? M12? I probably won't try the M12's unless I don't like the first two alternatives, or can't get the level balance. But since the RD's are pretty much resistive above 150 Hz, I figure I can even pad it down a little if I have to.

In the midrange/mid bass, an array of 6 RS270's properly configured should be ~ 96 dB for 2.83 volts, 1 meter, which gives me some room for passive EQ - almost 8 dB of dipole equalization, running the RD's flat out at 88 dB. It would be like having a first order passive roll off at 60 Hz, then kicking in a zero around 120 Hz, then the ~250 Hz notch to equlize out the bump in the forward response, and transitioning to a 3rd or 4th order transistion network at 600 Hz. I've demostrated the feasibility in LspCAD already, so I'm hoping in the real world it will work, too. Isn't actually as complicated as it sounds.

Think of it as a narrower range Amazing Carver speaker, perhaps. Maybe the stuff below that should be bi-amped.

[Bing Fung]

Ten, thanks for the link, I'll have a look :T

How did you know I'm a sucker for ribbon tweeters? :huh: ;)

[Dennis H]

Jon, I don't know anyone who has heard the Veracity 3-way. FWIW, there's a glowing review on his website. ;) Dennis Murphy, by all accounts, has a very good ear for voicing, backed up by measurements, and the components are top-notch so I imagine it sounds very good. And those cabinets are real works of art - in a guy sense - I suppose most spouses would want to cover them with lace thingies or paint them purple or something. ;)

[sfdoddsy]

Bing,

I have similar speakers to Arvos and have also built centers and surrounds to match:

http://doddsy.net/blackbob3.jpg

They match very well, but of course the center is really only suitable for front projection systems. I cross my centers and surrounds over to a monopole sub at 80Hz, but you could get more volume or go a bit deeper by doubling up the eight inch midrange, in effect making it just the MTM (or TMM)of the Arvos.

Steve

[JonMarsh]

Jon, I don't know anyone who has heard the Veracity 3-way. FWIW, there's a glowing review on his website. ;) Dennis Murphy, by all accounts, has a very good ear for voicing, backed up by measurements, and the components are top-notch so I imagine it sounds very good. And those cabinets are real works of art - in a guy sense - I suppose most spouses would want to cover them with lace thingies or paint them purple or something. ;)


Yeah, I just love seeing lace thingies on my hardwood speakers... ;)


I'm sure Dennis did an excellent job on the crossover design and voicing.

But I get an attack of "marketing-itis" when I read the prose put in on the web site:

Only the highest quality crossover components are used in the Veracity HT3's. Yes, they are expensive. But it makes no sense to compromise the capability of this exceptional crossover design. Rest assured, with quality crossover components, our crossovers are a perfect match for the quality of the drivers in the HT3 and provide the utmost in transparency.

Oh yeah? What are the highest quality components? Which ones? Why do you think they're the highest quality components? He sounds like some of our product marketing guys where I work, that just drive me crazy. Lots of claims, no specifics.

Oddly, though the MTM speakers are offered in three crossover variations depending on location and required baffle step compensation (a very good idea in my opinion - for example, I've heard the Usher 2 way design Dennis did, and they clearly weren't fully BSC- but the guy who built them thought they were, and was trying to use them that way), there's no mention of this, or optimum room placement requirements. Jim could do a more informative job on his site, me thinks.

Still, the cabinets are lovely. Especially without the lace. ;)

~Jon

[Al Garay]

I'm confident that the Veracity HT3's use Sonicap capacitors and Alpha-core inductors and Mills resistors. Hiqh quality yes. Highest quality assumes too much.

[JonMarsh]

I'm confident that the Veracity HT3's use Sonicap capacitors and Alpha-core inductors and Mills resistors. Hiqh quality yes. Highest quality assumes too much.

I have to second you on that, Al. I think the Sonicaps are very good as metalized caps go, but they're not the last word for crossover caps.

Now, I actually don't like the Alpha core inductors, because for their claimed equivalent guage, they have higher DCR than an equivalent wire wound inductor. They're also a pain to attach to (I built a kit crossover for a friend which used them). Lastly, within the audio band, the skin effect comparison between AWG 14 or 15 wire and the Alpha foil shows no meaningul benefit, especially considering the main effect would be slight impedance increase at very high frequencies- for shunt inductors in tweeter networks, no plausible value which I can see. I posted some calculations about this once before, based on the physics- I can do so again if anyone's bored and curious. ;)

Now, I'm a power conversion engineer, and do use foil windings in transformers, mostly for form factor reasons, but also to reduce proximity effect and skin losses at high switching frequencies- we're talking 100 - 200 kHz pulse frequencies, with harmonics extending to 10's of MHz. Not audio by any stretch. :B

My favorite inductors are the big honkin' AWG 12 jobbies which Geroge at NorthCreek can make; I can't do ones that big on my coil winder. They're great for series pass, and for shunt applications. DCR is critical to determining ulimmate low frequency attenuation at the tweeter.

~Jon

[Ten 99]

http://www.cell2000.net/~users/Ryan/MUSCLES.JPG

Hey Butthead, he said ...


Yeah, I just love seeing lace thingies on my hardwood

~Jon

[Dennis H]

Jon,

Yeah, I've wondered how the commercial guys deal with the BSC thing. You never know how someone is going to place them in the room. I think Dennis M. has said he does something less than full BSC, at least on his little 2-ways, because he assumes most people will have them sorta close to the walls. Maybe something like the old 70's style trim pots or switches are the way to go. Revel has a bunch of back-panel adjustments on their big speakers but I don't know if BSC is one of them.

[JonMarsh]

...with Mark Knofler and James Taylor.

The hybrid Arvo's are playting tunes again in ThomasW's living room, with TC2+ replacing the DPL12's, after sorting out a few glitches with frequency setting modules for the electronic crososvers. When these things happen, I'm reminded of why I'm doing the passive version... ;)

This set has one passive crossover point (between tweeters and midwoofers) and two active points, from Midwoofers to dipole woofers, and then to IB.


http://audioworx.virtualave.net/TWArvoTC2ASS.jpg

The Behringer EQ is just being used on the sub output.


http://audioworx.virtualave.net/TWArvoTC2BSS.jpg

Been doing some tweaking, now I'm going to hook up the laptop and see how well my ears and the test gear correlate. ;)

We've already "upgraded" from the Sony XA7ES to outboad DAC on the CD; still not running the "A" team on hardware, aside from one of the Ayres, but it definitely has that "you are there" thing I like with dipoles; kind of like big headphones. :B When it sounds like this, I don't worry too much about measurements. ;)

This setup is running a Sony E90ES FET preamp, a Marchand electronic crossover in three way mode, an Ayre V5 on the top end (175 Hz up), an Aragon 8008 on the dipole woofers (TC2+), and an Adcom 565 on the IB sub. In this system configuration, the TC2's are little more than "filler" drivers, as they're in at 175 Hz on the top, and out at 75 Hz on the bottom. In my own, they cover 200 Hz down to below 40 Hz. They seem to have a good operating range of ~ 400 Hz on down. I don't think I'd try to run them up beyond 600 Hz, so for adquate overalp, I'm investigating the RS270's first for the Saint-Saen.

~Jon

[Dennis H]

Lookin' good guys. I especially like the AV rack disguised to look like a fireplace. ;)

Seriously though, can you describe the process you went through to pick the 75 Hz XO? Is it because Thomas's IB is clean up to there and you worked from the bottom up so to speak?

[JonMarsh]

Yeah, that fireplace cum AV nook is just the setup for exhausing the heatwaves coming off of ThomasW's 36" Toshiba! And Thomas did a nice job with incorporating the fireplace hearth into a convenient platform for Balanced power units!

Actually, most of the gear is going to end up down in the basement, except for the preamp and CD player, and I have it on reliable authority that Thomas is already saving up his quarters for another Ayre K5x preamp for this system...

The IB sub crossover point is sort of a bottom up thing, but also considering the need to allocate SOME useful bandwidth to the TC2+. When Thomas was last running this system, the IB was NOT hooked in, and ThomasW was running the "Ultra" version to get a more ruebenesque bottom end. Now, the IB is hooked back up. It works OK to beyond 100 Hz, but as is typically the case, the manifold starts entering into the equation regarding the frequency response. My recommendation is to run IB's with manifolds crossed over at 60 Hz to 75Hz. At 75 Hz crossover, the 12's in the Arvo don't need any additional Eq to cope, so this sort of dictates the ideal crossover point from the upper side. So, 75Hz it is. And yeah, we're kind of under utilizing the TC2+ in Thomas's impelmenation- I think we could get by just fine with a couple of pairs of Extremis 6 on the 75Hz to 200 Hz area... ;) The system has a lot of "ease" - very low IM, of course, and piano and voice sounds great- which is the first thing I listen too in a new speaker.

My Arvo's will be run full range from one passive crossover, with a outboard LF EQ box that uses a non feedback buffer amp and Eq below 75 Hz. It will require a somewhat more powerful amp, but it will have a lot less cables... with what I save on cables I can spend on amps!

Actually, that's not quite true, as I have two sets of Palladiums... ;)

Hearing Thomas's again is spurring me to finish my own. OTOH, I just got some additional veneer and tools for finishing the M8ta's.

I just need a month off to get all of this done!!

~Jon

[Hank]

Well, well, something more than contemplating the stress-relieving medicinal side of tequila was accomplished last week! Yes, seems like almost a waste of the TC2+'s capability, but what the heck, progress was made. Thomas is obviously in for some fine music sessions.

[mante]

"You know, it's been a little painful doing this work on these other projects, and looking over and seeing the Arvo's in the corner, waiting for me to put new tweeters in and test..."

~Jon

Here I thought you were just waiting for the Avalanche 12s to try. I have this feeling when you get done with the tweeters, you will want to try the RS180 mid-woofer.....it never ends, does it. ;)

Enjoy,

Craig (Young Padawan, Mante)

[JonMarsh]

You HAVE identified one of the serious perils of speaker development. But actually, I'm pretty happy with what the M8a's do in this system- the range above 300 Hz is pretty dang pleasing, I think I got lucky with my simulations and hit just the right baffle width and offset.

The Avalanche 12 is something I'm very interested in for the Saint-Saens, but I hadn't really given thought to using those in this system until you mentioned it...

Since I'm trying for a mostly passive design, managing the sensitivity, Q, and all is critical. I don't know that the Avalanche 12's would work (they'd have the throw, for sure), they would definitely need some bottom end EQ due to the low Qts.

ThomasW's use an active crossover for the mid to woof crossover, and an IB below 75 Hz (we joke that his TC2+ are the worlds biggest, most expensive filler drivers, as they only cover 75Hz to 175 Hz in his system).

I'm just trying to optimize the passive version in comparison, and if possible reduce the need for an active/passive line EQ below 75 Hz, though I've got the bass shelving EQ design already done... both passive and active versions.

Now, if it was only for me, I'd just use whatever VC arrangement I want, and if it's a two ohm load in the bass, my Aragon's don't care... but most amps in the world besides Aragon and Krell DO care. ;) And wiring them in series would sacrifice some sensitivity.

I still have some experiments to do, I'm afriad... but getting close. Maybe we just put a second amp on the additional driver, with paralleled crossovers, for the more "conventional" amplifiers. I'll give this some thought... and try to forget for now your suggestion about the Avalanche in the Arvo... too many changes! ;)

[JoshK]

This is a fantastic project. I have been reading and contemplating building one of the various dipole designs (orion, pheonix, bob, NaO, etc) but hadn't come accross this one yet. This is close to what I had envisioned in my mind, with the exception that I am thinking DEQX (although I haven't bought one yet) and not passive xo.

I have a couple of questions if you guys don't mind. If you were to use active xo (be it digital or analog) would you still choose the HiVi M8a? Or would you want to use the Excel 8"? It seems you have ruled out the Dayton RS series 8" for this ap. I am trying to hone in on driver selection and am wondering your guys' thoughts.

Another is tweeter selection. The Excel millenium is a great tweeter, I've heard it before in Focus audio speaks and lot it for what I can tell.

The TC2+ 12" is pretty much settled on.

I have a really difficult room acoustically to deal with and am hoping the dipole approach will help me get my system sounding bearable again.


(I wasn't famililar with this forum until just today, so this is my first post)

[RonS]

Hi Josh,

First, welcome to the HTGuide. As to your questions, I'm hardly one of the resident guru's. However, I am building a dipole speaker using the HiVi M8A drivers as midranges, and I'm doing it with a Behringer DCX2496 digital crossover. The tweeter is the Vifa XT25 (I'm also going to be testing the Seas 27TDFC of which I have a pair on hand). The woofers are currently the Madisound 1252DVC, which I will eventually replace as funds allow (probably the TC2+, but will also look at Ascendant and maybe Peerless).

You can find a bit of info about a unique mounting system that I developed for the main panel at http://www.htguide.com/forum/showthread.php4?t=9579&highlight=orion+mounting.

I found working with the M8A and the digital crossover to be very easy. I'm using 8th order L-R for both lowpass and highpass, at 100Hz and 1.24kHz respectively. I added a parametric eq at 400hz, and a 6dB/octave shelving filter from 300Hz down to account for the dipole cancellation. Currently I've got the on axis response to be within +/-1.5dB from 100-10kHz, excluding a dip at 2kHz that I have to investigate. The off axis is pretty good on the M8A, but I need to add some felt padding around the tweeter to nock back the off axis response at it's bottom end.

Good luck on your project!

Ron

http://www.htguide.com/forum/attachment.php4?attachmentid=2411&stc=1

[ThomasW]

Hi Josh and welcome to HT-Guide ....

Jon's traveling so I'll chime in.

The design as you see it is finished. But Jon's trying to decide whether to run only one VC in each TC2+. His recent distortion testing shows a slightly lower distortion running one VC vs two. This possible change obviously only effects people running passive XO's. Here's the plot so you can see what the differences are......

http://pwp.t-3.cc/ttriff/TC2+distortion.jpg

Note that many other drivers were considered for this speaker and eliminated for one reason or another.

Yes the Excel millenium is a great tweeter. But there are products in the pipeline that are going to be pretty stiff competition for about 1/3rd the price.

In the far, far, far, far distant future, there will probably be either an upgraded Arvo, or a design with an identical layout but using different drivers. At this time it would be premature to discuss it because it's literally just a concept. Some of the drivers being considered aren't yet available at a retail level...... So it's vaporware for all intents and purposes.

[goskers]

Hmmm, I think the new tweet from PE would be the cloaked item. All the RS series has really made some great affordable designs a reality. :T

Oh yes, welcome to the club Josh.

[JonMarsh]

I hope to be tweaking the crossover and trying out the RS28A in the Arvo this weekend.

Still mucking about with the low frequency setup, as Thomas points out- for an active crossover approach, that would be easy to play with, but I don't want a tri-amped system, and have been specifically working on this speaker as a test bed for my next amplifier projects. There will be either a passive in line LF EQ or a buffered version using BUF03's; I've designed both (final EQ pending VC usage), but haven't finished the PCB layout for the latter. Got sidetracked at bit with these HT oriented projects (The Modula MTM-IV-1 and center channel).

I hope to finalize the current crossover with the TC2+, the M8a, and the RS28a fairly soon- hopefully within the next couple of weeks.

