Vifa XT25 Tweeter

**AndrewM** · 26 March 2002, 11:08 Tuesday

So are we going to see some Krell LAT-1 klone's over at Klone-Audio?

Andrew

**JonMarsh** · 26 March 2002, 11:46 Tuesday

How "Klone" do you think we should go, Andrew?

I've worked with cutting aluminum on a saw before; it's not something I really enjoy doing! :roll:

Also, I'm not as much into the Scanspeak paper drivers as some are. OTOH, we could do something with a combination of 8" and 6" aluminum magnesium drivers from HiVi, and put them in a sonotube enclosure

sort of reverse where we use the metal and the paper, huh?

Well, I dunno about Kloning a Krell- Thomas kind of wimped out on his MRS Klone project, but made a very nice Sonotube sub instead. I've got too many speaker projects in the air as it is- look for another announcement this weekend from a new division of those fine folks at AudioWors/Klone Audio.

And I'm still tweaking up the MkII M8 slightly. (It's really "lush" on natural sounding vocal tracks- I should rename it the J-Bro Model 1) (yeah, too much listening to Jonatha Brooke lately- another obsession, I'm afraid).

And the M8 MKIV has parts ordered, and a preliminary crossover design nearly complete.

And there's an MP DAC to finish building for ThomasW, and another for Dennis.

And if I don't get off my butt and finish building the first test board for the AragonX project, Thomas will probably make me sleep under an overpass when I come out to visit in May.....

Can you spell, "buried alive"?

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 26 March 2002, 12:54 Tuesday

hey, no complaints from me. it looks very promising

p.s. do you do im and harmonic distortion tests on raw drivers?

i.e. what is your testing philosphy on distortion?

**JonMarsh** · 26 March 2002, 16:15 Tuesday

Hi Mark, you scoundrel!

You probably knew what you were doing all along, suckering me into these with a "little taste"! :B

Yes, I do distortion tests on raw drivers, but not at meaningful levels on ones I don't own! So, today my own order went off for a pair of these to play with.

Tonight I'll send the data files I've got so far to you.

I'm probably going to wait on the new mic, and the preamp I'm building to do the distortion measurements. My 20 year old 4133 is getting a bit glitchy; I'm not adverse to reworking the electronics, but I need to have something else while doing that.

But I have to admit to being impressed with how these perform overall, and what they sound like with just the MLS test signal. From my perspective, I wonder how low they'll go with a steep crossover- considering what people get away with for conventional networks at 2.5 kHz, I'm expecting to pull off one of my 48 dB/octave elliptic networks at 1.25 kHz without problems. Then, the meaningful test might be a multitone test in a finished network (woofer could be dummy load), with 800 or 1 kHz, plus some in the passband of the tweeter. The trick is spacing the test frequencies so you can look for difference tones and harmonics. Something like Dean Jenson's spectral contamination test, and like the tests Sigfried Linkwitz has been doing on woofers.

ThomasW wants me to design an MTM around the M8 concept; it may be more practical with this tweeter than any other I've looked at so far. It would be a sealed box system, designed to be used with a sub or LF cabinet. (could be a top module replacement for my X1 klones). I'm thinking of another project though- might be too close to what SL does, as it's a mult-way dipole.

Best regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 26 March 2002, 17:24 Tuesday

Originally posted by JonMarsh

Hi Mark, you scoundrel!

You probably knew what you were doing all along, suckering me into these with a "little taste"! :B

perhaps

i was also very curious to see what your feelings are on lowish xover points for this driver. i think we should be trying each others' drivers more often. the reality is, you can't buy every interesting driver on the market, so what does everyone do-they read everyone elses opinion online. while this is pretty good, still it seems that often, everyone just repeats what someone else said. this driver has achieved a "don't cross it too low rap" on most forums, and not much objective data why.

anyway it's time to get off my soapbox. i'm glad to see you think enough of them to buy yourself a pair. i'm looking forward to my project with them. i haven't actually decided what that project would be, but therein lies the enjoyment.

i'm at home today-i've been sidetracked by pneumonia, believe it or not. anyway, it's not much of a pneumonia, just a dinky subsegmental right lower lobe infiltrate. i will be back to work soon enough.

**JonMarsh** · 26 March 2002, 22:40 Tuesday

Well, I wouldn't call *any* pneumonia dinky; my mom and sister used to have frequent problems with that, though I've been blessed with a fairly robust constitution compared with anyone else in the immediate family (good thing, considering some of the silly things I've done like motorcycle racing, and all the years living in Colorado). I hope you're feeling better soon.

Regarding crossover points, I tend to take anecdotal comments by other DIY types with a grain of salt (Yankee scepticism?). Too many of the designs seem to have odd priorities- you know, "let's buy a sixty or eighty dollar driver", then skimp on Fc zobels, padding network ("just use a series resistor, it only costs 1/2 as much"). There are a lot of things done in the name of cost reduction in commercial speaker designs, but then their goal isn't extending the boundaries of driver performance.

We'll see how the XT25 fares- I've got a design "on paper" now, (well, actually in one's and zero's on the PC) for 1.25 kHz. I've run the MB Quart MBTTR1 as low as 1 kHz with just "conventional" crossovers with reasonable results (so has Avalon Acoustics

), so I think this is doable. How it sounds and measures? Well, we'll see. With my elliptic filter, I don't really push the driver on cone excursion with a crossover that low- it's less stressful than a conventional 3rd order at 2.5 kHz.