As Thomas alludes to, there is also a new not yet available driver under consideration for the midwoofer; may just be a minor upgrade or tweak to the crossover, or a much more substantial crossover change. This driver offers the possibility of reduced midband harmonic distortion overy any other 8" I know of, and still moderate price, about double the M8a, but much less than a W22. But the proof in the pudding is in the eating, so until I've measured it, who knows? If it works well, it will also be used in the Saint-Saens, I think.

For those that want to spend the money, there's no reason not to use the W22, but once you talk about W22 and Millenium, why not just do an Orion? It's a proven design, fairly compact, and you might easily find someone within driving distance that has one you can audition.

OTOH, the Millenium Excel has considerably greater upper harmonic distortion than the RS28a (one man's "air" is another man's (and the test system's) distortion, and costs three times as much (I know whereof I speak; I have a set of those, too). The Peerless 10's are quite low distortion for the type of driver, but not lower than the TC2+, not even in the range between 100 Hz and 200 Hz, where many 12's fall apart) and without either comparable Xmax or Sd. The 12's are driven by a desire for more output without compromising the sonics.

There is another 12 which will be evaluated in June or thereabouts, having tested the 10" version....

Although I wouldn't say I'm trying to engineer the cost issues in the way a commercial designer might, I'm also trying to stay with best performance for the dollar, not the "most popular performance for a lot more dollars". ;)

That doesn't mean that others seeking to do something similar but with active crossovers shouldn't feel free to experiment- it's not like this is a production design department! I have enoguh of that in my day job... ;)

[Paul H]

... For those that want to spend the money, there's no reason not to use the W22, but once you talk about W22 and Millenium, why not just do an Orion? ...

Because that would be cheating ... :D :wink:

Paul

[JonMarsh]

Well, I suppose so... ;)

Instead of doing an "Ultra Arvo" like ThomasW did at one point, one could always do an "Uber Orion", say 4 10" XLS, or 12's... could even go to the MTM arrangement, which for higher output, if one used SL's crossover points (110 instead of 175), would be a good idea. But, then, you'd want to use the asymetric baffle arragement (flatter response), and before long, it's not looking much like an Orion.... :B

[JoshK]

Thanks for the responses and the warm welcomes!

I can completely respect your compromises, that is what designing is all about. I was asking to see if I understood them, more than calling them into question.

This looks like a very interesting project and one I just may build myself, although with active amping. I understand you not wanting to active amp, in my case I'd like to because I am building the amps now for such a purpose and want to try my hand at Fourier transforms (I studied math).

What tweeter did you use in the final Arvo Part? The SS tweeter or the Vifa XT? I think I saw a picture with either of them used.

[JonMarsh]

An early version used the XT25; the version built for ThomasW which has been operating (but not in pretty cabinets yet) for about a year uses the SS9800; mine have 9800, but will probably switch to Dayton RS28a. I didn't get around to testing the Dayton in this Sunday because the work on the Modula MTM HT type speaker took all day Sunday and still wasn't finished. Maybe some evening this week.

The RS28A will be out sometime late March or April; MarkK reports the Fs of the production version is lower, about 500 Hz, compared with the prototype at close to 1 kHz. The RS28a is a 28 millimeter tweeter, probably actually manufactured by Usher, with full copper Faraday shield on pole piece and pole plate, very low inductive rise, very low distortion. Dome and VC assebmly look quite similar to the SS9800, but larger. DLR did a nice tear down and photos, and also tested it with a modified difusser, using a hard felt ring on the outside of the grill, in place of the clear plastic disk on the inside of the grille.

http://www.speakerdesign.net/dayton_rs/tweeter/images/pe_rs_t_normal.jpg

http://www.speakerdesign.net/dayton_rs/tweeter/images/pe_rs_t_motor_2_small.jpg


http://www.speakerdesign.net/dayton_rs/tweeter/images/pe_rs_t_dome.jpg

Usher gets diaphrams and components from one of the same vendors Scanspeak uses, and it may be that it's the same supplier in this case. Response of this driver is quite smooth, and nonlinear distortion measured by MarkK was very low.

[JoshK]

Thanks for the info, looks like the tweeter to be on the look out for. Once I finish a number of electronics projects I have going I think I will purchase the Hi-Vi mids and probably by that time these tweets will be out. I have 4 10" woofs on hand from something else and I can knock out a prototype to see if I like the results.

*when I say knock out, I am trivializing how challenging this will probably be for me since this is my first real speaker and I haven't acquired the eq/xo yet, but I do have access to borrow some measurement equipment and have a lot of smarty pants friends who have built speakers.

[JonMarsh]

Thanks for the info, looks like the tweeter to be on the look out for. Once I finish a number of electronics projects I have going I think I will purchase the Hi-Vi mids and probably by that time these tweets will be out. I have 4 10" woofs on hand from something else and I can knock out a prototype to see if I like the results.

*when I say knock out, I am trivializing how challenging this will probably be for me since this is my first real speaker and I haven't acquired the eq/xo yet, but I do have access to borrow some measurement equipment and have a lot of smarty pants friends who have built speakers.


Then you should have a fun and entertaining time ahead of you!

And feel welcome to drop in and share your questions with us if you wish.

Obviously, the more you study on SL's site, the better prepared you'll be for your own project. With a little effort and persistence, I suspect you'll be rewarded with some pretty nice results. There's some real "magic" in a well working set of dipole speakers. Other's who have heard the Arvo's have been pretty pleased, to the point of suggesting we should be selling these on the Internet! Well, that's not really our interest; I'd rather see a more "Linux" like approach to speaker DIY, with shared info and projects, not proprietary kits. But, I understand why those who are trying to make some part of their living go the latter route- it's not an easy one in today's or yesterday's world.

Are you planning on using the Behringer DCX2496, or some other similar unit?

Welcome to HT Guide, and keep us updated on how your project is developing.

~Jon

[JoshK]

No, likely not the Behringer.

I have three viable options at this time for digital processing. #1 a friend is going to loan/sell me his Merlin DSP xo/eq, #2 sell my integrated amp and buy a DEQX with the funds, #3 build an audio PC (which I plan to anyway) and use BruteFIR and a Lynx L22 or similar.

I like the idea of the last option since it is fairly open source, like your Linux speakers. My training in mathematics had a lot to do with scientific computing and numerical implementation as well as dynamic modeling so hitting the books to "play around" with the filters would be a fun way to exercise my brain. However, there are issues with a PC approach that I am not sure I want to deal with now.

Anyway, I will definitely keep you guys posted. I have been browsing some of the threads and there looks to be some great info here, great guys and knowledgeable posters. Sometimes some of the other forums aren't like that.

[JonMarsh]

This is a fantastic project. I have been reading and contemplating building one of the various dipole designs (orion, pheonix, bob, NaO, etc) but hadn't come accross this one yet. This is close to what I had envisioned in my mind, with the exception that I am thinking DEQX (although I haven't bought one yet) and not passive xo.

I have a couple of questions if you guys don't mind. If you were to use active xo (be it digital or analog) would you still choose the HiVi M8a? Or would you want to use the Excel 8"? It seems you have ruled out the Dayton RS series 8" for this ap. I am trying to hone in on driver selection and am wondering your guys' thoughts.

Another is tweeter selection. The Excel millenium is a great tweeter, I've heard it before in Focus audio speaks and lot it for what I can tell.

The TC2+ 12" is pretty much settled on.

I have a really difficult room acoustically to deal with and am hoping the dipole approach will help me get my system sounding bearable again.


(I wasn't famililar with this forum until just today, so this is my first post)

I've been traveling a lot the last two weeks on business, and so missed some of these posts- just some more specific feedback for you, in case it helps.

Between the HiVi M8a and the W22, besides there being about a 1:3 difference in cost, the W22 has somewhat higher Xmax, but with the crossover frequencies I use (~175 Hz), Xmax for the dual 8's isn't a factor for the most part, not even when pushing it fairly hard. Both drivers get into issues with energy storage if you try to use them higher than 1400 Hz, even though the primary cone break up peak is higher on the Excel, so I recommend (and use) a steep crossover (LR-8 equivalent) at 1250 or 1200.

I'm planning to evaluate the Peerless Nomex exclusive series also, before finalizing this design; it may have some advantages in midband distortion without the Excel sticker shock.

Regarding tweeter, the Millenium Excel is a well regarded tweeter, easy to apply (it's very flat in overall response), and holds up well as regards distortion in the lower frequency range, down to 1 kHz. OTOH, I keep looking at the SEAS 27TDFC and wondering why there's a 5X difference in cost. The 27TDFC is NOT as flat, and requires a little contouring work to achieve the best "voicing" (a small inductor and additional resistor for the lpad circuit), but distortion performance is quite comparable, even down low at high playback levels.

http://audioworx.virtualave.net/SEAS_ContendersSS.jpg

Here's an overlay distortion measurement MarkK did- where ever you see red, that's the Millenium Excel distortion, which is higher than the 27TDFC.

http://www.htguide.com/forum/attachment.php4?attachmentid=1790&stc=1

At ~ $50, the Dayton RS28a will be the tweeter to beat, IMO, based on sheer performance, and value. It's measured performance puts it on the "A" list, near the upper end of that list; upper harmonic distortion is lower than the Excel. It's the first commercial tweeter I've seen under $100 that would give my Hales Transcendence tweeters a good run...

But a lot of folks do like the Millenium; it does sound pretty nice... but I suspect I'll wind up selling the pair I have rather than using them in a system. We'll see how the RS28's sound... I'm expecting some pretty good results, based on measurements. Maybe by NEXT weekend I'll know! ;)

[JonMarsh]

The danger to these modern simulation and design tools is that you can spend too much time wrapped up in doing what if's and variations on the design. OTOH, it gives one the opportunity to explore in detail a lot of possibilities, and with luck eliminate most of the clunkers before you build hardware.

I've just finished another round of modeling and crossover design for the Arvo- though by no means comprehensive, I think progresss is being made. For this latest effort, I went with the TC2+ using only one voice coil per driver, and wiring the two drivers in series. This changes the Qtc, and also lowers the net inductance per driver, though adding the two in series puts you back up to what one normally shows. This sacrifices some thermal power handling, but for a dipole, it's the mechanical limits that are the issue, not thermal below 100 Hz. This required a complete revamp of the crossover, and I took a little more care about the impedance curve, in the hopes that a relatively "normal" amp might drive these. (I've been thinking about picking up a "quality" "Budget" amplifier like the Marantz 95w/channel integrated that's so popular in Europe, as an "audio check" that designs work and sound well with something less than an Aragon or Ayre).

I've also used data from one of the prototype RS28a's to evaluate how easy that driver would integrate into the 1250 Hz LR-8 crossover using the cauer-elliptic method, and it was very easy to get it tuned to the exact response I wanted. At least on paper... er, digital bits, that is. :W

While at it I tweaked up the midrange a little more; just some slight adjustments improving the driver blend to the low bass and the tweeter.

I'm not going to finalize the network until I do another batch of measurements, and have some production versions of the RS28a on hand. But it's looking quite promising, and not all that different from the last one, except some bigger inductors in the LF section. Thank heavens for NorthCreek and their AWG 12 inductors.

http://www.htguide.com/forum/attachment.php4?attachmentid=2494&stc=1

Note that the tweeter was NOT measured on the same baffle; it was on a big quasi IEC baffle, and doesn't show diffraction effects from the untreated Arvo front panel which are evident in the midwoofers.

[JoshK]

Have you ever published the schematic of the passive xo you are using? I wouldn't mind trying to build it passively first, or active only between the TCs and the HiVis.

[JonMarsh]

While I have posted pictures of one set of crossover boars (earlier in this thread), come to think of it, I haven't directly published a crossover schematic, because of the state of flux I see the development in. I did have a link to a PDF with it, but changes in my hosting company's file name support killed that.


And I wouldn't want someone to go out and build an earlier version, only to see a later version with many revisions due to more extensive testing. The version shown earlier in this thread has been listened to for a couple of weeks, then set aside because of other projects- and look at the date- six months ago... I was very pleased with my listening auditions, particularly on voice, acoustic standup bass, and piano, but this was without the low bass equalizer, and without the recent changes which should extend the bottom end before the use of the equalizer.

Note that the LF woofer crossover, though it looks like a standard 4th order at first glace, isn't, and is designed to work with the baffle and driver acoustic interaction as well as provide some equalization for the bottom end below 100 Hz- this sacrifices some sensitivity, and I may elect to change that in favor of more EQ in the inline bass EQ unit in development. Right now, the design only calls for ~ 8 dB of EQ below 60 Hz. It doesn't have a subsonic filter, and I may add that, too, depending on how the system handles program material. But I'd rather not, as that will mess up the LF group delay.

So, understanding that no measurement verification has been done yet, and levels and other things will probalby get more refinement, this is just presented for illustration purposes- what the concept and component count looks like- and it should be kept in mind that it's dependent on the specific acoustical setup for the drivers; i.e., don't go thinking to use this with a symetrical baffle layout (ala SL's Phoenix or Orion) without modifications.

The driver models are the lumped measurements for the midwoofers and TC2+, wired in series in each case (only 1 VC on the TC2+). Note that the LF inductors used in the first protos and will be used for these are NorthCreek AWG12 inductors; I just buy 10mH and 5 mH and unwinde to the value I need. Gives some flexibility in design. The tweeter shunt inductor is also a NorthCreek AWG12; minimizes the DCR, and improves the very low frequency roll off (at the frequency where the inductor XL is equal to the DCR, there's no more LF attenuation below that frequency- you want DCR quite low on that shunt inductor.)

When this is further along (i.e., nearly finished), I'll start a sticky thread like the Modula- we've been asked to do this a few times already.

http://www.htguide.com/forum/attachment.php4?attachmentid=2497&stc=1

[JonMarsh]

Some folks don't have no luck at all, but sometimes I have some with speakers.

The borrowed prototype RS28a, when I measure it, seems a little more friendly to this design than the original proto measurements Mark K made. Don't know what's going on here, but now the Fs, based on impedance peak, is around 600 Hz, quite close to what the production units are reported at, and quite compatible with the existing design, as well as others using the SS9800. Hmmmm.

http://audioworx.virtualave.net/RS28aZSS.jpg


This may be an easy drop in other places, like the trusty old M8 two ways and MTM's. :B

Should have some listening experience later this weekend.

~Jon

[Paul H]

This may be an easy drop in other places, like the trusty old M8 two ways and MTM's. :B
~Jon


Some of us are listening with interest .. ;)

Paul

[Jam_Master_J]

Me included, I might hold up on buying drivers. Will that tweeter change my crossover at all? You mentioned it as a drop in upgrade so I probably can treat it as a simple part swap?

[JonMarsh]

It looks very close to that for the XT25 and SS9800; might need a small level tweak (LPAD adjustment). By next weekend I should know for sure. The RS28a is expected to be available within the next 30 days or so.

[Jam_Master_J]

Nice, so I don't need to hold off on buying the crossover compents for the MTM design? I want to do up the cross overs and cabinet first and that way when I have the money to get the drivers I won't have to wait as long to finish.

[JonMarsh]

Yes, that's basically correct. There might be an adjustment of a uF or two on cap values, and maybe a tweak on resistor values. I should know better this weekend-

...stay tuned.

[Cdub]

Jon,

I know this isn't a completed design yet, but I was wondering if you had the driver layout and frame plans available? I checked both the audioworx site and this thread, but both seem unavailable. :(

thanks,
Chris

[JonMarsh]

It's all part of my calculated plot to build suspense and frustrate the DIY community...


Seriously, I'm trying to avoid a situation where the design isn't finished and enthusiastic DIY'er are already building them for themselves, such as the M8ta. Some details, such as the degree of U baffle, might change- such as being reduced. Then again, may not. If you want some PDF's of current configuration, I can probably generate that and mail you- send me a PM.

I'd been hoping to have more done by now, but work has been freaking unreal since before Easter- doesn't show any signs of lightening up for a while- the latest today was being assigned an urgent assignment to present at an IBM Symposium, for which dates I'm already committed to be in Montreal doing training for Future Distribution. And this is in mid May... don't know who they're going to wind up sending... maybe the Marketing guy should go, huh? ;)

Then there's those pesky little distractions, like finishing up the RS180 based MTM's- they do sound sweet, though- I think you'd be impressed to hear them- very nice on music, even though I "designed" them for HT (well, for me- that was mostly a size thing).

Last, 4 of the new Peerless Exclusive Nomex series are on order from Solen, but they won't ship until May sometime- that could be late May, I'm told.

http://www.d-s-t.com/peerless/data/foto/830884.jpg

~Jon

[ThomasW]

I know this isn't a completed design yet, but I was wondering if you had the driver layout and frame plans available Here's what we have. Use it at your own risk, understand that everything in the design is subject to change/upgrade.

http://pwp.t-3.cc/ttriff/ArvoData/

Although Jon is changing aspects of his design, my intent is to use the design as is (meaning same drivers and same baffle design). So there will be some support for it should you choose to start building.

I've been using the prototypes for over a year, and am quite pleased with their performance.

There will be a matching center channel designed and built later this year. It will use the same SS9800/M8a/12"TC2+ as this version of the Arvo. One BIG caveat, none of these drivers are shielded. So the center should not be used near a direct view CRT.

[JonMarsh]

I'd forgotten about those being up there...

Definitely Don't use the crossover schematics on the link Thomas provided- as I mentioned, I'm investigating 4 ohm VC mode with the TC2+, and the crossover, as recently posted on HG Guide has changed from these PDF's, and could change further IF I decide there's merit to the Peerless 830884's. Should be valide for cost estimation; note the most recent design is predicated on the RS28a-4, to try to get costs down. The proto is pretty much a drop in for SS98000. Still, there will be a version with optimized design for the M8a's, which cost about 1/2 the Peerless.

Note that the front panel should be mirror imaged; shown is a speaker that would be on the listener's left.

There's a bunch of rough pics I took during the assembly of this current mule- not that I'd necessarily use that as a final guide for how to do things. Also, I do have Bolivian Roswood purchased and an idea in my head about how I'll finish things up, but that's still in the formative stage- not that you couldn't do something different, anyway. Rockler up in Pleasant Hill is the source of the wood. There's also an LF equalizer, to boost things below 100 Hz a bit, but that design wouldn't be finalized (except in topology) until the LF configuration is set, regarding voice coils.

Too many balls in the air, huh, Chris? ;)


~Jon

[ThomasW]

I'd forgotten about those being up there...That's because they weren't there. I uploaded them 15 min ago ... :roflmao:

[Paul H]

Ya think someone might be due for a vacation ;) :lol:

Paul

[JonMarsh]

Past due... way past due... like a library book that's been under your bed two years! :E

[Cdub]

Thomas and Jon,

Thanks for getting the tentative plans up. I was planning on getting all of the pieces cut out and then wait for any further changes before putting everything together. This wood working class I am taking has me wanting to build stuff. My plan is to build cabinets for a few projects while I have access to the tools, and then take my time finishing the cabinets and gathering the drivers and crossover components as finances allow. Right now my project is to build enclosures, I can route for drivers at a later time if necessary.

Its good to know that the current Arvo design will be supported in some fashion.

I'm in no hurry with regard to the final crossover design or final driver selection. Do you have any guesstimates as to whether the actual baffle dimensions of the Arvo projectwill change if the new drivers are good candidates?

Also, I know you will be evaluating the Peerless drivers for the arvo, will you also eval them for use in the MTM as an upgrade from the M8a?

Hey! Do you guys have any tantalizing tidbits on the design plans for the center channel design?

BTW, I thought I needed a vacation, but I decided to give you a few of my vacation days:) I'll fed ex them right over!!!

Chris

[JonMarsh]

Thomas and Jon,

Thanks for getting the tentative plans up. I was planning on getting all of the pieces cut out and then wait for any further changes before putting everything together. This wood working class I am taking has me wanting to build stuff. My plan is to build cabinets for a few projects while I have access to the tools, and then take my time finishing the cabinets and gathering the drivers and crossover components as finances allow. Right now my project is to build enclosures, I can route for drivers at a later time if necessary.

Its good to know that the current Arvo design will be supported in some fashion.

I'm in no hurry with regard to the final crossover design or final driver selection. Do you have any guesstimates as to whether the actual baffle dimensions of the Arvo projectwill change if the new drivers are good candidates?

Also, I know you will be evaluating the Peerless drivers for the arvo, will you also eval them for use in the MTM as an upgrade from the M8a?

Hey! Do you guys have any tantalizing tidbits on the design plans for the center channel design?

BTW, I thought I needed a vacation, but I decided to give you a few of my vacation days:) I'll fed ex them right over!!!
Chris


Now that I could really use- I'll be watching my Fed Ex guy like a hawk!

But the problem isn't vacation days- the problem is work load and being able to use vacation days- as I pointed out to my boss around December, I had enough vacation time to take a week off every month starting in January, and do that for nine months- and by September, I'd still have 4-5 weeks off due, because of additional accurals.

Back to business...

The Arvo baffle design won't change - that's a function of what works flattest for driver loading and staggering the edge and boundary effects- would be the same, regardless of midwoofers used.

MTM's could be modified, too; who knows; MKV, like the original M8, might be dual peerless woofers with the RS28A-4 tweeter. Crossover would get tweaked, of course, especially since the Peerless woofers have much lower inductance, due to Faraday shields.

The deciding issues will be energy storage, midrange distortion, and T/S parameters and LF performance. That and whether they look as cool in the flesh as in pictures. ;)

The existing baffles are 1" thick, with backside reverse bevels behind the midwoofers. They're laminated up from 1/4" HDF. Other materials could work, but might not be as machinable- like 1/4" HDF over baltic birch bly.

Backside view may give you some idea...

http://audioworx.virtualave.net/AP2S04.jpg

An intermediate assembly stage...

http://audioworx.virtualave.net/Arvo12Empty2.jpg

and

http://audioworx.virtualave.net/ArvoBack1.jpg

and

http://audioworx.virtualave.net/ArvoBackWoofers.jpg


http://audioworx.virtualave.net/ArvoUpperBack.jpg


Now, this is all top secret info, that we don't want Linkwitz or John K to get a hold of, so don't go posting this stuff around on the internet.... ;)

Driver sizes are the same, if I remember correctly; both are 8-1/2" woofers, as are the Dayton RS225, which in all likelihood is NOT a candidate for this design (unless we go to a waveguide load for the tweeter... not really planning that at this point).

~Jon

[jdybnis]

But the problem isn't vacation days- the problem is work load and being able to use vacation days- as I pointed out to my boss around December, I had enough vacation time to take a week off every month starting in January, and do that for nine months- and by September, I'd still have 4-5 weeks off due, because of additional accurals.

~Jon

Just think of those days as an investment. Every time your salary increases those days go up in value. :D

-Josh

[Dennis H]

I'm gonna tell John K. Everybody knows he invented the U-baffle (along with the internet) and holds the exclusive rights to it. :roflmao:

[JonMarsh]

Well, I can show prior art- first Arvo post with V1 was pretty much the day Linkwitz put up the Orion challenge (an H baffle) - a LONG time ago! ;) :B

Still, we gotta keep those secrets, ya know....

[Dennis H]

Aw heck, ages ago SL had on his web page how you could build a cardioid by putting a broadband acoustic resistor (if you could find such a thing) behind a U-frame. John, errrr..... borrowed the idea, using fiberglass as an imperfect broadband absorber. ;)

[Davey]

Hey now! John K. didn't invent the internet, that was our favorite politician from Tennessee. Let's get the facts straight Dennis. :)

SL told me that he started experimenting with "H"-baffles and variants way back in the 70's.

In case you guys missed it, there's some updated information on the Pluto at SL's webpage.
http://www.linkwitzlab.com/Pluto_1.htm

Cheers,

Davey.

[Cdub]

Jon,

what your supplier for 1/4 hdf? would MDF still work OK? I was kinda leaning toward HDF if I can find a source for it...might as well go all out if I am going to build! I know you guys wouldn't expect anything less :W

Chris

[JonMarsh]

Jon,

what your supplier for 1/4 hdf? would MDF still work OK? I was kinda leaning toward HDF if I can find a source for it...might as well go all out if I am going to build! I know you guys wouldn't expect anything less :W

Chris


Gosh, Chris, that's more info that we can't leak out over the Internet less John K get that, too!

The easiest way to get manageable size pieces of 1/4" HDF is your local big orange, unless they're out of stock, which is quite frequently in my case. I don't remember if the Rockler store has any (have you ever been to it, up in Pleasant Hill? They've got BB ply, too, lots of other stuff- a lot more than the Dublin Woodcrafter's store). Getting HDF through ordinary retail at > 1/4" thickness is hard to do- that's why I laminate up what I need. It's a PITA, but what's a guy to do? Thicker sections are used for flooring underlayment sometimes. But it's hard to find outside of wholesale outlets.

If you'd like to be all ecological in your construction method, I'd consider Medite II (manufactured by SierraPine ) over BB ply- it's a non formaldahyde MDF that's recommended for hospitals- but it's not recommended for "structural" applications. Rabetting or routing driver setbacks would be a piece of cake, and it would finish nicely with paint or veneer.

~Jon

[Steve Goff]

Aw heck, ages ago SL had on his web page how you could build a cardioid by putting a broadband acoustic resistor (if you could find such a thing) behind a U-frame. John, errrr..... borrowed the idea, using fiberglass as an imperfect broadband absorber. ;)
And Roger West has been using sound absorbing baffles behind his largest electrostatic speakers for a very long time.

[JoshK]

I am getting way ahead of myself here but I REALLY want to build these speakers. Problem is my wife took one look at the pictures of Thomas' (I think) and said, "way too modern, won't go with our decor, I like your current speakers better". Well of course that doesn't deter me. :roll: However, it does get me thinking about how to make them more classical in looks.

I thought i'd through out some ideas I have been brainstorming and get your guys' feedback and additional ideas.

The background: Our house was built in 1873 and is a traditional victorian/neoclassical row house. We have 12' tin covered ceilings :blink: with large crown moldings and many other 'classical' looking features. My wife went with a more traditional/classical looking furniture setup.

http://www.htguide.com/forum/attachment.php4?attachmentid=2814

- One idea I have is the finish the baffle in leather to match the couch. Our furniture dealer already told us he can sell us all the fabrics/leather in to make our own matching accesories. My wife is handy, can sew and stuff, so I can utilize her help in finishing.

- My wife also objected to the swooping (orion) or angular (arvo) side profiles. Hey, she is the one looking for a boxy look! So an open rectangular frame like MFK's dipoles comes to mind.
http://www.mfk-projects.com/images/RS%20dipole%20project/Resize%20of%20loading%20drivers.jpg
The open portions could then be covered with grill clothe and trimmed with nice mahogany to match our furniture, somethink like MFK did.
http://www.mfk-projects.com/images/RS%20dipole%20project/cheery%20trim.jpg

These are comprimises obviously, but would love to hear your opinions. How might this affect the sound? What other good ideas do you chaps have?

Also, I am still not quite sure how the U frame and H Frame designs would differ in this project. I still need to read up on their properties. MFK again has some really nice looking angled H Frame bass 'enclosures' that look really nice. Of course I can do this with a U frame as well, except for the angle mounting.

http://www.mfk-projects.com/images/construction%207.jpg

[Evil Twin]

Well, I'm a Danish modern kind of guy for a lonnngggg time, so I'd be clueless about what to suggest- I'm drawing a total blank when trying to imagine a neo-victorian styled diople speaker.... ;)

The biggest problem with MFK's dipole design isn't the cabinet or baffle layout, it's the limited Sd and Xmax of the RS270's- they're nice drivers, but with no where near the swept area needed for dipole bass (not with just two per side- 4-6 is where they need to be). If substitute the sound splinter 10" based on the TC2, it would be closer- better still to upsize the whole thing and use real TC2+ 12's.

So if you like MFK's design esthetically, consider upscaling it, getting rid of the edge stuff on the front panels (diffraction, you know), and using an asymmetric baffle layout (the frequency response is flatter, distributed boundary effects, you know).

While you're looking at alternative dipole concepts, consider the Gradient revolution, too. I've been toying with that idea in my head...


http://www.restek.co.uk/images/revolution.jpg

Keep in mind, if you like strong bass below 50 Hz, you'd might need to consider a configuration more like their Revolution active, with addtional woofer modules (I don't think their 12 has any more throw than a Peerless XLS)


http://www.gradient.fi/images/revo_act_01.jpg

But this isn't neo victorian, either, is it? Maybe with some art deco flourishes in the woodworking? ;)

Like I said, clueless....

[JoshK]

I threw the picture of the couch to better represent what I am getting at in style. Victorian is technically lots of color and whatnot like the french do, but quite often it is missused as a descriptor in a neo-classical look of the late 1800s which is how I am using it. I think the couch hits the style spot on. Anywho...

ET, I was planning on just that, not using MFK's design, but rather JM's Arvo design (just ordered the RS28a's!) and modifying the aesthethics to be closer to that of MFK's 'look'. I too have considered the diffraction affect that the lips on the side of the baffle would create and often wonder why SL and MFK decided to do that. I have a pretty cool 2" roundover bit and router setup that I could pull out for this project to make the trim flush with the baffle and rounded.

So far my ideas don't really get me to neo-classical, but my wife has already resolved to accept my current speakers (VMPS RM40s) which are far from that, but are squarish and veneered in natural cherry. I still would like to see how matching I can make the Arvo design aesthetically.

[Dennis H]

You can do a lot to get that Neoclassical look with moldings to sort of simulate the lines of a classical column, especially at the top and bottom. Moldings similar to the ones in your house would be good. You can do quite a bit without messing up the front baffle (especially near the tweeter) and overall dimensions. While not the style of your house, I thought this Stickley (Craftsman) style Orion on SL's page was pretty clever.

http://www.linkwitzlab.com/images/photos/Naples-2c.jpg

[Cdub]

I think I'm going to sandwich 3/4 birch ply between 1/4 hdf for my construction. Would this be OK? Also what is the glue of choice of all you cabinet makers out there? I was planning to spread the glue evenly with a brush unless there are better alternatives. Will it be necessary to apply any pressure to the "sandwich" once I glue it up. I imagine it would be knda hard to clamp up.

Any advice from any of you experts would be appreciated.

Thanks,
Chris

[JonMarsh]

I used BB ply with 1/4" HDF for the M8ta's side walls, back, and braces.

For that application I apply titebond II with a small squeegee. Spreads well, cleans up easier than a bruch.

Yeah, you definitely need pressure on your sandwich, unlike a club or BLT!