I've got another friend of a friend who's sending some OW1's for me to look at next- he's just finished building a set of the M8's with the help of ThomasW. I've heard good things about them; I'm curious to see them also, though I'm not normally a "soft dome" kind of guy.

Take care of yourself, and get well!

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 26 March 2002, 23:50 Tuesday

(edited 3/27 at 4pm--the deal's over)

fyi, if anyone else is thinking of buying a pair of these they are the 3/27 deal of the day at PE and are 49 a piece. the deal ends at 3 pm tomorrow pacific time.

i'm going to see if they pricematch

(edited again) PE in fact graciously pricematched my recent order, so kudos to them for that!

**Patrick Sun** · 30 March 2002, 07:50 Saturday

I'm going to find your "crossing low" experiment findings very interesting because I know some of the advocates for not crossing the XT25 low (most will stay in the "3KHZ and higher" camp) who say this from simply listening to them crossed low. People have tried the resonance traps/et al and found that it kills the tweeter's dynamics.

Just because it has a ultra low resonance frequency doesn't always mean it's suitable for being crossed low. IIRC, the Seas 27TFFC's are more of the indestructible tweeters that can handle being crossed low.

I've heard the XT25 in a 3-way (crossed higher in the 3.5KHz range) and it true of its reputation: it disappears, nary ever drawing attention to it. This was a design done by an in-town friend with the full LEAP/MLS system.

But I am always open to new ideas that buck the conventional wisdom. Good luck!

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**JonMarsh** · 30 March 2002, 11:04 Saturday

Hi Pat,

Do you have any feedback or information about what kind of crossover topology they were using, re comments about peformance of the XT25 with crossover points lower than 2.5 - 3 kHz? What midwoofer/midrange were they using up that high? Did they make any effort to find a midwoofer with pistonic operation to the upper passband?

I'm also a little curious about what people are hearing, and what they mean when they say that an impedance matching zobel "kills the tweeters dynamics". Obviously they're hearing something different, but is it really wrong or right? In my experience over the years, there isn't a tweeter made that doesn't perform more "accurately" with a conjugate matching network than without. Only if the impedance the crossover network sees is frequency linear, will the network perform as expected. The attenuation curve in the driver Fs region can't be accurate (i.e., it actually looks rather "lumpy"), for a tweeter with a relatively high undamped resonance at Fs. (The XT25 fits that bill- look at it's curve compared with a Focal TC120, for example.) Hitting the tweeter with extra energy at the Fs will be more likely in my experience to push voice coil to higher excursions- which all things being the same, will increase THD and IM. I've found that makes it sound a little harder and brighter. The question this raises is how good a reference any of us has for determining system voicing- depending on the associated behavior of the rest of the system, and the expectations of the user, different "affects" may be desired. That's one reason looking closely for driver behaviors apart from linear frequency response that influence the perception of tonal balance is so important. Sigfriend Linkwitz is one of the few, if only, people discussing those kinds of driver behavior in detail on a web site- most people don't want to give that kind of information away, if they have it.

I've got two weeks of business travel coming up, and after that I want to do some burst sine testing and look at ETC behavior at different frequencies, as well as actually build them into a network and listen to them. I'm intrigued by this part for running a little lower than average, because it does have a lot of radiating area compared with a conventional 1" dome.

So, we'll just have to see what we see. I've been able to successfully use the Focal TC120 at a 1.5 kHz crossover point with this new network design- (keeping in mind that it's in a speaker not intended for the dynamic range of something like an X1 SLAMM

) -that's what the M8 speakers documented in other threads on the DIY forum use in the MkIII incarnation. ThomasW and a friend of his have built these now, too, and they seem pretty pleased for a parts cost of roughly $600 for a pair.

While the ultimate proof is in the listening, I also like to look into the driver behavior as closely as possible to see if I can understand and correlate the two. I've got a new quarter inch mic on order to help with that!

Sure wish I didn't have this business travel coming up...

Regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Patrick Sun** · 30 March 2002, 15:13 Saturday

Jon,

Much of the stuff I've read is off the Madisound and the PE Tech boards. There's a guy named Ken Perkins who's big into series XO's (loves the old fashion trial-n-error method of voicing a speaker), and he built what he calls "The Nebula" which is a XT25 and PL18 (Vifa's) 2-way with the new-fangled Acoustic Reality series XO (but it's sort of all-mucked up with an inductor tossed into the mix on the midwoofer side. Let's see if I can find the link...

Link 1

Here's a PE link to the introduction of his project:

Link 2

You'll need some sort of Yahoo ID to view all the stuff in there. But he and John K worked on the XT25 and came up with the final design. And they gave up crossing it under 2.5KHz IIRC.

Ken's the kind of guy who'll try to cross any tweeter he gets his hands on as low as possible, but as this year progressed, he's leaning more to just getting the midrange "right" and all the other pieces will fall into place, so he not as agressive in crossing as insanely as possible anymore.

I met Ken at the Atlanta DIY2001 last year. He's in the boonies (near Columbus GA), so he loves listening to music out in the sticks, and has plenty of time to build/design speakers.

Here's John K's home page (if you've never encountered him before):

Link 3

This man loves speaker testing/designing (but isn't the world's greatest proofreader).

And then there's PE Paul, one of the moderators on the Parts Express Tech board who also gave up crossing this XT25 tweeter low.