I use cast iron angle extrusions for alignment aids, and lots of clamps at the edges. I also put big Plitron power transformers in the center, but you know, I think bricks or cinder blocks might work just as well... ;)


http://audioworx.virtualave.net/Panel2ss.jpg


http://audioworx.virtualave.net/Panel3SS.jpg


This is representative, though here I'm gluing up the four layers making up the M8ta front panel- same process, smaller scale, in your case.

~Jon

[PMazz]

I'll add to Jon's advice....

Make the panels oversized so you can throw a couple screws into the sandwich to keep it from "squirming" when glueing up. An inch or 2 on each end is enough. If you move fast enough with glue, try to do several sets of glue-ups at once. We use paper sheets in between layers to keep them from becoming one big set ;). Use as much flat scrap material as you have for cauls to spread the clamping pressure. You can use a 2x4, cut to proper length and wedged against a ceiling. Make sure you use cauls.

Another method, using 2x4s as cauls, would be to clamp the stack between lengths of 2x4. If the 2x4s are crowned, which they usually are, use the crowned side towards the stack to apply more pressure in the middle of the panels. Just make sure they stay flat....

We use steel I-beams as cauls and clamp the stack on horses.

Pete

[Captain Cojo]

A couple of 4x4s run lengthwise directly over the driver holes joined with alllthread makes a great way to clamp the middle.....

[Paul Ebert]

I'm considering building a pair of these, or perhaps, the Saint-Saens. Either way, I'm going to need a power amp or four. Any recommendations? My budget is below $3000 (more like $2000). I've very open to used or DIY. Should I go with a pair of Pallidiums? Or, an Ayre amp or two?

How about UcD modules? They might be inexpensive enough to make the Saint-Saens affordable.

Thanks for any thoughts!

Paul Ebert

[JonMarsh]

I've been pretty happy with the used Aragons I've gotten (Audiogon is your friend ;^)

But the odd thing is they sound markedly better with my Ayre balanced preamp than with a passive single ended preamp using Shallco attenuator- I have no idea why, unless the cables are a bigger factor than I'd expect, too (I'm using Ayre Signature Balanced 3 meter interconnects with the Ayre preamp- the Ayre cables are custom manufactured for Ayre by Cardas, to Charles Hansen's design- it took almost a year to get them the way he wanted - I know, I was waiting and following along through that process).

I have bought some UCD modules to build up and test, but until I do that, no opinion; like an analog amp, I think the impelmenetation details (power supply, wiring, ect) will be important. Take a look at Mac's implementation; I'm doing something similar with Par Metal Chasis and Avel-Lindberg transformers, Panasonic caps. I'm planning a custom PCB for the power and interfaces.

I like the V5/V6 and V1 Ayre amps a lot, but I'm not as fond of the older V3. THe X and evolution upgrades for the newer amps are quite worthwhile.

In my own case, these big speakers are partly being done so I'll have stuff to test my own amplifier designs (heavily modified non loop feedback jobbies starting with an Aragon chasis, tossing all the Aragon active circuits - that's the Aragon Xmod project - on the third iteration of that design)

For a multi-way speaker, an HT style amp might not be a bad way to go, as you don't have bass on all the channels- maybe on of the newr HT style amps from Aragon, which are often for sale on AudioAdvisor?

There are many, many other good conventional amps you might consider- Parasound, Bryston, Rotel, etc. I like the sound of the Theta Digitals about as much as Ayre, but they're also expensive- similar non feedback design, and very good build quality.

If you're not in a rush, I'd check Audio Advisor, watch the sales on Audiogon, and hang tight with what happens with the UCD stuff. In our company, we're evaluting class D designs internally, and are looking at a cooperative effort with the real biggie in Class D as regards publisehd IP. More than that I can't say. We're evaluating their moduels, too. They're pretty "cool". ;) Our PM's and TM's in Villach are raving about the sound, compared with their standard gear (best one guy has is a Naim), but they're young and not as experienced as some.



~Jon

[JonMarsh]

Listen to Pete, Chris- he's the expert with the woodshop stuff, I learn something new everytime he posts.

I'm just a guy from the electronics lab that's bought too many power tools over the years! ;)

~Jon

[Cdub]

Thanks for the clamping tips guys. I went to the big orange store today to look for 1/4" hdf, but came up empty. I asked one of the workers and he thought it was in the building supplies area as opposed to next to the MDF, where I assumed it would be located. But no luck. If I can't locate anything I was just thinking about using 1/2 " MDF instead of 1/4" HDF, would this be an acceptable alternative?

Chris

[JonMarsh]

Most of the big Organge stores only have 2'X4' pieces - it's in the precut section. And very often that is out of stock, so when it is available, I stock up.

But 1/2 MDF is OK, too- just stay away from that PB stuff! ;)

Another alternative would be to try a non-big bix lumber store-

[capslock]

Jon, your HDF looks a lot like MDF on those pictures. The HDF we get here is usually mid-to-dark brown, vs. light brown for MDF. Also, our HDF is coarser, you can sometimes see little regions which have one fiber grain. One side is usually very smooth (it's coated with some resinous stuff that you need to grind carefully to make the paint stick), the other side can be very rough with a square grit structure, or also smooth.

I'd have to go back to my old lab book to get the figures exactly right, but quoting from memory, HDF has a density of about 1.1 g/cm³ vs. MDF around 0.7 - 0.8. The thinner pieces measured higher, which makes sense, because even in MDF, the top and bottom layer are more compressed than the bulk.

Seeing that you are making sandwiches anyway, you could route away large inside areas of the inner layers and fill the sandwich with sand. Bernd Timmermanns of HobbyHifi magazine did comparative tests a while ago. He got the best attenuation from a fill of 50% sand and 50% lead shot.

Personally, I prefer a sandwich of MDF, 4 mm glass fiber filled roofing tar and 10 mm tiles....

[JonMarsh]

Our "tempered" hard board looks like what you describe, Eric.

I've only used layerd construction on two projects- the M8ta and Arvo.

[Cdub]

I got lucky and found a piece of 48" x 96" 3/4 HDF which I decided to laminate to 1/2" baltic Birch ply. The next step is to laminate the pieces and get to cutting!

Chris

[Jonasz]

Hi Jon!

Why are the mids and tweeter placed like they are on the baffle? What happens with the horisontal radiation? Won't there be different characters to the sound if you're listening a little to the right compared to the left of the 0 degree axle (if you know what I mean ;) )?

Jonas

[Evil Twin]

You're referring to the Arvo's, correct?

They're placed the way they are for the smoothest overall frequency response on axis; in practice the speakers are toed in just slightly.

As a few others have found, very similar principles apply for midwoofer and tweeter driver positioning on the baffle for dipole as for enclosed operation. The Phoenix type orientation does not produce the smoothest loading and frequency response for a dipole anymore than it would for an enclosure style speaker.

I found this through a combination of modeling and construction and measurement, as Troels Gravesen did through construction and measurement alone.

Try it for yourself and see. Both versions of Arvo have been built this way.

[Jonasz]

I'm sorry if I offended you Evil Twin. I should have known the "strange" look of the baffle geometry wasn't a random thing done by the Dark Side... I so much fear and appreciate your vast knowledge of audiotechnology that I will follow you and do the same in my new WMTMW (all Dayton RS) design. :twisted:

I'm just trying to learn something here... :B

Jonas

[JonMarsh]

Youv'e gotta be careful around that guy, you know.... give him a cross look, and the next thing you know, it's getting kinda difficult to breath... at least there's none of that lightning stuff...

Those BDS tools are pretty handy, just not the easiest to use.. but it's amazing what they did with just a spreadsheet. I'm pretty sure I could model the same thing in LspCAD6 these days, as I think I can model the dipole baffels now, but I just haven't had the time what with the insanity at work.

If it weren't for the being older part, I wish I was retired so I could work on this stuff nearly full time!

[oneoldude]

Hi all,

I am a newbie here and have a couple questions because I am in the process of designing and building something like the Arvo. Hope you can help.

What is the designed target maximum SPL for the Arvo system? By that I mean the max SPL that the system is designed to produce, say from 50 Hz to 20 kHz with all drivers staying within their respective Xmax?

I am especially concerned about the tweeter. With a crossover in the region of 1.25 kHz, I suspect that it will need a whopping Xmax to meet any reasonable SPL in a linear manner 8O .

For example, if the design SPL is 110 db (to handle musical and HT peaks and spikes without distortion while listening at say, an average of 90 dB) and if the crossover is at 1.25 kHz, the tweeter should be able to handle 104 db or more at 1.25 kHz and not exceed its Xmax. I assume here that the mids will contribute 104 db and the design point of 110 db will be met (if my numbers are right). At this high a frequency we cannot count on in-phase contribution from the other channel so each channel must stand on its own. I might add that no matter how sharp the crossover is below 1.25 kHz, the tweeter will still have to perform to 104 dB and stay within Xmax at 1.25 kHz.

So now to more questions. What is the Xmax of the chosen tweeter? And what SPL will the tweeter produce at 1.25 kHz at Xmax?

The reason I am asking these questions is that I am facing the very same issues in my design and am having trouble finding answers. Is there a nomogram, chart or some software utility that will help me figure out what the max SPL a tweeter of a given diameter and given Xmax will produce at a given frequency? It simply will not do to have the tweeter go into gross distortion right in the band where human hearing is most critical. :M

Linkwitz has a spreadsheet program that will do calculations of this sort, but only for woofers and midranges. It does not cover tweeters.

Can someone help? It would be great if there was a utility out there that simply required Effective Driver Diameter, Xmax and F and produced dB SPL for the driver into half-space. :T

Thanks :B

[Dennis H]

Linkwitz's spreadsheet works for tweeters if you copy/paste the rows to cover higher frequencies. The frequency is calculated from the frequency in the row above so the copy/paste works fine. Add 6dB for 2pi space and another 6dB for the contribution of the mids. Using his spreadsheet, the Seas Millennium will do 98 + 6 + 6 = 110 dB at Xmax at 1280 Hz.

[JonMarsh]

Yeah, what Dennis said- good, quick answer!

While the Millenium Excel is known for it's heavy duty specs and output capability, let me say that's not your ownly choice. Based on the testing we've done at Mark K's, including a re-test on the Millenium and 27TDFC, and using the rather punishing 3 tone cluster centered at 1 kHz, with 800 Hz and 1.2 Khz at 50% of the 1 kHz signal, the 27TDFC hangs in their like a little monster- it's certainly the class champ in $30 tweeters. It doesn't meausre quite as flat, so it needs a little more voicing work.

OTOH, there's also the RS28A-4, which also is an over achiever in the low frequency end, AND has much lower high order distortion than the Seas, due probably to all that copper in the gap keeping inductance and inductance modulation very low.

Any other potential candidates you're interested in, I SERIOUSLY recommend you plan to purchase and measure them. Spec sheets for tweeters as a tool to estimate there performance have limited value- you can eliminate some obvioius clunkers, but there's no substitute for your own testing, except maybe using someone else's testing, if they've covered the drivers you're interested in.

BTW, my nominal design goal for the Arvo's was only about 104 dB peak, though obviously in some frequency ranges they have more left in them than that. There's another system in the works for the 110 dB range, actually, two possibilities being investigated, but things get rather pricey as you get up to that turf. I look at it more as an excercise in distortion reduction, and I've been spending more time than I orginall expected to complete the Arvo just for that same reason- distortion reduction through optimal driver selection.

OTOH, it was obvious the configuration had a lot of merit when I first listened to the first test mule for the concept with just the Dayton Titanic MKII 10" woofers, M8a midwoofers, and Vifa XT. Since then it's evolved a lot in the intervening years. The last step probably is evaluating the Peerless midwoofers (Exclusive Nomex series) which are on order. I'm hoping for midband distortion similar to Excels, but at half the price. Yes, I can afford Excels, but good engineering is not spending money unnecessarily. REALLY good engineering is getting $1 worth of parts to do $10 worth of work. I'm not in that class.... ;)

[oneoldude]

Thanks for the replies guys.

Dennis - I already tried some swaping and pasting to the Link excursion program but never got it to work. I guess its back to the drawing board. :roll:

Jon - I was thinking of the RS28A-4 based on your experience. But I was unable to find the spec for Xmax. Do you know the spec? BTW are you the Marsh that wrote some articles for the Audio Amateur some years back?

I am interested in what you are doing here and hope to learn from your experiences. So I hope you don't mind if I ask too many questions. :T

I have IMP-M, True RTA and Calsod as my primary audio tools. While IMP and RTA are quite slick, I cannot say that of Calsod. I have yet to model and optimize a crossover or notch filter or (anything else) on Calsod and have it work worth a $#$%^#! :M Yea, I know, GIGO. :roll:

I like the idea of a passive crossover for the M and T and then active for the W and SW. I can then use my DCX2496 as a three-way crossover (MT-W-SW) and still have the luxury of equalizing the M and W dipoles and the SW as well. :B

If I felt confident that I could integrate and voice a decent T to either the RS 8" or the HiVi 8" I would have them ordered as we speak. But for now I will wait on you. A final design will be a blessing. But I suspect when those new Vifa drivers finally arrive there will be other new drivers in the wings possibly worth waiting for. And the beat goes on. It will never end. :B

Also, where are you crossing over the M to W. It seems to me that if the crossover is in the 150 Hz region you have no need for two M drivers. You should be able to get to that 110 dB level with only one. Also you would not have to deal with the roughness of the MTM design. Indeed, the same can be said for the W. If a SW is cut in high enough, say 40 or 50 Hz you could get by with a single W. As someone here suggested, it makes no sense to use two expensive long throw woofers as filler drivers. Could you elucidate? :T

Hope I have not been too much of a pain.

[oneoldude]

Jon,

One more thing. I agree that you must test drivers to see if they are useful in a given application. But that is not to say that specs are totally meainingless. Most manufacturers do not under spec their products. So we can conclude that the specs are at most a best case scenario. Therefore, if a given driver specs an Xmax that will not do the job, I will not consider it. But if a given driver specs an Xmax that will do the job, it MIGHT be worth considering. That is why DIY'ers like me rely on dedicated persons like you to do some of the filtering for us. You help us all get a higher quality project completed at a lower cost than most of us could do alone. We count on you. And we thank you. Kudos to you.

[JonMarsh]

Lots of interesting questions and inputs here! I'll try to take them in order and keep things straight....

No, I'm not the Marsh that's written articles under than name in AudioXpress. The only article I've written for them was a series on "An Unorthodox Eight Inch Two Way Speaker", a three part article published in fall of 2003. I'm not the Marsh guy that has a high end electronics company, though I've done some design consulting in that area. To give you an example of how confusing the "name" thing can get, there's a long time contributor to AudioXpress by the name of Charles Hansen, who is NOT the Charles Hansen which is an old friend of mine, and founder of Avalon, and Ayre Acoustics.




I have IMP-M, True RTA and Calsod as my primary audio tools. While IMP and RTA are quite slick, I cannot say that of Calsod. I have yet to model and optimize a crossover or notch filter or (anything else) on Calsod and have it work worth a $#$%^#! :M Yea, I know, GIGO. :roll:

I think everyone has to find the software that works best with their workflow and way of doing things. I do have TrueRTA for messing around a bit in room, and I've used CLIO Standard for years. I've had past versions of SoundEasy (two of them), and they worked pretty well, but now I'm quite happy with LspCAD 6, used 5 for a long time, too. Praxis is on my purchase list for next month. ;) I've been looking forward to that for a while.



I like the idea of a passive crossover for the M and T and then active for the W and SW. I can then use my DCX2496 as a three-way crossover (MT-W-SW) and still have the luxury of equalizing the M and W dipoles and the SW as well. :B

This is what ThomasW does, but I'll be running my Arvo's pretty much full range, though I'd describe the 8's as midwoofers. ;) What you propose should be very flexible, and I'd hardly be the person to talk you out of it. When I started the Arvo Part project (quite some time ago, unfortunately; as our former director of power semiconductor devleopment used to say, "Slow work takes time"), my plan or desire was a full range speaker with passive crossover, partly as a listening tool for amplifier designs I've been working on. I've done multi-amplifer systems many times in the past, but I wanted this one to be driven from a single amp, with at most an inline passive line level EQ or buffered EQ. That's my story and I'm sticking to it, but I'll help others realize the system what ever way they like- not personally dogmatic about this stuff.



If I felt confident that I could integrate and voice a decent T to either the RS 8" or the HiVi 8" I would have them ordered as we speak. But for now I will wait on you. A final design will be a blessing. But I suspect when those new Vifa drivers finally arrive there will be other new drivers in the wings possibly worth waiting for. And the beat goes on. It will never end. :B

There's undoubtedly a lot of truth to that - new drivers in the wings? Well, here's the status from my perspective.

For the tweeter, I have decided to "freeze" the design with the RS28A, based on the performance of the prototype; I haven't listened to the production parts I have yet, but they appear to be essentially identical. At most, a minor tweak to the SS98000 crossover, as noted elsewhere.

For the midwoofers, I am still waiting to evaluate the 830884 Peerless (I think that's what you meant referring to Vifa above?) If those weren't available, it would be the HiVi M8a (for the budget oriented builder) or the Seas W22 for those wanting the lowest distortion regardless of cost. I hope to bridge the difference in price and performance with the Peerless. We'll see pretty soon, sometime in May, I hope.

For the woofers, I haven't seen anything better adapted to the design requirements than the TC2+ with one voice coil in operation. There are some other woofers I expect out in June or July which are also very promising, but they may go into the Saint-Saens or another project idea I have. My requirements for the woofers because of the desgin parmeters aren't easy- moderate sensitivity, low Le, good distortion behavior out to 200 Hz, good non-linear distortion down low. I've only seen one other driver that does what I'd like, and it's not in production yet.



Also, where are you crossing over the M to W. It seems to me that if the crossover is in the 150 Hz region you have no need for two M drivers. You should be able to get to that 110 dB level with only one. Also you would not have to deal with the roughness of the MTM design. Indeed, the same can be said for the W. If a SW is cut in high enough, say 40 or 50 Hz you could get by with a single W. As someone here suggested, it makes no sense to use two expensive long throw woofers as filler drivers. Could you elucidate? :T

Crossover is in the 200 Hz range. See last devlepment simulation plot.

http://www.htguide.com/forum/attachment.php4?attachmentid=2494&stc=1

If you're familiar with Linkwitz's spreadsheet, then you'll see I'm kind of overbuilding in the midwoofer range. But, OTOH, because those drivers carry so much of the "music", I wanted them to not be pushed close to the Xmax limits. Overall, my SPL goals were 6 dB higher than Linkwitz's design- and without the "benefit" of W22's. The current test mule has a very nice balance and VERY at ease midrange an vocal sound, even when there's a lot going on with drums and bass instruments like bass guitar, or standup bass. So, that's a specific "voicing" decision on my part, which you might argue is overdesign, but I don't feel so for the full range version. Now, for ThomasW's, where he's crossing to an IB sub, the TC2's are totally coasting- could be replaced by one or two 10s, probably - unless something really vigorous just comes along in that range...

MTM's are "difficult" when you try to implement a crossover in the wrong frequency range, as most folks do. If you keep in mind the idea that the driver centers should be no more than 1 wavelength apart, they start to get a llot easier to "tune up". But for a typical 6-1/2" driver MTM, that means a crossover frequency no higher than 1400 Hz. How often do you see that?

If you have a subwoofer cutting in around 50 Hz, then 1 12" might be OK- depends on your SPL goals and how "at ease" you like your system to sound. Certainly there's no reason not to try that first. OTOH, which might sound better- 1 12" staying within 15 mm of Xmax, or two 12's staying within 7-8 mm? You'd be surprised (well, maybe you wouldn't) how much distortion some of the 12" "subwoofer" drivers generate at just 4V RMS of driver... quite audible.

I encourage anyone to take these ideas and play around and modify them and try them out- as Chris and others have noted, I'm a little particular about how I do things, but then I like a system that fools you- you turn it up, and up a little more, and you don't really realize how loud it's playing unless you check an SPL meter, or you notice a little ringing in your ears later if you've listened too long. My old X1 SLAMM Klones were like that.... and I've been looking for lower and lower distortion drivers ever since, as it's the distortion we here as things are cranked up with a lot of speakers that both tells us it's getting loud, and tells us it's just "reproduction" we're listening to, not something a bit closer to the original sound. Really large speakers have that quality, but the trick is to get those dynamics and "cleanliness" in something of more managable proprotions that doesn't weigh 400 lB or more per side! ;)

~Jon

[oneoldude]

Jon,

Thanks for the reply. I really appreciate the in-depth discourse. And yes I did mean Peerless.

I understand the 200 Hz crossover. Also, I now understand the direction you are taking for the ultimate reduction in distortion. So I must presume that the RS28A-4 must be able to handle the low crossover point with the same low level of distortion at high power as the M8a MW pair does at the T-MW cross. Right?

Now I need a pointer. Can you point me to a quality two way crossover using the RS28A-4 and one M8a or one RS225S-8? Indeed, one of each would be better so I can make comparisons. It would be great if plots were included. The bottom end does not need to be lifted or crossed over because I will lift it and cross it over with my DCX. So an excellent two way will do just fine as a start.

Also can you please point me to the RS28A-4 specs where Xmax is specified. That spec must exist somewhere, but I cannot find it. Does anyone know that Xmax spec?

[JonMarsh]

Xmax is not frequently specified for tweeters, and when you consider flux fringing effects in the gap, wouldn't be easy to specify by a simple overhang measurement. Yes, the RS28A measures very well down low; here's a comparison Mark K measured and overlayed between one of my Millenium Excels and the RS28A-4; the RS28A-4 is in black, and overlayed on the Millennium Excel, which is red. Test tone was a three tone cluster at 1 kHz, with 800 Hz and 1200 Hz at 25% of the total RMS level, 1 kHz at 50%. That's a nasty thing to do with the tweeter. If you change the overaly mode layering (which is in foreground vs. background, the RS28A "disappears". Level is pretty high, as you can see. The Excel is clearly exhibiting more "distress".


http://my.starstream.net/mk/tweeter9_usher/graphs/prototype_millenium_nonlinear_1k.GIF


The current working network design for the Arvo is as shown below, but this could change a little more- might bypass the series attenuation to pick up the 12 kHz + level on the RS28a, as Mark did. Need to do a final sanity check on the tweeter section- this is with the M8a. As ThomasW and I have discussed, there will be a completed design with the M8a regardless of whether I chose also to use the Peerless 830884 or any other driver in a "souped up" version- it's a cost thing, for those that want to go this way, but have saving $40 somewhere in the costs in a few places makes a difference in affordability.

http://www.htguide.com/forum/attachment.php4?attachmentid=2497&stc=1


I'll draw up one sometime soon with the high pass elements on the midwoofer removed, and no EQ in the passband (there's not much here anyway- the asymmetric layout helps with flattening the axial response, as it would for an enclosure style speaker).

I'm really hoping that by June I have all the drivers in, tested, and sorted out to where I can go back to the fun part, figuring out what I'm going to do with the Bolivian Roswood I bought for "prettying up the design". ;)

~Jon

[Cdub]

Jon,

Quick question: If I plan to place a solid wood trim around the front baffle should I reduce the front baffle width to accomodate for the width of the wood trim? Also, how much can the front baffle width be increased without affecting sound quality?

Chris

[JonMarsh]

Keep in mind that small increases in the width (an inch or so) don't really make any significant difference in the baffle frequency response- Thomas was asking me the same question the other day, as he want's to start his "nice" looking Arvo's soon this summer. What I haven't decided yet is exactly how I'm goin to do my baffle edges, whether I'll stay withing something like the MkI Arvo (a thin wrap strip on the front, and 3/4" edge flush to the back), or if I'll go with something more like I'm mulling over for the Saint-Saens or Mahler, which I'm envisioning as a wider tapered edge. I'm thinking I would like to establish a "family esthetic" for the dipole systems and stick with it. However, that may not be easy in the case of the Mahler if it materializes as I'm considering. (that one is out there because of driver issues and fabrication questions- it's a two way - I know, how Retro. It's something I'm considering because of Hank Frankenberg, as an exeperiment. Otherwise, not ready to discuss).

Target on finished width is 16"; obviously, 17" wouldn't change the "tuning" significantly. I used 15.5" for the HDF part.

[oneoldude]

Hi all,

Thanks again for all the replies. It is good information.

Also thanks to Mike Van Den Broek who sent me the following info on the RS28a-4

"The voice coil winding length is 3.0 mm
The height of the magnetic gap is 2.0 mm
Therefore the "Xmax" is .5 mm (one-way)"

I am not sure who Mike VDB is, but thanks Mike!

Now I know one of the many reasons the RS28a-4 has good linear response in the low end. It's nice to know what the throw really is. It gives me a warm and fuzzy feeling. So much so, that I am ordering some ASAP.

[JonMarsh]

Some? Just some? ;) Heck, you're not considered serious in some quarters if you aren't ordering in half dozen or more lots! ;) :B

Try it, I think you'll like it.

~Jon

[mante]

Dear Evil Twin,

Did Jon just mention that he is working on a top secret speaker system code named "Mahler"? Also, is 1/4 HDF the same as hardboard, or peg-board without the holes? Finally, do you know if Jon wants to sell his current NEC projector in order to buy the projector he is currently lusting after?

A Young Padawan wanting to get into the HT projector scene. ;)

[goskers]

Sorry to get off topic a little around here but I have a couple questions concerning the dipole TC2+ used in the arvo part.

Are you trying to achieve a semi-cardiod response with the arvo? If so, how because you have to absorb a great amount of the backwave if I am not totally off tilt.

John K. specifies the usage of 'pink fluffy' in his u-frame which I don't believe is adequate if a goal is to dampen low frequencies. It is also stated that the length of the u-frame is supposed to be greater than 1/4 wavelength at the desired xo point.

I ask this as I am trying to implement some u-frames with the same woofers but am not completely sure of what to focus on.

Thanks

[JonMarsh]

No- not trying to do a cardiod type- no deep U baffle with damping here. Reason for taper is to distribute the termination frequency - like Apogee's and old Carver's did.

As a dipole, either an H or a U can be used to extend the effective baffle dimension- they have different concequences regarding esthetics and construction issues.

[goskers]

Hi Jon,

Could you please explain the differences in theory between the two designs?

I would love to hear some benefits and drawbacks between the two.

Thanks

[Dennis H]

The voice coil winding length is 3.0 mm
The height of the magnetic gap is 2.0 mm
Therefore the "Xmax" is .5 mm (one-way)

Interesting. Aside from the Xmax, we now know it's an overhung motor while most tweeters seem to be underhung (voice coil shorter than the gap). Maybe that helps explain its "grace under fire" when you overdrive it at low frequencies.

[oneoldude]

Jon,

Wish I had reasons to buy stuff by the palet. That would be great fun. But I just quit my job. :T Now I will have lots of time to play and not as much cash to blow. :roll: Time for me to get into the tight budget engineering paradyme. :B

[cjd]

Ok. I'm gonna ask a real question here.

Why cut the dipole to monopole at such a low frequency? (or rather, why not dipole for the tweeter)?

I'm not yet sure whether my brain is able to fool me into thinking I can or can not tell the difference by 1000Hz. Only testing will tell. But given that a violin goes up to 3kHz and I've played that for over 25 years...

C

[Dennis H]

Because above there, it's acting like dual monopoles i.e. no dipole nulls to the sides. You could add a rear-firing tweeter if you wanted but many people think that adds too much HF energy to the power response of the room. Many people with big Maggies or electrostatics end up mounting absorbers behind the panels to kill some of the rear HF.

http://www.soundlab-speakers.com/accessories.htm

Maybe the way to go is with a rear tweeter with an Lpad on it like JohnK did with the NaO. Then you could tune the rear HF to taste. SL has a live room and John has a dead room so that might account for John preferring the rear tweeter.

[JonMarsh]

My experience with the actual speaker mirrors what Dennis points out- seems to me the ideal place to go back to monopole is somewhere between 1200 and 1500.

The first dipole we did was only dipole from 300 Hz on down; still, that was pretty impressive, and sold me on the concept. (Nod and a tip of the hat to Jeff Duddleston, creater of the original Whisper Legacy, which we reverse engineered years ago). The Whisper got around controlling dispersion by having a large array for the 300-2.2 kHz stuff so that the directivity was controlled by beam width. But I do like the sound of a true dipole doing up higher- but much above 1 kHz doesn't seem to confer any advantages. Not saying you shouldn't try it for yourself, but then getting a driver with a small enough rear shadow to work well in that range isn't easy, either.

~Jon

[cjd]

I may just have to experiment with something like the Neo8 where it's easy enough to play with as a dipole or a monopole. I think. ;) But that's another thread.

C

[JonMarsh]

Careful, or you'll wind up like me, spending 12 hours a day writing threads...

[cjd]

Careful, or you'll wind up like me, spending 12 hours a day writing threads...

If I can keep occupied for the 8 hours I'm at work I would be happy. :| :P

I usually have a whole team of developers to document for and develop with. There's one, whose attention is waning due to Friday being his last day. And the back-end coding is just enough over my head... (which is how I like it! :T I hate coding, it gets boring. I do enough UI coding, which I similarly dislike. Gimme morons in a room with all their stupid ideas and room to swing... just might end up with something that works, WELL, and even does almost everything they need)

C

[JoshK]

Man I am getting itchy to build this.... Just ordered my DEQX PDC-2.6P couple days ago. Must resist urge to purchase all the drivers (already have a pair of RS28a's). I also need to finish up my UcD amps so that I have enough channels to drive the resulting speakers.

The cool thing about the DEQX is that in addition to being able to do all the xo work easily in the digital domain, it comes with a measuring setup. Not inexpensive, but in my mind worth the cost since it is both a component and a design tool.

Besides the crossovers themselves is there any particular EQ needed for the midrange drivers in dipole? I know you have to EQ the woofers, but what about the 8's?

[JonMarsh]

Very small amount of EQ for midrange bump, compared to what Linkwitz sees in the Orion- I believe it's the asymmetrical layout that's the benefit.

I'll try to get around to posting the current crossover transfer functions sometime soon, but as things are pretty crazy at work (long days) probably won't happen before the weekend.

~Jon

[JoshK]

Thanks Jon,

That is very helpful. I did look at SL's diagram and saw his 6db/oct boost in the lower mids. Strangely I am quite interested in deriving the transfer functions, guess it is the mathematician in me. I just want to see, since I am just learning all this stuff, whether I understand the predictable EQ needed based on the theory & design.

[JonMarsh]

Theory often makes simplified assumptions about how it is constructed or implemented, I'll try to post the raw driver responses, too, as measured on the baffle.