Link 4

I'm not sure if WayneJ (aka "The Rev") has messed around with the XT25. You can find his stuff at:

Link 5

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**Mark K** · 30 March 2002, 16:35 Saturday

hey, i don't mean to make a stink about this low frequency issue.

i just haven't seen any objective measurements to make me think that it needs higher xovers, though it is certainly possible. initially, i thought the published curves might be suspect, but it doesn't appear so from the jon's measurements.

i'm assuming that these drivers are similar to true ring radiators a la jbl 076J(more or less, obviously, the xt25 isn't a horn). the diaphragm is clamped in the middle as well as the sides, forming an annular radiating surface.

generally, these ring radiators are very high frequency drivers, so it's possible that, for some reason not measured yet, the xt25's don't do well lower, just like other ring radiators. maybe there is some sort of distortion issue or?

hence my interest. all the other posters may yet be right.

**Patrick Sun** · 30 March 2002, 18:36 Saturday

If those guys could have crossed the XT25 low, and been happy with the sound quality, you bet they'd be tauting this development. As it stands, their ears tell them that the XT25 is happier being crossed around 3KHz or higher. They got the fancy gear and the XO sim software (LEAP/LspCAD) to guide them (and they know how to use these tools).

There were a lot of "disappointed" DIYers when the XT25 didn't emerge was the go-to tweeter for extreme low XO points.

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**JonMarsh** · 30 March 2002, 19:41 Saturday

Hi Pat,

I'm familiar with John K and Ken Perkins, though I don't frequent the PE boards.

For anyone using the Acoustic Reality crossover, I'd certainly agree that 2.5 kHz is probably the lowest reasonable range. This crossover gives slopes of about 6 dB/octave. The LC Audio series crossover gives about 9 dB/octave. Frankly, at 6 dB/octave, I'd be inclined to go with 3 to 3.5 kHz, except for the difficulties of getting a midwoofer to work that high cleanly. With a 6dB/octave network at 2.5 kHz, the attenuation at 1.25 kHz is about 9 dB, and at 600 Hz about 14 dB. I'll certainly agree that's pushing that driver pretty hard in the low end.

On the other hand, with a standard L-R 24 dB/octave crossover at 2500 Hz, the attenuation at at 1.25 kHz is close to 30 dB, and about 54 dB at 600 Hz, using a 2500 Hz LR network. For the 4th order L-R, my feel is that the lowest usable frequency might be in the range of 1.5 kHz, where it would be down about 28-30 dB at 600 Hz. Not a level I would be especially comfortable with, but actually much better than the 6 dB/octave network's attenuation with a 2.5 kHz crossover.

For my high slope network network (which achieves results similar to the Joeseph Audio networks, but using a different technique to achieve it) the attenuation at 600 Hz is about 40 dB using a crossover point of 1.25 kHz. Power handling and distortion in the passband, tested on an accuton C23, is quite good- very similar to a 4th order L-R at 2 kHz. The notch in the transfer function for the high pass is tuned to the tweeter resonance, for the low pass it's tuned to the first woofer cone mode.

The only driver's I've seen that can go high enough to mate properly (pistonic) with a tweeter in the 2.5 kHz region would be the W18 Excel woofer, who's resonance is at 5 kHz and still requires a high order network. But one might be able to use an asymmetric combination of 3rd order on the woofer, and 1st on the tweeter, using the tweeter flare to fill in for the limitations in dispersion of the woofer. But that would stress the tweeter a fair amount- it wouldn't be my choice over a conventional high order network. To work well, the series crossover would require a woofer free of cone modes up to 6 kHz, without any peaking- that's a pretty rare bird, in my example. The Scanspeak revelators do a good job of controlling cone modes that high, but they aren't pistonic.

But you know, the great thing about DIY speakers is that everyone gets to share their ideas, and in the end, chose what they want to build and what they want to listen to.

Best regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Patrick Sun** · 31 March 2002, 00:08 Sunday

Jon,

Don't get me wrong, I'd love to see how you fare, and what it took to get there. I'd be tempted to build a 2-way with the XT25 (perhaps experiment with a few of the proven XO's and decide which one *I* preferred). I confess, I'm drawn to the XT25, it must be the pertuberance of its form...

Do you have in mind what midwoofer you'll be mating with the XT25?

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**ThomasW** · 31 March 2002, 00:39 Sunday

Pat

Jon's initial plan is to use these midwoofers
Hi-Vi M8n

Most of the M.A.D. top guns and others are intimmidated by the cone breakup peak (see .pdf file). And that would be an issue if it weren't for Jon's newest XO design.

I've got a pair of the M8a-MKIII 's in the HT and the integration of the 2-way is seamless. This is without doubt the most transparent 'box' speaker I've heard. Had a friend over to audition them with my Sony SCD-777ES >Marchand Passive pre-amp>Ayre V-5. His jaw hit the floor when I powered them up. When I told him the price to build them, he mistakenly thought the per pair price was the price for an individual speaker.

FIY, the M8a-MKIII uses the Focal TC-120 tweeter not the XT25. But it has the same XO topology that Jon is describing in this thread. The details of the design are going to be available in a pending Audio-X-Press article

theAudioWorx
Klone-Audio

**Patrick Sun** · 31 March 2002, 12:10 Sunday

Ah....now I understand the reason for crossing the XT25 low given that muy colorful 8" woofer.

If anyone cares, I read that the Rev. (WayneJ) is about to also embark on messing around with the XT25 in a MTM (he had a speakerbuilding sabbatical brought on the birth of his 4th or 5th child recently - the man is prodigious!).