~Jon

[JoshK]

Absolutely! In fact in my day job, it is all about empirical results rather than theorectically implied results.

[Cdub]

Jon,

I just wanted to ask about the new peerless drivers
that you hope to be evaluating for the arvo project. I'm getting
close to cutting the holes for the drivers. My arvo frames are coming
along slowly, I just want to get them in a finished, but naked state
before, my woodworking class ends. The finish is still to be
determined.

I wanted to know if you have any info on the cutout dimensions for
peerless drivers. From the pictures posted it appears that the
Peerless drivers would not be a direct drop in for the hi-vi units
because of what looks to be "flattened" edges of the frame on two
sides, unless I am mistaken. I just want to make sure because if
indeed the cutouts are different I will await evaluation on the
Peerless units before making the cutouts.

BTW, as you have mentioned to me before I realized the arvo is a work
in progress and all design aspects are subject to change. I am in NO
rush to finish these anytime soon as these are not you average "budget
bookshelves" :) so I'm saving a little bit at a time.

I hope everything is well with you and that the traveling isn't
getting to you to much!

Thanks for your time,
Chris

PS Maybe you should get overtime for all the extra time zones you work
in!

[JonMarsh]

Regarding the Peerless woofers (which haven't shipped yet to the best of my knowledge), the cutout hole diameter is 19 cm, and the overall diameter is 21 CM. They way they're made with a tapered lip, you probably could justify using them without a flush rabbet. OTOH, using one deep enough for the M8a and using a gasket or felt ring would be OK, I'd think. The Peerless is nominal 8.27" diameter, the M8a nearly 8.5" if I remember correctly. Cutout hole of 7.5" should work with either.

I'm sure hoping the Peerless aren't going to be vaporware. Have some measured data on production RS28A, and have been further refining the crossover design with the RS28A and M8a. I need to build some new networks, but I'd like to have Praxis measured data before finalizing the "theoretical" network design. My guess is in June sometime. Late June.

~Jon

~Jon

[Cdub]

Quick question: How deep should the recess be for the tweeter, mid and woofers so that they are flush with the front baffle?

Chris

[JonMarsh]

For the tweeter (the most important, the only one emitting frequencies in a wavelength where diffraction from driver/panel interaction will be significant) it about 0.125- 1/8th an inch.

I take the rubber gaskets off the TC2+, so there it's about 0.250. For the 8" midwoofers, about 0.20.

Have a further refinement of the crossover for RS28A-4, M8a, and TC2+ which I'll post today or tomorrow, depending on time- which is tight, due to magazine article deadline (worked on that all weekend!)

~Jon

[oneoldude]

Hi all :T ,

Can anyone tell me the max SPL at one meter the RS28A will produce at around 1 KHz before audible destortion begins :B ?

[JonMarsh]

It depends on what you consider audible. Also, MarkK has seen some variability in the distortion performance at low frequencies; we suspect it may be a diaphragm/VC/gap alignment issue. I've seen this in other tweeters, and sometimes gotten the best results by hand tunning using an FFT analyzer with test tones.

The last tweeter torture tests I saw Mark run were on the SEAS 27TDFC and Millenium Excel, at levels over 100 dB with a three tone cluster (800 Hz, 1 kHz, 1200 Hz). The RS28A is in pretty much the same category- OTOH, I do try to get off of it pretty steeply, so that it's down as much as possible at 1 kHz. For reference, with the specified crossover, it's down -6 dB at 1200 Hz.

[oneoldude]

Jon,

Defining audible distortion is like what the supreme court justice said about defining pornography. Like him, I don't know how to define it, but I know it when I hear it. I bet you do too. :B

Problem is, that when you cross a tweet as low as in the Arvo, it is carrying a great deal of musical info right where the ear is most sensitive. The trade off between off-axis response, SPL and distortion is a never ending issue at that juncture. I am simply trying to get a handle on how good, or bad, the RS28A is in that part of the spectrum. I have asked Mark K about it. I am awaiting his response. :roll:

BTW, do you have ASCII FR and Z measurements of the RS28A and M8As taken from the Arvo baffle that you will share with me? I would like to model the system in CALSOD so that I can understand it better. :T

[Mark K]

Hi old dude,

I got your email. I'm not sure I can answer the question. The audibility of distortion is a very complicated question. Audibility depends on frequency, number of frequencies, number of distortion products, duration of signal, and I'm sure a bunch others that I can't think of on top of my head.

And the distortion results can be quite varied.

So, you have to rephrase the question in more detail, what tweeter can reproduce a 1k sine wave at 110dB@1m with no more than -30 2nd order HD, -30 3rd order and all other products less than -60. Otherwise it becomes very hard to know.

It's much easier to design an arbitrary test signal and then see how different tweeters perform. My tests are designed to look at distortion signatures corresponding to levels somewhere around 94dB@1m. While this isn't what I would consider a peak level, the test tones are quite brutal and it would be unrealistic to expect any consumer 1" dome to do well reproducing a three tone continuous signal at 105dB. Trust me. Perhaps a horn or pro driver. But that's a different set of compromises.

Now simple HD testing could probably be done at 100-110, and see how low one could go. I just don't think this is relates much to audibility.

You had mentioned in your tests the dynaudio d28. I've not tested that directly. But dynaudio has developed a bit of a mythology since it exited the diy market. Is it warranted? Certainly the dynaudio 8" I recently tested was nice, but by no means heroic in its nonlinear distortion.

So, all that to say, I don't know if the RS28 will do what you ask. Horn loading is a complex bit, and would help. How much, I don't know. The only answer is to do the testing for yourself.

I do however, put the RS28 in the upper echelon of tweeters that I have tested, irrespective of price.

[JonMarsh]

Jon,

Defining audible distortion is like what the supreme court justice said about defining pornography. Like him, I don't know how to define it, but I know it when I hear it. I bet you do too. :B

Problem is, that when you cross a tweet as low as in the Arvo, it is carrying a great deal of musical info right where the ear is most sensitive. The trade off between off-axis response, SPL and distortion is a never ending issue at that juncture. I am simply trying to get a handle on how good, or bad, the RS28A is in that part of the spectrum. I have asked Mark K about it. I am awaiting his response. :roll:

BTW, do you have ASCII FR and Z measurements of the RS28A and M8As taken from the Arvo baffle that you will share with me? I would like to model the system in CALSOD so that I can understand it better. :T


I'll try to get the data files uploaded and linked before I take off on my next batch of travel- actually, I'm out on business all week next week, but also have to drive to San Diego and back this weekend- gotta pick up an NEC 10PG CRT projector I bought.

If you REALLY want to get the distortion spectra as low as possible without going to something insane like my RD50's at 1 kHz, a waveguide approach is probably going to be the way. I'm thinking about trying that if I ever get any spare time in the next six months! The Gedlee software support design with that; should cut down on the tweeter excursion quite a bit.

I do understand your concerns; sometimes it just comes down to "what is the lesser of evils?" at a given price point? I hear a lot of improvement in the midband sound quality from the Arvo (compared with enclosure style speakers using the same midwoofers) because the 8's aren't hardly trucking at all, so distortion is pretty low- but could be lower in the upper midband with a better motor, of course.

It's all evolving...

~Jon

[jhatch]

aaarrrrrrrrgggghhhhh! The tc2 is listed as sold out.......what will be the
recommended alternative?
Thanks

[oneoldude]

Mark,

Thanks for you reply. And special thanks for your web site. It is great! Kudos! :T

Pardon me if I was unclear. I did not do the testing on the D28. That was done by Joe D'Appolito. But I did use the tweeter (crossed at 2 kHz) in a high SPL system over 20 years ago. The system was actively crossed and equalized and at times I was pumping 1,600 Watts into it. It would play LOUDLY and cleanly. But it was a bit analytical and sterile. It lacked that certain "j’ne se qua" of a really good system. So I got rid of it long ago. :roll:

I have tried a lot of things since then, including ESLs, Maggies and dipole line sources, but if its not one thing, its another. So I am back to the drawing board looking for a useable tweeter that will not run out of gas on peaks. That is why I have an interest in the RS28A. If it is as good as the best of them at low frequencies, then it can be crossed low, maintain off-axis response and blend in nicely in a dipole as well as other designs. :T

I agree that one or another simple HD test on a single frequency is not the gold standard of quality. But it is an indicator. If the tests have not been done, then they have not been done, irrespective of any standards not employed. No problem. :T

The ultimate test is, "How does it sound when crossed at 1.25 kHz and pushed to 110 dB SPL?" That is pretty much a go no-go test. It either sounds good or it does not. Apparently, no one has pushed the RS28A that far and listened to it. Kind of reminds me of the Medieval philosophers theorizing on how many teeth a horse has and never counting teeth. But since you have made comparison tests and find it to be a top tier tweeter, I guess I will buy some and see if I can count those proverbial teeth with my ears. :B

And thanks again for all the data you have made available to DIY'ers like myself. :T

[oneoldude]

Jon,

Let me know where to get those files when you post them. And thanks for making the data available. It will be fun to see if my old IMP can duplicate your results. :T

I see you can hear the upper midrange fall apart on the Arvo. It seems you attrubute it to the motor in the M8A. How do you know the problem is not the RS28A? After all the RS28A is arguably crossed in the upper midrange and is more likely to be the culprit because it is busting its hump at those low frequencies. Hmmm. Six of one, half dozen of the other. :roll:

I agree that this hobby is a balancing of compromises. You trade one problem off for another. For example, the ribbon line source you contemplate will probably result in very low distortion, at least at moderate levels. But from my experience, if you get any distance away from it, your imaging will result in singers with mouths as big as your listening room. Kinda gives you the impression you are a dentist working on a singer while she is singing. That is not realistic to me and is why I quit line source experiments. The same sort of thing can be said of horn loading a dome tweeter (I guess that is what you mean by a wave guide). There tends to be a loss of clarity with such a technique and you must fiddle even more with the crossover to get the thing flat. But you can get more efficiency and less strain on the tweeter itself. Yup, it is another trade off. Heck, the tweeter low-end strain can be relieved by upping the x-over freq. Then you must deal with more variation in the off-axis response (its gonna happen anyway). Which is worse I wonder? :roll:

FWIW, the improvement to the sound of the M8A in a dipole over a box is less likely due to the effort the motor is making than the use of the dipole method itself. In the midrange, unlike the lowest range of that driver, the driver is working just as hard to accelerate the cone in or out of a box. But without reflections from inside a box, or box resonances flowing through the cone, etc. the dipole sounds better. That's my guess and I'm sticking to it. :B

[Arctophile]

Hello all,

I'm a member of some other DIY audio fora and just came across yours, what an awesome resource!

Anyway, I have my own MTM dipoles based loosely on the phoenix prototype design, and I'm looking to upgrade my drivers (Vifa P22WP and D26NC). I was all ready to order some dayton RS225s and RS28s till I read this thread - can some kind soul please explain in noob language just why these woofers have been rejected for the Arvo Part project?

Would I perhaps be better off sticking with my current woofers (for which I have unfortunately not been able to find either a freq. graph OR distortion figures) and try the dayton tweeters? So far I'm not getting that "ooh, it's so transparent" factor everyone's talking about and I don't have enough money to buy several sets of drivers and test them (that's what this forum's for though right? ;) )

Thanks heaps,

Arcto (Adam F)

[oneoldude]

Hi all,

This is my current dipole system. It was built about four years ago as a proof of concept after my dissatisfaction with dipole line sources. It is the Pheonix panel attached to the Pheonix W woofer. The drivers are what I had on hand at the time. D27TG-45, P17's, and Swan 305's. None of the drivers were optimum for the design then and are less so now with the wealth of excellent drivers available. But they do show the benefit of dipole design, especially in the clarity and naturalness of the mids on down. :clap: Nevertheless, the drivers are not as good as they should be and the system is limited in max SPL. So, its time to move on! :T

I prefer the Arvo and Orion upright woofer designs due to better integration with the mids as well as a direct shot to the listener for the upper bass. The result is a cleaner and more immediate upper bass. :hump:

I use the system for both music and HT purposes. Unfortunately, the drivers in my system are not shielded and must be kept well away from rear projection crt and direct view crt sources. :M The magnets are powerful. :yesnod:

Right now I am leaning in favor of the RS28A, either the RS180 or RS225 for mids :huh: and I don't know what to use for shielded woofers :huh: . My intention is to cross to serious subs (like the SL Thor) that can be moved away from my video sources. That sub does not need to be shielded, but the rest of the drivers do. :roll:

So some questions:

Is the shielded version of the RS28A a drop in for the unshielded version? :huh: Are there any sonic differences between the two drivers? :huh:

Any suggestions for shielded woofers that can bring me down to a frequency that will work with a distant sub? :huh:

Crossover point suggestions for the woof to sub? :huh: How about the mids to woof? :huh:

All help is appreciated. :amen:

[oneoldude]

Oops, forgot to post my system. Here it is:

http://www.htguide.com/forum/attachment.php4?attachmentid=3429

[Evil Twin]

Hello all,

I'm a member of some other DIY audio fora and just came across yours, what an awesome resource!

Anyway, I have my own MTM dipoles based loosely on the phoenix prototype design, and I'm looking to upgrade my drivers (Vifa P22WP and D26NC). I was all ready to order some dayton RS225s and RS28s till I read this thread - can some kind soul please explain in noob language just why these woofers have been rejected for the Arvo Part project?

Would I perhaps be better off sticking with my current woofers (for which I have unfortunately not been able to find either a freq. graph OR distortion figures) and try the dayton tweeters? So far I'm not getting that "ooh, it's so transparent" factor everyone's talking about and I don't have enough money to buy several sets of drivers and test them (that's what this forum's for though right? ;) )

Thanks heaps,

Arcto (Adam F)

The RS225 have a first cone mode that is a dip at 1500 Hz; it's visible as a glitch in the impedance curve, also. This is too low, and combined with the masking of the large sheilded magnet, we don't believe it's an optimum choice for a dipole midrange.

http://audioworx.virtualave.net/RS225NFC.jpg

See that sharp dip at 1.5 kHz? That coincides pretty well with the impedance glitch. It also is very, very similar to Mark K's measured results.

Here's a more conventional 1 meter measurement; since it's not on an IEC baffle, it shows the characteristic drop between 800 Hz and 200 Hz. This is why you need baffle step compensation.

http://audioworx.virtualave.net/RS2251M.jpg



IMO, you want a mid bass driver relatively free of energy storage and dips or peaks (must consider nearfield measurements as diagnostic tool, or shaped sine bursts as Praxis can do). Most 8's don't do all that well. W22 is good. M8a is fair. (costs much less). Waiting on new Peerless to evaluate- and don't have any data on your Vifas, but many other Vifa midwoofers tested have so-so nonlinear distortion- another reason to cross them off the list.

I haven't tested P22WP- a polypropylene 8"- I understand that was used in the Mackie 824 monitor, but not readily available through DIY channels. Poly woofers tend to be a bit "floppy" and have energy storage issues, in spite of what seems like fairly smooth farfield response.

[Arctophile]

Thank you Evil Twin, that explains it perfectly.

I guess even more saving up is in order, as per usual. I hope the Peerless drivers come up trumps - the W22s are simply out of my reach.

I'll be very interested to hear people's opinions on whether the RS28 listening experience matches up to its great measurements.


//Adam F

[Eric S]

Hi Guys,

Just got this from O-Audio about the TC2+ 12" driver:

"Due to substantial supplier price increases we were not planning to carry it in the near future although we continue to negotiate pricing so we may strike a good compromise, too early to tell at this point. Thank you for your interest.

Thank you,

Todd Oiesen
www.oaudio.com"

[Eric S]

I received an update from Todd at O-Audio today:

"...it looks like we will likely carry the TC2+ 10", 12" and 15" versions. Keep an eye on our website for details. Thank you for your interest. Let us know if you have any questions.

Thank you,

Todd Oiesen"

[JonMarsh]

Well, folks, the Peerless Nomex Exclusive Series is officially not vaporware any more. Well, at least that is, not in Livemore. Got four drivers sitting here in the living room... too bad I'm so strung out at work, and busy all weekend. :cry:

The packing by Solen was pretty fair, but the Peerless boxes and driver wraps were pretty thin compared with what I've seen from them previously. As a result, the sides of the inner cartons had the driver edges punching slots through them from the edge of the frame. 8O Overall construction quality is OK, but doesn't look to my eyes as nice as the pictures- no one will confuse these with Seas Excel drivers.

Compared with the sheilded PE RS series, these drivers are pretty "light weight", as they're lacking the wrap around sheilding basket. OTOH, the relatively small looking magnet and nice large open back spider means there's not much rear masking, which should be good for a dipole application. Cone is moderately smooth paper/nomex texture on front, pulpy texture on back. Foam gasket is integral to the back of the frame.

Film at 11.

~Jon

[Interious]

Jon--I don't think I've asked you why feel compelled for replace the M8a. Is there a specific complaint, or are you motivated to find something superior based on the general desire for improvement?
Dave

[Arctophile]

Well I suspect it's the latter, but as someone who can't easily get HiVi drivers (the company confirmed for me that my nearest dealer is Parts Express, which wants US$180 for shipping), I'd be very interested if the Peerless were to turn out to be a good replacement.

Peerless is one of the few quality brands available here at reasonable prices, with other high end 8" drivers (e.g. SS, Seas Excel, Dayton, Focal, Jordan) being ridiculously expensive.


//Adam F

[Interious]

Frankly, I'm unsure of Hi-Vi quality control.

I have 4 - M8a's. One had something rattling in it, like a blob of solder or something. Another has a surround that is smashed sort of flat. I assume this is a PE shipping artifact. 3 of 4 have surrounds that were not glued to the cone in a circular fashion--rather oblong.

But this is anecdotal, of course. This forum has gone through what I gather to be a lot of M8a's and I've not read of problems. Additionally, my luck with drivers lately has been abyssmal.
Dave

[ThomasW]

Frankly, I'm unsure of Hi-Vi quality control.

But this is anecdotal, of course. This forum has gone through what I gather to be a lot of M8a's and I've not read of problems. Additionally, my luck with drivers lately has been abyssmal.
Dave Jon and I have purchased no less than 3 dozen M8a's, and have had no QC issues.

Perhaps it's time that you offered a sacrifice to the speaker gods..... :wink:

3 of 4 have surrounds that were not glued to the cone in a circular fashion--rather oblong If you're referring to 'ooze' (over flow) on the backside of the cone that's pretty typical and hasn't proven to be a problem. If this is on the front of the driver have PE replace them.

[Eric S]

Its good to see this thread moving forward again :T I thought I had inadvertantly killed it with my messages from O-Audio...

Thanks for all of your efforts Jon & Thomas :B Can't wait to see the final product!

Eric

[JonMarsh]

Jon--I don't think I've asked you why feel compelled for replace the M8a. Is there a specific complaint, or are you motivated to find something superior based on the general desire for improvement?
Dave


There's a number of places scattered around in different threads where we've talked about the Peerless Nomex Exclusive Series; the hope is that they offer performance similar to the best of the HDS series with phase plugs. Those drivers are characterized by having both good nonlinear low frequency distortion, and due to the copper Faraday shields in the VC gap, very low inductance modulation and low inductance rise, which translates into markedly lower midrange distortion (say, from 250Hz to 1 kHz).

The phase plug series is even markedly better than the more conventional HDS drivers that I've used like the 850439, which has good LF numbers but just average midrange numbers. The Phase plug HDS parts had distortion numbers up to 1 kHz that would hang in with pretty much anything made, including Excel, but at about half the price.

Take a deep breath and repeat after me: Low distortion is good. Low distortion is good. Low distortion is good. (MarkK, you're excused from this exercise for all the obvioius reasons... ;)

http://www.d-s-t.com/link/peerless/data/foto/830884.jpg



http://www.d-s-t.com/link/peerless/data/freq/830884.gif


These seem to be nicely made for the diopole midrange applicaton- they might work well in other locales, too- I may test in an M8ta. I'm very curious about the 7" version, also.

http://www.d-s-t.com/link/peerless/data/foto/830883.jpg


http://www.d-s-t.com/link/peerless/data/freq/830883.gif

But I don't have any of these on order.



:T :T :T


~Jon

[ThomasW]

I thought I had inadvertantly killed it with my messages from O-Audio... Not a chance, we're customer's of O-Audio, and really like Todd's products.... :T

[Interious]

I'm returning the M8a's for a refund and will wait and see what kind of results Jon has with Nomex cones.

During the interim, I may build George Short's new Pegasus. He totally sucked me in with the advertising copy. In the mean time, I have my 1801b's DCX - paired with my Arvo bass sections and it sounds fantastic. Who would have thunk?

Todd O is first rate. He's tolerated me, which is a feat.
Dave

[Arctophile]

The phase plug series is even markedly better than the more conventional HDS drivers that I've used like the 850439, which has good LF numbers but just average midrange numbers. The Phase plug HDS parts had distortion numbers up to 1 kHz that would hang in with pretty much anything made, including Excel, but at about half the price.

So was there a phase plug 8" before the current model (830884) that you've already tested? If so, how come you decided to wait for the new one?



//Adam F

[JonMarsh]

The whole of the other HDS series were discontinued... phase plug and non-phase plug- now, wouldn't that be useful for you guys, if I offered you a new design to build with drivers I have stockpiled that you can't get?

I've got a stash of 850439's, too.

Could still do that, Thomas has about 4 pair of Focal TC120dx2... yeah, I can see it now, 850439's with TC120dx2 crossed in at about 1800 Hz. Kine problema, as my Deutsche friends would say... ;)

[Arctophile]

Aaah the penny drops - you're talking about 6.5" 850467s (sorry I should have searched). As far as I can determine there was no phase plug 8" HDS until the "Exclusive" series was released.

My local dealer only stocks the 850490 in 8" HDS. Looks as though they have quite a spike at 700Hz, so I guess I'll be waiting with everyone else for the results of some tests on the new Exclusives :)

Cheers,

//Adam F

[JonMarsh]

Yes, you've got it now. The previous various 8" models from Peerless have had their first "problem" around 700-800 Hz, so we'll have to see how these turn out. May be the the 7's are the ones to go with, but I wanted to give these a shot first.

And even if I don't wind up using these in the Arvo, or the 7" version, the 8's are a very good candidate for the midwoofer array in the Saint-Saens, due to the expected low distortion, smooth response beyond 800 Hz, and light weight (ever moved a baffle with a line array of 8" woofers around? Even with small magents like an Eton 872 it's a bear!) (Can you imagine what 8 RS225 would weigh?!?!?!?)

~Jon

[JonMarsh]

He told me on the phone a week or so ago that he'd started laying up the panels for the main baffles, but I dont' know how much further he's gotten than that, or what his timeline is. I'll take some spy pictures when I'm out the week after next if nothing else... ;)

[ThomasW]

I'm working on finish carpentry in the kitchen and occasionally working on the Arvos.

So far the front baffles are laminated up. 2 layers of 3/4" BB ply with a cover layer of 1/4" hardboard. And man are they heavy 8O

I'll post a pic later today....

[mante]

Hi Jon and Thomas,

Jon, I hope that is a vacation coming in a few weeks. Will you take the new Peerless drivers with you?

Thomas, thank you for the quick response. Two-inch thick baffles that size will be heavy. Did you just lower the MTM by a few inches from the original Arvo blueprint in order to lower the T to a better-seated listening height? Also, did the dimensioned front baffle and side drawings from twelve months ago get changed.

Looking forward to seeing the progress.

Thank you,

Craig

[ThomasW]

Craig,

I'm working on 3 completely different projects, all involve laminated BB ply. So at this point the baffles are just glued up blanks 60" tall.

I plan on lowering the tweeter, and may do some other things such as round over the top, and have the baffle go all the way to the floor (meaning no visible front plinth). None of this is set in stone because this coming week the 'job' is finish carpentry in the remodeled kitchen (only a couple years overdue 8O ).

Jon will be in Denver for a week starting Aug 15th, because I was able to snag 2rd row seats for the "Trio" concert featuring Stanley Clarke, Jean-Luc Ponty, and Bela Fleck. I have to use high powered 'bait' like this to get him to take some time off..... :wink:

[Jonasz]

Would it demand a complete redesign of the crossover if you would want to use the W22 in place of the M8a? W22 and RS28 sounds like a mouthwathering combo... :P

Jonas

[JonMarsh]

It would require some adjustment of the low pass for the midwoofer, due to slight differences in impedance in the transistion region. Not really a biggie, just takes bucks...

[Jonasz]

If I understand you correct one could use the existing crossover (mids/tweeter) without any changes? Would the W22 perform better 175-1250 hz than the M8a (without regarding to cost oc ;) )?

Jonas

[JonMarsh]

The W18 has lower midband distortion; it certainly ought to for 3X the cost! ;)

I'm hoping that the 830884 is similar in performance, with regards to mid band performance, but at 1/2 the cost of the W18. If it follows in the footsteps of the 850467, that may be the case.

~Jon

[carlb]

Long time lurker...

Anyways a quick comment. Lot's of people doing laminations of either BB or similar, ie appleply and sometimes in combination with MDF. Has anyone had any experience doing these laminations with green glue?

I have no affiliation whatsoever with audio alloy but they have a viscoelastic damping glue that would seem to be a natural for producing a far superior resulting panel with far less ringing. Would seem to be a very good fit for producing exceptional speaker cabinets.

I have a future project for doing a HT room and this product has recieved a lot of attention in some of the HT or listening room forums and thought I'd bring it to attention here.

In any event, if someone takes it upon themselves to experiment it would be interesting to hear their comments.

Maybe I should have put this in Graham's Extreme MT Design and Construction thread but he already has put in way too much effort on his cement filled cabs... I bet green glue if what they state in their website is true, would have been potentially a better route... (then again maybe not - pure speculation on my part.)

Carl

[GrahamT]

Long time lurker...

Maybe I should have put this in Graham's Extreme MT Design and Construction thread but he already has put in way too much effort on his cement filled cabs... I bet green glue if what they state in their website is true, would have been potentially a better route... (then again maybe not - pure speculation on my part.)

Carl

Welcome Carl,

A big reason that I went with cement is that it was different, and free. The latter being a big factor for a poor student like me. I like the brute force technique too.

The green glue is very spendy, and I'm a luddite like many others here. :W

[AndrewM]

I have no affiliation whatsoever with audio alloy but they have a viscoelastic damping glue that would seem to be a natural for producing a far superior resulting panel with far less ringing. Would seem to be a very good fit for producing exceptional speaker cabinets.

I don't know anything about the product other than hearing about it from time to time, but I have to wonder what would give this product such a "superior resulting panel" over something like regular wood glue, or if you want to get fancy (and stoned) good 'ol fashioned contact cement.

[carlb]

Hi Andrew,

Check out the comparative section... it is if true it's pretty impressive.

Like I said, just curious in that on paper (even if virtual paper) it is interesting.

A small cabinet could probably be done with just a tube or two so any of you with an active or about to be started project wants to try and report it would be cool... If it makes a substantial improvement and you have to laminate any it would be relatively cheap for the advertised improvement it makes.

Just a long time lurker looking to contribute... :-)

Carl

[AndrewM]

Check out the comparative section... it is if true it's pretty impressive.

Yea, I saw those, but I didn't really expect to find charts showing that their stuff made no difference at all ;)

A small cabinet could probably be done with just a tube or two so any of you with an active or about to be started project wants to try and report it would be cool... If it makes a substantial improvement and you have to laminate any it would be relatively cheap for the advertised improvement it makes.

Considering that a small cabinet will have less of a problem with panel resonance than a larger cabinet I wonder how much of a difference it would make even if this stuff does what it says it does. I helped a friend build a project with some smaller 5"'ish 2-ways, built with 3/4" MDF and the panels don't vibrate at all even at upper SPL levels by using the old hand test.

Contrast that with the first sub box I built with a Tempest and a whopping 250W, single layer 3/4" MDF with no panel span bigger than 8-10" or so, and stuff would walk across the top of it when the sub was going and gave a nice soothing vabra-massage if you left your hand on it.

[JonMarsh]

Yea, I saw those, but I didn't really expect to find charts showing that their stuff made no difference at all ;)



Considering that a small cabinet will have less of a problem with panel resonance than a larger cabinet I wonder how much of a difference it would make even if this stuff does what it says it does. I helped a friend build a project with some smaller 5"'ish 2-ways, built with 3/4" MDF and the panels don't vibrate at all even at upper SPL levels by using the old hand test.

Contrast that with the first sub box I built with a Tempest and a whopping 250W, single layer 3/4" MDF with no panel span bigger than 8-10" or so, and stuff would walk across the top of it when the sub was going and gave a nice soothing vabra-massage if you left your hand on it.


Part of that may be due not just to panel resonances, but to unbalanced forces; i.e., for every action, there is an equal and opposite reaction. I think "enclosed" subs with balanced force (dual drivers firing opposite directions) are interesting and preferrable for that reason.

OTOH, I'm pretty much either a Titebond II or Epoxy kind of guy myself. Especially the latter for cabinet and panel construction, the former for veneering.

~Jon

[ThomasW]

I recently 'glued' up the front baffle sections for my Arvo design. I laminated two layers of 3/4" BB ply using contact cement (a recommendation from a project in ShopNotes magazine). There is an obvious amount of damping occuring between the layers compared to my previous technique of using standard yellow glue. (The effectivenss of the damping was validated by using the industry standard 'knuckle rap test' .....:wink: )

Contact cement is quite a bit cheaper than Green Glue, and IMO probably a better choice for loudspeaker construction

[Paul H]

Thanks Thomas, I'll try that next time. For some reason it never occurred to me to use contact cement with 2 layers of ply, but it makes sense.

Paul

[carlb]

It would be interesting to have 2 baffles of exactly the same size, one with contact cement the other with green glue to then be put under the industry standard, brutal, "knuckle rap test".... and see if spending a few extra bucks is worthwhile...

Hey, can't argue with whatever works.... just seemed like an interesting use of a product that is being heavily recommended by many highly respected accoustic room designers. A room afterall is but one large speaker enclosure...

Carl

[Dennis H]

That Green Glue does look interesting. If you believe the company's graphs, it has much better damping than conventional vinyl construction adhesive. Spendy stuff though. $177/case, Ouch!

http://www.audioalloy.com/c06-00_comparision.html

About Thomas's comment on contact cement being a good sound damper, I didn't take it too seriously at first but I changed my mind. The material sheet on Green Glue indicates it's a latex (e.g. rubber) polymer. Well, heck, so is contact cement, more or less. So, if you slathered on some really thick coats of contact cement, it might work pretty well.