Waybe's more into using less expensive drivers and also squeezing the more performance out of them as well. But his latest project were pretty pricey, but got good reviews by those who tried his designed.

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**ThomasW** · 31 March 2002, 13:07 Sunday

Pat

There are currently 2-TM versions of the M8a-MKIII. The larger ported design using the Woodstyle 123Rev cabinet. And a smaller sealed version using the Woodstyle 803Rev (these are my bedroom speakers).

I'm pondering a sealed MTM version in the 123Rev cab. These should be an outstanding design that would be suited to using a sub for max benefit. The current TM version in the 123Rev cab has enough bass in the 30 Hx region to shake the floor in my HT

The MTM version would kick the price up to about $800/pr if the Woodstyle cab is used. If a raw MDF box were built the price to build would drop $160.

The performance target for the M8a-MKIII design was to be competitive with the $5000 Avalon Acoustics Ascent. Jon's design doesn't quite lay waste to the Ascent but it certainly gives one pause to ponder given the $600 price tag

theAudioWorx
Klone-Audio

**JonMarsh** · 31 March 2002, 16:55 Sunday

Well, actually it was the Eclipse - but I get their names mixed up frequently, too.

It looks like there will be two MTM's that we're going to work on using the Xt25 and M8a's. The sealed box MTM for ThomasW was attractive to me also as Mini Maxx sort of speaker; imagine that as the top module, and a "bass bin" with either two Shiva's, or Shiva and a Tempest (a la my X1 Klones).

We have another project though, somewhat hush-hush at this point

which we'll probably have an announcement about later today or tomorrow- a dipole speaker with the M8a's, the XT25, and a pair of Shiva's. Code name is "Arvo Parrot". This will have to be covered in a separate thread, though.

One last comment; there's a lot of discussion about the merits and benefits of transient perfect crossovers. It's an interesting challenge, one few commercial companies have met successfully. Dunlavy comes to mind as someone who has focussed quite successfully on optimizing wideband performance with well built WMTMW types of speakers- remarkable peformance and prices.

Dunlavy is one of the few who actually uses square pulse waves to document and quality control the performance of their systems.

In the seventies I built with a friend several transient perfect designs which would reproduce pulse waveforms on axis (the only place it will happen for any speaker) with high fidelity. What we did find was that there was a fairly high price to pay in terms of uniformity of power response into the room (i.e., crossover lobing off axis), and power handling and distortion, resulting from the use of 1st order crossover networks.

When setup in an LEDE room, they would produce some very "live" and detailed sound, with excellent imaging. But we found out that when we compared these speakers with some "non transient perfect" speakers that were otherwise well engineered and designed (some early Snell Acoustics models, for example) the Snells, in our listening room, also imaged and played very well, with fewer drivers and a somewhat simpler cabinet. And when placed in a more conventional reverberent room, I generally preferred the Snell's; in this case, the off axis lobing of the first order crossovers in a reverberent room increased the overall coloration and room sound. This is part of why Dunlavy so religiously uses MTM's; as this at least results in vertically symmetric crossover lobes. The inescapable conclusion I drew at that time was that a lot more listening and evaluation was necessary to see if transient perfect really made a difference in imaging and other performance characteristics that was superior to a comparably engineered system wtih high order crossovers.

BTW, Joeseph Audio, who is practically the poster child for high order crossovers, has won "best sound at show" several times in the last few years. Maybe the way human hearing works, transient perfect isn't quite the be all and end all some would think?

Just some food for thought....

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Steve Goff** · 31 March 2002, 17:52 Sunday

Jon,

I'm following this thread with great interest. The MTM design will likely be something I'd like to try to build, if you'll allow. Your comments about first order crossovers echo those of Kevin Voeckes (or is it the other way around?), who started out using first order networks, but decided that transient pefection was not as important as other considerations.

Steve Goff

**Patrick Sun** · 31 March 2002, 22:17 Sunday

Well, to me, the transient perfect designs are pretty narrow in driver selection, while the normal higher-ordered XO networks allow designers more lattitude in shaping the driver response to their design goals while being able to pick from a larger selection of drivers (especially helpful when costs is a consideration).

One more thing: I sure do wish I had y'all's disposable income!

PatCave; HT Pix;Gear;DIY Projects;DVDs; LDs

**JonMarsh** · 01 April 2002, 02:55 Monday

Hi Steve,

I think Kevin has done a lot of interesting work over the years; you can certainly see an evolution of philosophy, which is based for the most part on a pretty rigorous combination of measurement technology, manufacturing precision, and extensive listening. There are some folks who criticize his current designs, but I think part of the problem is the wrong associated equipment; the same thing can happen with Avalon's and some other "revealing" speakers.

If you're interested in the MTM design that we finalize, that's cool. This is not a "for profit" endeavor that ThomasW and I do; Thomas was a fine arts major, and he understand the creative side of audio is a lot like painting; you don't stop painting just because one wall is full!

Hi Pat,

You've got my point exactly; there's few drivers that will really work well or properly in a transient perfect design (especially a two way!). In the seventies, the minimum way to do it was a three way with a dome midrange. ALL cone midrange drivers available at that time had too much cone breakup and resonance. When you're looking at a scope and trying to recombine a 2 msec pluse wave (which requires bandwidth from 1/10th to 10X the nominal frequency), you really get acquainted with the faults of a driver in a hurry!

I know there are a lot of guys who spend a lot of time trying to wring the best performance out of a $30 6-1/2" driver. If you're trying to make a living by building a speaker that performs well for a minimum amount of money, that's the name of the game.

In DIY, I've often posed the question, "How much is your time worth?" Is it really worth saving $30 on a midwoofer, or using an inexpensive ferrite bobbin inductor instead of a low DCR air core coil? Obviously, we all have to answer that question for themselves, and for the application. A well designed system using what might be considered modest drivers can be fairly musical; last night I was listening to a small set of MTM's I built in the late eighties, using dual SEAS woofers and MB Quart tweeters, and currently connected with a compart Shiva based sub in my girlfriends sewing room. That system is very pleasing musically, compared with a lot of what you hear in the local stores, but it doesn't touch the recent builds using Eton, Focal, and HiVi drivers for midrange resolution or dynamics. A lot of that stems not from basic development (this was a 4th order L-R design, fairly sophisticated for it's time), but from limitations of the plastic cone SEAS woofers, which though they decouple relatively gracefully with rising frequency, also have issues with various cone modes smearing the response, and creating a certain "eehhhh" coloration that is characteristic of these drivers.

One of the set's of Vifa drivers is going into a small system which will be used for computer playback, with a carbon fiber 6-1/2". I'm probably going to experiment with the AR crossover for it. It's actually pretty similar to some stuff Sonus Faber has done for years. Over kill for the computer? Not once you look at the cost of all the other parts - enclosure, etc. (plus, I just about retched the last time I heard some so called computer speakers in a store! :roll: . I've got a ten year old set of AR Powered Partner 570's; those are fairly decent computer speakers (powered).

Plus, one is much less likely to become disatisfied and want to toss the project at a later date and start over if it is built to satisfy the first time around. One of the questionable aspects of our society is how much stuff is designed and produced with planned obsolescence and a limited lifetime. It's not really very friendly to the environment, or the pocket book in the long run.

I have to chuckle hearing you "envying" my disposable income- I don't really think of myself has having very much of that, but then unlike many of my "peers" at my age, I don't drive a late model SUV or import, I don't have a ski boat, my car has been paid for for several years (and I ride a train to work); my biggest financial challenge is my daughter's upcoming college education, and figuring out how to eventually buy a house since my divorce 12 years ago. In California, that's no small challenge, but the "big challenges" make a few Vifa XT25's look like dinner out and a movie.

One thing I will agree on with many others; DIY is a very enrichening experience, but it's rarely a way to save money! (considering the tools of the trade and other "investments").

Best regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**EMT** · 01 April 2002, 06:50 Monday

Jon,

It is always interesting and educational to read your posts. Thanks for sharing your projects with us.

I see that you use CLIO as your measurement system. Just out of curiosity, what crossover design software do you use?

Ellen

**JonMarsh** · 01 April 2002, 12:57 Monday

Hi Ellen,

Thanks for the very kind words.

I have bought both SoundEasy (Bodzio software, Australia) and LspCAD; I find myself using LspCAD most of the time, particularly since the version of SoundEasy I have doesn't work with NT based OS's because of the driver for the copy protection dongle. LspCAD really has a lot of small, detail oriented features (like being able to save multiple sets of network values in the same file, being able to export nets, being able to build up snapshot composites, live updates on calculations when using arrow keys to modify component values) which make it very productive in the design process- it just suits my working style very well.

I have both CLIO for DOS and CLIO WIN; the latter is more powerful for measurements, but a bit harder to export data into other programs. I keep telling myself I'm going to try out the "JustMLS" module that comes with LspCAD, but I haven't had the time yet.

If I just getting started in speaker DIY, on a tight budget, I'd probably just get the standard version of LspCAD, and use JustMLS for measurements. That's a combination that would be hard to beat for about $125.

Best regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 03 April 2002, 11:07 Wednesday

hi jon,

not sure you got my email, so i'll post the question here. what's your definition of a driver's piston operating range-specifically, what defines the upper limit?

thanks.

ps did you send the drivers to SL yet?

**JonMarsh** · 03 April 2002, 17:31 Wednesday

Hi Mark,

Man, am I glad this morning is done. I always feel that way about big deadlines.

Anyway, what do I mean by pistonic operation for a loudspeaker? The pistonic operating range of a driver is that range for which the whole cone moves in synchronization with the voice coil. Now, you'd think at first blush that should always be the case, but for larger cones, or higher frequencies, or a combination of the two, some rather less than ideal behavior can occur. Factors affecting this are the stiffness of the cone material, which affects wave propagation speed, the mass (it's somewhat intuitively obvious that there's some potential for the cone edge to get out of sync with the center due to lag time), and whether or not the material is to some degree self damping, and if the designer has sought to mechanically decouple parts of the driver from the rest mechanically, due to operating frequency.

Very stiff drivers (such as metal cone drivers from SEAS, HiVi, ceramic cone drivers from Accuton, metal dome tweeters from Focal, MB, etc) tend to maintain pistonic behavior up to a fairly high frequency, where a primary cone breakup mode hits. This often results in a large peak in the response, sometimes with a small dip below the peak. However, you can also have radial bell modes which don't show up as an obivous farfiedl amplitude swing, but which still result in glitches in the impedance curve, and irregularities in the ETC behavior. Some composite cone drivers, like certain Eton models, will have a first mode without a big amplitude glitch, but it's visible in the impedance curve, and still produces an audible smearing. (cone cry). In a lot of cases, though, there is a clear correlation between amplitude behavior and impedance issues. Some speaker reviewers, such as John Atkinson of Stereophile magazine, seem to have assumed that irregularities they see in impedance curves in many speakers are due to resonances in the enclosure walls; my measurements and experience shows it's generally the driver, because they exhibit these same problems free air.