Another one I've used in construction is a urethane based adhesive caulk that remains flexible. Sorbothane is urethane and it's about the best sound damper around.

[Eric S]

Hi Thomas,

Since I don't have much experience with very many "glues", can you please elaborate on "contact cement". Is it a spray, or does it come in a can so it can be brushed on. Are there many differences among brands, or are different varieites of the stuff relatvely the same in terms of performance?

thanks,
Eric

[ThomasW]

This is a process where you should have someone experienced show you how it's done prior to attempting it alone. For larger pieces it's mandatory to have a second pair of hands......

Sold by the quart, gallon, and 5 gallon buckets.

I apply it with a roller, use a tray liner then throw everything away. The H20 soluble works, but doesn't hold as well as the solvent based stuff which is VERY TOXIC USE OUTDOORS with a respirator. It can be sprayed if you have a gun and compressor. Spaying necessitates use of a fullblown fresh air supplying respirator and massive amounts of ventilation since the aerosolized cement can explode.

You put a coat on both pieces of material and wait for it to dry, then apply a second coat and again wait for it to dry. The carefully place the 2 coated surfaces together. AS SOON as the pieces touch THEY ARE GLUED and CAN'T be repositioned. The trick is to make them over sized and cut them to size after laminating.

I've found most brands to be similar, but since our fellow HTG'er Hank works for 3M, I use their Fastbond 5 when possible...:wink:

You should have some Acetone around to clean up hands and spills.

For very small jobs 3M makes "Super 77" which is contact cement in a spray can. It's very spendy compared to buying in quarts...

[Eric S]

Thanks for the details, Thomas. Much appreciated!

Eric

[mante]

Hello Jon and Thomas,

Time for me to build the Arvos. Anything else I should know before I start?

Your trusted student and young Padawan.

Craig
p.s. I do not have any testing equipment........ ;x( and am flying blind :boohoo:

[ThomasW]

Jon's flying to Denver monday, then we'll discuss some ideas I have about some minor cosmetic modifications (they're detailed a few posts above this one).

Other than that nothing has changed if you're building the version using the M8a drivers, so no test equipment is necessary

[mante]

Hello Thomas,

If you are finishing up on your kitchen project and are gearing up to proceed with your Arvos, I'll wait, watch, and take notes. After Dave's comment about driver height, I am interested in what you will do.

I was planning to use the new Peerless driver. Madisound is five minutes away, and someone has to try them. 8O I will use my DCX until the final crossovers are completed, unless Dave says no. Oh, why birch and not SierraPine Medite MDF?

Also, I've asked Jon to ask you if you have any extra drivers available for my project.

Thanks

Craig

[ThomasW]

Oh, why birch and not SierraPine Medite MDF?Strength.... :wink:

[Aaron M.]

From what I've read so far, it sounds like the lower midrange and upper bass are the things most positively effected by dipole radiation. If that's so, here's my idea:

Get the Behringer Ultradrive Pro digital XO, and four to eight big mutha woofers so as to build a U-frame woofer setup that doubles as a speaker stand...not exactly revolutionary, but It might give me a taste of dipole goodness without having to build a whole new system.

So what say you, is the TC2+ 12" the only high excursion beast with descent midrange qualities?
I'm hoping the Dayton RS subs will meet these requirements.
Other woofers I have questions about:

Vifa ma26wr09-04: A ten inch alu cone with 23mm P-P X-max...And look at that things frequency response! (check it out at Madisuond) I'm awfully curious about this one 'cause it looks like I could cross it over higher than the rest and therefore get more dipole effect.

Peerless XLS: I have the ten incher in a PR box and I'm pleased, I also noticed it makes very little air noise from the back and that's always handy for dipoles. I don't know jack about it's midrange, however.

I curious about all less than $200 subs (10"-12") for use in passive radiator boxes and dipoles, so any other recomendations or comments are welcome.

[JonMarsh]

The TC+ is the current champ for a long travel driver as regards upper range distortion and LF non-linear distortion, besting even the Peerless XLS.

BUT, the new Dayton RS sub drivers, not yeat shipping, are it's match and then some in the upper regions. I hope to be able to use the 10's up to 450 Hz or so. Until I get final production parts to test in Praxis, I can't say. The proto's looked very good, very low non-linear distortion in the LF region and upper bass.

I find the advantages of dipole operation to hold up and be worthwhile up to 1200-1500 Hz. We've done systems with just lower range dipole (say, 40 Hz to 300 Hz, like the Whisper Klones), and that's quite good, but not as good going dipole up to the 1 kHz area.

Just my 0.02. YMMV, all the usual disclaimers. ;)

~jon

[Aaron M.]

Do you have any measurement info on the Vifa unit?
I mention it 'cause the freq. response graph is so impressive, the first sign of cone breakup occurs at 2.5KHz or so.
You could cross it pretty high, like 1.2KHz if you can get good driver spacing.
If you really think it sucks, then I'm glad I asked before I got myself in even more trouble.

But let's assume it doesn't suck for a minute: maybe a guy could make a two way dipole (MTM) on a large L-frame panel that could reach 95-100db (at one meter) down to 70Hz or so.

Possible pluses: Simple 2-way X/O, fewer drivers for less cost...

It's a compromise, that's for sure, but I'm trying to think of a way to get maximum dipole effect with very little fuss.

Am I missing anything? (probably) Am I in for an aural assault? eek!

[Paul H]

Aaaron,

I don't have any info specifically on that vifa, but can note a few rules of thumb. Any 10" will start beaming above 1300Hz, so that would be ok. However, if you wanted to go to 1200 Hz crossover, max recommended spacing of drivers should be around 11" - pretty tough to get with 10" drivers in a mtm. To get a 13" centre-centre spacing, a good target for 10" mtm's, the crossover should be about a 1000Hz - which is hugely stressful on almost any tweeter. This thought process leads to using 8" for mtm's, as Jon has done in a couple of designs, which provides a good balance of the most bass, least beaming/breakup balanced with acceptable stress on the tweeter (if it's a good low-freq tweeter).

I'm building a large full-range dipole noted here in another thread, and have therefore done much careful listening to sub, bass and mid dipoles recently. If I was in your shoes and wanted to build dipoles, possibly in stages, I'd start with the bass, just as you suggested a couple of posts ago.

My 2 cents,

Paul

[Aaron M.]

Thanks for the input, mind if I throw some more ideas at you?

Check this out, this wave guide seems to damp dome movement around the X/O frequency.

http://www.zaphaudio.com/hornconversion.html

Line source idea:
Also, what about Newform ribbons? Everyone talks about the BGs but nothing on these guys.

http://www.newformresearch.com/bare-ribbons.htm

With a pair of the 45" ribbons and twelve of the Vifa tens, a guy could make a fairly slender, elegant 5' tower with good output to 60Hz or so all for an estimated $3000.
If I had the skills and equipment Jon has, I'd pobably just get some samples, test 'em and report my findings here. (sigh, someday I'll figure out how Soundeasy works...)

When I visuallize my idea, it looks really great, but one or both of the drivers I mention could be totally unsuitable in some way (nasty distortion and all that).
I really hope someone has had experience with these babies...think I should start another thread on the subject?

[Dennis H]

Until someone does some distortion testing on the Vifa 10", I'd take that frequency response chart with a large grain of salt. Even an unsmoothed impedance plot would tell us a lot but we don't have one. That driver isn't on the Vifa page and the measurement isn't the standard factory one so we really don't know much about how it was done and how much smoothing was applied. As it is, the plot is "too good" to be real for a big metal cone. Looks like they smoothed the begeezus out of it possibly hiding a lot of bad behavior.

[Aaron M.]

If the distortion tests yield good results, I still have hope that it can be crossed at 1K with a 4th order transfer function or better...I should pester Madisound adout those graphs, though.

[ThomasW]

Line source idea:
Also, what about Newform ribbons? Everyone talks about the BGs but nothing on these guys. We have talked about them.

1) Monopole
2) XO point, 1kHz or higher
3) Price, R45's are $1230/pr excluding shipping

So they don't really fit our needs/wants

Check this out, this wave guide seems to damp dome movement around the X/O frequency. Yep covered wave guides a while ago.
http://www.htguide.com/forum/showthread.php4?t=13074&highlight=wave+guides
http://www.htguide.com/forum/showthread.php4?t=14232&highlight=wave+guides

[JoshK]

I got my Seas W22s today. Wow, they look much cooler in person that they do in photos. They aren't cheaply made to my eyes.

I have been kicking around the idea of using the MCM shallow waveguides on the RS28a in my Arvo Pärts. That would mean they are drifting further and further away from Jon's design. (I stuck 100% to Jon's recipe on the modula mtms, so I am not going to blame jon's ears if I blow this one myself).

The problem I foresee with using the waveguides is they are 6.5" in diameter. This means that even with the tweeter center set 33º off the vertical line between mids the c-t-c distance between mids is 13". => 1000khz x/o point. Would the RS28a handle this? probably not I assume.

[Interious]

Josh--Do your copper phase plugs appear to be properly centered in the VC?
Dave

[JoshK]

Dave,

I only looked at two briefly and I would say it seems so. Why, has there been an issue with this?

[JonMarsh]

From what I've read so far, it sounds like the lower midrange and upper bass are the things most positively effected by dipole radiation. If that's so, here's my idea:



Other woofers I have questions about:

Vifa ma26wr09-04: A ten inch alu cone with 23mm P-P X-max...And look at that things frequency response! (check it out at Madisuond) I'm awfully curious about this one 'cause it looks like I could cross it over higher than the rest and therefore get more dipole effect.



As Dennis points out, the FR chart does seem to be rather smoothed, and doesn't show anything below 100 Hz. Curious for a "sub" driver, since on a moderate size open baffle you can get pretty good FR data with a 6" close mike measurement that rejects most of the room stuff due to precedence.

Also, it's odd that the impedance curve isn't published; this is a very useful item for evaluating driver behavior, and Madisound is certainly smart enough to know that.

Now, for a dipole application, this part looks like it could have potential. The Qts is fairly high, which is a good thing. OTOH, the Xmax is rather low, at 11 mm Pk, especially considering that the sensitivity is rated quite on the low side, 83.3 dB/2.83 VRMS, NOT 83.3 dB/1Watt - remember, this is a four ohm driver, so it's seeing two watts. So, nominal 1Watt sensitivity is ~80 dB. Wasn't there someone asking a question this week about how much power you need to drive a dipole woofer? This little puppy will take more than average.

On the plus side for dipole applciation, it has a small magnet and relatively open basket. On the minus side, we have no idea about the rear venting or chuffing. Small magnet often means smaller pole piece hole. No sheilding, which doesn't much matter for dipoles, but might be a problem in more conventional applications.

In a series/parallel line array, their might be hope to get the sensitivity up to perhaps 86-88 db best case; this doesn't leave much to work with for a passive crossover. The RD's are 88 dB, and I don't plan to give any of that sensitivity away.

Regrettably, there's no data to be had online from Vifa for this driver.

I'm not sure I'd be willing to plop down my own nickels to meaure one of these later this fall, but who knows...

~Jon

[Interious]

Josh,
I've got a pair of W18s. Their phase plugs are not centered with respect to the VC former. This does not appear to affect performance. I just wondered if this anomaly appears in the W22. Not a dig deal.
Dave

[Asterduc]

Until someone does some distortion testing on the Vifa 10", I'd take that frequency response chart with a large grain of salt. Even an unsmoothed impedance plot would tell us a lot but we don't have one. That driver isn't on the Vifa page and the measurement isn't the standard factory one so we really don't know much about how it was done and how much smoothing was applied. As it is, the plot is "too good" to be real for a big metal cone. Looks like they smoothed the begeezus out of it possibly hiding a lot of bad behavior.

For what it is worth:
I bought 2 of them last week and I am just measuring them. Measuring at 10 mm cone distance (in a 30 liter closed box) the curve is flat from 25 up to 350 Hz with a first resonance at 2830 Hz. So yes, the curve is real. I wish I could show the curve but I can't get the curve out of my clio that runs on an old laptop / docking station and bring it into my internet PC. Attached is a .txt file with the measured data.
Distortion shows approx. around -45 and -65db. Only below 100 Hz it gets to -20 db.

[jdybnis]

I see a lot a perturbations in the impedance from 40-150Hz. That's probably cone breakup. It might be a measurement artifact. Can you do a higher resolution impedance measurement? Something like 1/24 octave.

[Dennis H]

Thanks, Asterduc! I plotted your data in Excel and it does look pretty good. It would be nice to see more resolution. 1/6 octave is still a bit coarse.

[Dennis H]

For what it is worth:
I bought 2 of them last week and I am just measuring them. Measuring at 10 mm cone distance (in a 30 liter closed box) the curve is flat from 25 up to 350 Hz with a first resonance at 2830 Hz. So yes, the curve is real. I wish I could show the curve but I can't get the curve out of my clio that runs on an old laptop / docking station and bring it into my internet PC. Attached is a .txt file with the measured data.
Distortion shows approx. around -45 and -65db. Only below 100 Hz it gets to -20 db.

Okay, I read that a little more closely. A nearfield 90dB signal will be pretty low SPL in the far field. Even at that low level, the distortion goes up to -20dB (10%) below 100 Hz. That implies that not much of the published Xmax is actually useable. So, if low distortion is of interest, I don't see any particular advantage over the Seas L26. The L26 will probably perform better in the real world and it's $10 cheaper. Another thing is the Vifa Mms is 3x that of the Seas. That will give you big vibrations on an open baffle. Me, I'd buy the Seas.

[Asterduc]

Thanks, Asterduc! I plotted your data in Excel and it does look pretty good. It would be nice to see more resolution. 1/6 octave is still a bit coarse.

and here you go.

VFMA26FQ: near field with stronger signal, 24 points / octave.

VFMA26I2: this is the impedance curve of the other driver in free air. The strange behavior between 40 and 50 hz is still there.

VFMA26ML: this is a MLSSA measurement taken at 1 meter (measurement not calibrated. Efficiency is much lower as seen on these figures. Guess will be in the area of 85 db.

I tried a distortion graph too, but my Clio 3. does not allow me to export the data. Maybe someone can tell me how to do so.

The type description on the unit I measured is MA26WR09. I noticed that these drivers are also found on the German market with slightly modified parameters http://www.lautsprechershop.de/hifi/index.htm?/hifi/vergl3hh.htm

Ed.

[Dennis H]

Pasted the new numbers into the Excel chart.

[JonMarsh]

VERY odd looking impedance curve- could be a signe of box wall vibration. The upper range impedance curve is quite clean, 100 Hz to 1 kHz.

I've gotten to where I do mostly dipole style measurements on a 24" wide baffle, at about 6" distance nearfield; you get impedance curves that only reflect what the driver is doing, and FR being fairly nearfield tends to swamp out room and other issues.

What the heck- when I order Praxis later in September, maybe I'll order one or two of these to meaure, too. Distortion would be very interesting to see...

BTW, in the DOS version of CLIO4, you can print to a BMP file on disk; this is the easiest way to get graphs out of the system. It was only with the most recent 6.5 update of CLIO WIN that I've been able to stop using the DOS version.

~Jon

[Dennis H]

Jon, I think the latest impedance curve is in free air while the FR is in the box. Strange looking beasty.