Some midwoofers and mid drivers are designed and crafted to have as smooth a decoupling as possible between the cone center and outside, using softer plastics with high self damping properties. This succeeds (in my opinion) in "spreading the problem around", and softening the sound, but doesn't result in very good dynamics or low coloration. On of the best examples of that is the SEAS 6-1/2 PP drivers, like the P17REX and RCY.

Some manufacturers have caught on to this phenomena fairly well; Avalon Acoustics, founded by a friend of mine, was one of the earliest. They figured after they introduced their first models, to considerable critical acclaim, that within a year or two everyone else would figure it out, and their advantage would diminish. That was in the early nineties....

In the 80's and 90's, some researchers in larger companies made a fair amount of ballyhoo about investigating the non-pistonic behavior of less than ideal drivers using Laser Inferometry. Without a doubt, this technique will give very specific information about the physical behavior during breakup. However, it is much simpler than that to measure a driver to determine it's pistonic range. If you have two mics, and are in a "hurry" for results, feed the outputs into a scope setup for X-Y measurements. Put one mic nearfield at the cone center, one at a cone edge. Check several points around the circumference. As you do frequency sweeps, if the X-Y plot remains a straight line, the operation is pistonic. As soon as you see any curving over or fluttering as you raise the operating frequency, you're getting into a problem area.

Another way to check this which takes a little more time, but just one mic, is to make successive MLS measurements or sine sweeps very nearfield, preferably with a 1/4" mic, though for woofers and mid woofers a 1/2" mic will usually do. Make successive sweeps, starting from the center to the outside, storing and overlaying them. Where the sweeps deviate, the diaphram is no longer in pistonic operation.

You won't like what a lot of drivers do, tested this way.

A unique thing several people have commented on, building designs which use drivers in the pistonic region only, is that even listening to just the woofer or midwoofer through it's crossover, the sound is cleaner and more distinct.

Food for thought?

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 03 April 2002, 18:40 Wednesday

very interesting jon. i'll have to try your technique someday when cliowin arrives (the folks at audioexpress say it's held up in customs :roll

anyway, if i can pick your brain. it's my understanding that the ideal goals of infinite rigidity and infinite self dampening are somewhat contradictory-you know, the more rigid, the more likely to resonate like a bell, so to speak. do you have an opinion-while metal cones might be on the rigid end of the scale, won't they also ring?

second, in general, with the drivers you've tested, can you make any broad characterizations of breakup frequency and it's relation to driver diameter, power response or some other variable.

say, if i pick an upper frequency limit well before power response drops off, say with a wave number*radius = 1 , will i be safe in avoiding a region with significant breakup, or is it basically driver dependent and can significant breakup occur well into the flat power response area of the driver?

p.s. i'm sure you know what wave number is, but i didn't until i read john eargle's loudspeaker handbook and i don't see that many references to it, so, for anyone who is wondering, wave number = [(2*pi)/wavelength] i'm not sure how practically important this is, but it certainly helped me understand driver upper frequency range power and off axis response better. (since i can't do anything but read anyway

)

**JonMarsh** · 03 April 2002, 19:30 Wednesday

Hi Mark,

My experiences evaluating drivers indicates there's not a set way to predict behavior based simple driver parameters like the relationship between wavenumber and operating frequency. The specific physical paramters of different cone materials enters into this, as well as the geometry. By geometry, I mean the differences in behavior between a straight sided conical diaphragm, a curvilinear diaphragm, and a concave diaphragm.

It may seem in principle that having as high as possible a first breakup mode would be best (SEAS Excel drivers, Monitor Audio, for example), but then you see certain behavior's that I'm not sure I like- due to the "Q" of the resonance and the severity of the mode. One clue I look for is the behavior below the first resonance- is their a pronounced dip, for example? This occurs in many of the larger Accuton ceramic drivers, as well as some of the Seas Excel series. This is analagous to the behavior I've seen in some electrical circuits, where there's a peaking after the start of a roll-off- part of the cone is already operating out of phase, or the cone is so large that either the wavenumber is an issue (but even with an ideal massless perfectly rigid diaphram, you have the dispersion/power radiation issue), or it's a time delay between different parts of the diaphragm which is causing the roll off. I haven't studied that in detail yet; I just make a point to stay away from it!

A couple of reasons I like the M8a driver from HiVi is that the Q of the cone mode is relatively low, and it doesn't manifest a bunch of weirdness before that mode. On the other hand, the actual frequency isn't all that high, typically 2.8 kHz, but that is higher than most 8's- an Eton 8-800 has it's first mode at about 1.4 kHz.

Looking at 7" drivers, you can find parts with their first mode in the 5 kHz region, but the real usable range is about 2-2.3 kHz. The modes should be at least 20 dB down (if they're minor modes), and more if they're high Q. At least, that's my opinion, a guideline I try to use these days.

If we couple that with outstandingly stiff and dead enclosure design, then you really can work some magic with dynamic drivers, even compared with electrostatics, which have their own diaphragm problems.

To a degree, everything is a compromise; the difficult task is finding which set of compromises really do the least damage to the music and the listening experience which you desire. And is at least somewhat affordable! :LOL:

Best regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Mark K** · 04 April 2002, 01:06 Thursday

thanks for your well thought out replies.

edited 4/4/02

thanks for sending out the vifa's to sl. i apologize again for nagging you so much about that.

**Mark K** · 09 April 2002, 23:12 Tuesday

p.s. i am looking forward to your impressions of the ow1, having a rough idea of your tastes. it seems to be the current favorite of many out there. i almost bought them instead of the vifas, and may yet down the road.

p.p.s. still no cliowin. it's enough to make me want to cancel and get praxis. oh well.

**JonMarsh** · 20 April 2002, 14:37 Saturday

2 weeks travel is over, now I need two weeks off!

I'm going to setup a couple of adapters in the next week, so I can try out both the OW1's and the XT's in my test box, if all goes well. Unfortunately, there is a lot to catch up on at the office, so I may not have the spare time I'd like in the evenings!

I've also taken some look at the published data for the high end ScanSpeak version of the XT-25, and the odd thing is, I don't think it will work as well for me as the "standard" one; it's constructed quite differently in some regards, and doesn't have the relatively flat bottom end.

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**JonMarsh** · 16 June 2002, 16:00 Sunday

The Vifa XT-25 is available from Parts Express, often for as little as $54.95.

Parts Express: Speakers, Amplifiers, Audio Parts and Solutions

http://www.partsexpress.com/

Your Audio Solutions HQ Since 1986. Shop Audio Parts from Speakers and Subwoofers to Home Theater and Pro Audio. Top Gear at Low Prices and FREE Shipping

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**clm811** · 06 November 2002, 17:54 Wednesday

Jon, Thanks for the informative posting about the XT25TG30-04, which I just read today. You mentioned the need for an "LCR Zobel to tame the peak at resonance". Do you have the values for such a filter, or can you help me to determine the correct LCR Zobel? I will be using my pair of XT25's to replace the aging Scanspeak D2905-9000 tweeters in my (Lynn Olsen) Ariel/ME2's, which use a parallel second order electrical highpass with an RC Zobel for the rising impedance ABOVE crossover. Thanks in advance for your kindness in responding. -chas

Charles Moore

**JonMarsh** · 06 November 2002, 18:32 Wednesday

The LCR zobel for resonance which I use is 700UH+136UF+4 ohms.

To control inductive rise in impedance, I suggest 2uF+12 ohms. Both of these series networks are wired in parallel with the tweeter; they are connected after any LPAD network you may have in the crossover.

Have fun with your Vifa's!

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Andrew Pratt** · 06 November 2002, 20:27 Wednesday

jon is that 136 or 125 uF in the zobel? I thought the new version called for 125?

**Dennis H** · 06 November 2002, 21:05 Wednesday

Hi Jon,

Flipissimo!!!! (I always wanted to do that) :LOL:

I started to type a question about the Zobel cap and quickly thought of a dozen more questions about the Great Cap Debate.

So, first the specific question: can you use a cheap cap for the Zobel as it's parallel to the tweeter? A 136UF premium cap would cost much more than the tweeter.

The more general question: could you give us some insight into how you go about choosing caps for various crossover applications and what types you favor? Do series caps need to be better than parallel caps? How about bypassing big caps with small caps, what's up with that? Can bypassing a cheap big cap with an expensive small one give most of advantages of an expensive big one?

So many questions, so much dubious "audiophile wisdom" floating around out there. :LOL:

**Andrew Pratt** · 06 November 2002, 21:19 Wednesday

Dennis I'm about as far from a expert on this but in building my crossovers for my M8 speakers I ended up using Solen's 400 volt caps for my XO's. For the big values simply add a bunch of smaller caps together in a cluster to get the value you need. For example here's my high pass XO (before I added the L-PAd)
M8 High Pass Cross Over

**JonMarsh** · 06 November 2002, 21:57 Wednesday

Andrew,

The first values used in the M8 for the Vifa was 136, from two 68uF Solen; for Thomas, I checked 125uF, which could be assembled from a less expensive combination of GE caps. Almost no measurable difference in the impedance, and no measurable difference in the overall response, as the crossover is down over 40dB by 600 Hz. It's less than a nominal 10% difference, which is the typical tolerance for the capacitors.

Dennis,

You know, you'll get a lot of different stories about what you should use, and what is audible. Some will encourage cutting corners in zobels, particularly when the frequency range is well attenuated. Other's say you've gotta have premium caps all over the place, and would sneer at the choice of Solen's or GE polypropylene; if it isn't an Auricap or whatever, they don't want to hear about it.

After a lot of building crossovers and listening to systems, and also being fairly familiar with the quality issues with lesser caps line non-polar electrolytics, I'd have the following guidelines.

1) No electrolytic caps, anywhere, period. I'd like to think I can build a speaker that, barring abuse, will last as long as a Bryston amp, (at least!), and non-polar electrolytics just aren't in my toolbox.

2) Tweeter and midrange caps in the signal path should be a good low ESR grade polypropylene at the least. Our "budget" systems have used the GE polypropylene caps which are available inexpensively through some distributors. Solen caps are available through PE and Madisound also.

3) Zobel caps where the effective signal level at that output range (like tweeter resonance zobel) can probably be mylar or polycarbonate if you want to save some bucks. But, in most things I build I use Solen's or GE's even in these zobels.

4) Bypassing caps in a power supply, when you try to stagger and minimize total ESL makes some sense. But I think the biggest problem with caps in the series portion of crossover networks is ESR and DA (dielectric absorbtion). A small high grade cap will not have a sufficiently low ESR to make up for a big crummy one. A big crummy cap with mediocre DA will not be cleaned up by a small low DA cap in parallel. That's like thinking you can put a noisy power amp in parallel with a quiet one and reduce the system noise floor.

4) Contruction quality and grounding are very important also- as well as solder quality. Personal recommendation is single point grounds on all crossover boards, and Cardas Quad Eutectic solder. Mmmmm, mmmm good!

If you ever saw the crossover of any of the original Avlon's, which use full conjugate networks even on the woofer resonances, you'd know why (besides the woodworking) they cost as much as they do. (parts times the usual markup of course) Obviously, to some folks it may seem a little over the top to have a zobel built as well as I describe above- but it's consistent, and it avoids being penny wise and pound foolish in the overall quality of execution and performance.

The M8 is a "budget" bookshelf design by my standards, but the components, design, and execution aren't remotely "budget" by the standard of mass market speakers. But if you ever compared the crossover for a B&W speaker (much less an Infinity) with an Avalon, you might understand why. You may not agree, or you may not be able to afford, but I find that I'd rather have a really well executed natural sounding 8" two way speaker, then a mediocre three or four way with oodles boom, squawk, and tinkle.

h:

h:

WHOOAAAA, I'd better chill out on the caffiene and climb down off my soap box!

Regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Dennis H** · 06 November 2002, 23:19 Wednesday

Thanks, Jon. Stay on that soapbox. You're preaching to the choir and we're hanging on every word. :LOL:

For the acronym challenged, please explain ESR and ESL - what they mean and what kind of caps are "good" and "bad" in that regard. As well, what are the advantages and disadvantages of film vs. foil caps. I guess I'm looking for a little basic theory (that a former carpenter can understand) about this whole cap thing - what to look for and what to avoid. I understand what a cap does but I don't have a clue what makes one better than another, other than somebody claiming one kind or another sounds better - with verbiage suited to a wine tasting review, of course. :LOL:

**JonMarsh** · 07 November 2002, 00:19 Thursday

Sorry for the acronyms- it's an engineering disease, you know.

ESR - Equivalent series resistance - this is the distributed resistance in the capacitor which acts like a parsitic series resistance; this lowers the "Q" of the cap, and causes a small voltage drop. ESR should be constant with frequency, but often isn't. ESR depends mostly on the sheet resistivity of the conductors making the capacitor (foil, metalization on film, plating on mica, etc), and the connections (usually by low temp solder) to the plates from the wire.

ESL - equivalent series inductance- the parasitic inductance of a capacitor, which is due to it's wire leads and to the winding and connection method of the capacitor sheets. Short wide capacitor plates make for low series inductance. At some high frequency, the ESL and the nominal C actually resonate. Note that older electrolytic style capacitors often had enough inductance that the self resonance would be in the range of a few kilohertz to 10 kHz.

Caps made with plastic dielectrics like polypropylene, mylar (polyester), etc., are divided into two basic types- metalized film, in which a thin layer of metal (tin or aluminum) is vapor deposited on the plastic, or film and foil. Film and foil are bulkier and more expensive to make, but have the lowest impedance. But with good manufacturing techniques, metalized film caps like the GE's and Solen will have ESR well under 10 milli-ohms- i.e., under 0.01 ohms.

Dielectric absorbtion coefficient is the "memory" factor of the insulating material. Many materials will absorb a little bit of the charge when a voltage is applied, then release it later. This is though to be responsible for "smearing" of sound. To demonstrate this, charge a cap up to a specific voltage, say, 10V or 20V; then quickly discharge it, then leave it open circuit. On materials like ceramic insulators, aluminum electrolytics, and even mylar, a residual voltage will "re-appear" to various degrees. Some of the best materials for this characteristic are polystyrene and teflon. Polypropylene is one of the best moderate cost materials, hence its popularity.

Modern polypropylene caps like the GE's and Solen's have been typically developed for switch mode power supply appliations, where low ESR is quite important, along with low ESL so that the impedance remains low to several hundred kilohertz. Some metalized film caps I've measured have ESR as low as 3 milliohms.

I'm going on the road tomorrow AM, so won't be able to post till the weekend. Have a good rest of the week!

Regards,

Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Dennis H** · 07 November 2002, 12:30 Thursday

Thanks for the clear and informative response, Jon. Enjoy your trip.

**Hank** · 07 November 2002, 15:56 Thursday

Originally posted by JonMarsh

I'm going on the road tomorrow AM, so won't be able to post till the weekend. Have a good rest of the week!

Hmmm...would you be driving a white van loaded with speakers?

**ThomasW** · 07 November 2002, 17:44 Thursday

Hank

The sorry thing is that Jon's social life is so limited he probably doesn't know about 'white van' speakers

theAudioWorx
Klone-Audio

**JonMarsh** · 07 November 2002, 22:12 Thursday

My social life seems to consiste of saying "Hi" to the stewardess when I get on the plane, and saying "Bye" to the stewardess, and "Nice flare on the landing" to the pilot after I get off the plane.

:? What are white van speakers? Is that like when Thomas and I met in the Utah/Nevada border in the desert, and I (driving a white rental van) picked up my X1 cabinets which we built in Denver, to haul them back to CA? Are my X1's white van speakers?

-Jon

Earth First!
_______________________________
We'll screw up the other planets later....

**Andrew Pratt** · 07 November 2002, 23:37 Thursday

Jon I highly doubt the people selling speakers from the back of a white an could spell X1

Vifa XT25 Tweeter

Vifa XT25 Tweeter

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment