Three Way Evil Design Study

[dar47]

Shouldn't mention this, but I have 4 of the NS10's, too!

Don't go there Jon they will be asking for a narrow baffle.

[Hdale85]

That would be a MONSTER center channel.... 2 NS10's wow....

[ColoradoTom]

I think the only question left is if Jon is going to do a killer centre to go with all these great options. ;x(

Probably shouldn't mention this... but with my Smyth Realiser I can "clone" speakers, and for headphone listening could actually have seven of these scattered around the virtual space of my headphone listening!

If this were to actually happen, I imagine I would have limited interaction with the outside world for an extended period of time and would eventually look like Gollum! :B

Tom

[JonMarsh]

If I'm going through the exercise of building the Ardents, I'm never going to hide them behind no stinking screen. :B

No kidding! Not the Ardents or this project - Isiris.

[JonMarsh]

Wooo Hooo!

The explosion proof exhaust fan for my knockdown/fold up spray booth baffles arrived today!

[sdl2112]

Wooo Hooo!

The explosion proof exhaust fan for my knockdown/fold up spray booth baffles arrived today!

Jon...I hope it's multi-speed or your finish may never reach your cabinets :W

[JonMarsh]

Jon...I hope it's multi-speed or your finish may never reach your cabinets :W

Has variable speed control!

[sdl2112]

Has variable speed control!

I think you must be like me in that it's funny how far off from our initial intentions we can get :E I think I remember you saying you need a sprayer for maybe one time use...now you are ready for production :T

[JonMarsh]

I think you must be like me in that it's funny how far off from our initial intentions we can get :E I think I remember you saying you need a sprayer for maybe one time use...now you are ready for production :T

Yeah, Scott, it's gotten completely out of hand! My GF has always had a liking for some unfinished furniture stuff, and now there's the possibility of running amuck there... I even found some intriguing solid maple chests, plus there's a couple of her pieces we want to refinish. And don't forget, the original Ardent's, too.

It's a whole new world opening up!

[CADman_ks]

Has variable speed control!

Not to bust your chops, but that's not a fan! The fan that I used for temporary booth was a fan! It's big! REALLY BIG!!! Like 3' in diameter big!

Here's a picture of mine when I used it in my make shift temporary spray booth this last summer. That's a 10' wide door that it's sitting it!!! :E



I am jealous, though. I do wish that mine had variable speed. It could suck up small mammals if they get too close!!!

[Hdale85]

Those fans are blowing in or sucking out? Just curious as I had heard you want to keep positive pressure in the spray booth.

[CADman_ks]

Those fans are blowing in or sucking out? Just curious as I had heard you want to keep positive pressure in the spray booth.

Mine is NOT high tech by any means, as indicated by the photo!

I used it to pull air OUT. This particular "booth" does not have a top. So, all the fan is really doing is drawing air from the top, and pushing it outside. More importantly, it's keeping the overspray off of everything in my shop. That was my goal, and it worked pretty well.

My plan at some point is to build a dedicated "room", that has air filters on one end, and this fan on the other. Then all of the air coming in is filtered, and clean in the room, and the exhaust goes outside and stays off of things...

[JonMarsh]

Well, CADman, I think you've taken the Russian approach; as they say, there's a certain quality to quantity (by their military doctrine!).

Somehow, I think I'll be able to get by with my 20" diameter variable speed fan. I've got knock down booth ideas from several sources, including the "Spray Finishing Made Simple" book by Jeff Jewitt, and some good online posted concepts. This will just be a 4 piece affair (main panel with exhaust, two side panels, and a shallow roof panel to limit overspray), which in use will be placed in the garage opening (two car garage), with air intake from the doors behind.

Dang, I shouldn't have popped into Amazon to check the spelling of Jeff's last name- new Hiromi Uehara album ("Moving") coming out in the US beginning of March! Amazon has these pop up ads that are targeted at stuff you already buy, and they're too effective at getting me to spend more money!

Speaking of spending more money, I ordered a new DAC. Not going to say more until I get it, and figure out what I got myself into, but I will admit it required an international bank wire to pay...

Yes, it's intended to be the ultimate solution to the Isiris system problem. We'll see if that pans out or not. Cross fingers.


Yesterday I also ordered chasis parts for my nCore 400 build; two monoblock chassis from Hexateq.











Happy Friday everyone!

[tf1216]

Those are the coolest Ncore housings I've seen to date. Nice find Jon!

[JonMarsh]

Those are the coolest Ncore housings I've seen to date. Nice find Jon!

I got lucky on this one; they're not usually available as piece parts, but Richard Krol from Hexateq posted on the nCore thread in the DIY Audio forum a couple of days ago that they had a few they could sell, so I jumped on it. They sell an amp in this chassis based on a UCD module with a linear supply- I suspect it's not selling so well since the nCore modules came out the beginning of 2012.

[JonMarsh]

Shouldn't mention this, but I have 4 of the NS10's, too!

Now, if one were to desire a lower crossover component count crossover with a Duelund transfer function, imaging a three way scaled down from this one with two 8's or 10s (either NS10's or Revelators), a Satori MW16P





or SS 12MU/4731T for the "midrange", and take your choice of candidates for the tweeter:

• D2608-9130
• D3004/6600
• D3004/6640

Speculation is cheap, of course; time to build something is not.


In totally unrelated news, just got the next storage item for my Music DAW and next server to go with the new DAC; it's a WD Velociraptor Duo, Thunderbolt interface, which uses two enterprise grade 10K drives. I've got a couple of other Thunderbolt drives, but they're not this fast- write throughput is about 300-350 MB/second; I copied a folder with 45 GB of high res music files in about two minutes. Awesome, when you're trying to get stuff done in a hurry; makes dealing with 4-5 GB album files feel more like copying CD level files on a standard system.

[CADman_ks]

Well, CADman, I think you've taken the Russian approach; as they say, there's a certain quality to quantity (by their military doctrine!).

Somehow, I think I'll be able to get by with my 20" diameter variable speed fan. I've got knock down booth ideas from several sources, including the "Spray Finishing Made Simple" book by Jeff Jewitt, and some good online posted concepts. This will just be a 4 piece affair (main panel with exhaust, two side panels, and a shallow roof panel to limit overspray), which in use will be placed in the garage opening (two car garage), with air intake from the doors behind.
....

LOL!!! I have a motto, if some is good more is better! May not be true in this case, though, because MORE is a LOT!! :E

As for the 20" fan, you'll be fine. We had a room very similar to the size that you're talking about when we did the finishing for the trim in our house. My father in law has a fan that's probably around 20" in diameter, and it moved a TON of air, and really did a decent job of pulling the air out of there.

One thing with my set up is that when I finally get the actual booth done, and I have filters on it, I can actually use the filters as a way to slow down the air. I could even double up the filter thickness, making it harder to pull the air through, and slowing it down. My fan is also belt driven, so I could actually change the gearing to anything that I want to slow it down, if I could figure out the speed that it needs to go. I have options, just not quick ones like you do...

[JonMarsh]

LOL!!! I have a motto, if some is good more is better! May not be true in this case, though, because MORE is a LOT!! :E

As for the 20" fan, you'll be fine. We had a room very similar to the size that you're talking about when we did the finishing for the trim in our house. My father in law has a fan that's probably around 20" in diameter, and it moved a TON of air, and really did a decent job of pulling the air out of there.

One thing with my set up is that when I finally get the actual booth done, and I have filters on it, I can actually use the filters as a way to slow down the air. I could even double up the filter thickness, making it harder to pull the air through, and slowing it down. My fan is also belt driven, so I could actually change the gearing to anything that I want to slow it down, if I could figure out the speed that it needs to go. I have options, just not quick ones like you do...

Options are good- I'll be using furnace filters for trapping big stuff going out. I can see that loading up on filters would slow down the airflow, but why not just get a variable speed drive? Changing the pulley ratios is probably a better idea.

The plan in my case is something made out of foam core panel used for insulation with selective light framing, that I can put away in off site storage.

We'll see how THAT pans out- a lot of "experimental' stuff going on here, I realize...

[CADman_ks]

Options are good- I'll be using furnace filters for trapping big stuff going out. I can see that loading up on filters would slow down the airflow, but why not just get a variable speed drive? Changing the pulley ratios is probably a better idea. ...

The motor that's on there is like the kind that you would find on a belt drive table saw, whatever they are called.

I thought that it was not a good idea to use them with variable speed drives?

Changing pulleys would be simple enough, IF one could figure out which RPM's they wanted to go to without needing to try a whole bunch of different pulley combinations. That could get expensive!!!

[dlneubec]

The project is looking excellent, Jon!

BTW, I have 4 of the Aura NS10's that I want to do something with at some point. Have you ever done any modeling with them in an OB dipole?

[JonMarsh]

The project is looking excellent, Jon!

BTW, I have 4 of the Aura NS10's that I want to do something with at some point. Have you ever done any modeling with them in an OB dipole?

I actually did some construction with the NS12's for OB, dual driver configuration, but switched from those to a Ciare 15" pro driver.

The NS10's were too good a deal to pass up, but was planning to do some kind of monopole system with them- I think of them more as woofers, with an unusually linear motor (underhung, distortion declining all the way up to 1 kHz). Now, I'm still wondering about a junior version of this system, which would be easy to engineer, but just don't have the time to build it!

[JonMarsh]

Some more measurements and other work done

Stayed nose to the grind stone, got Smith and Larson WT2 running on my MBP under Vista, and got all the ZMA files done, including for the N26C-A, new ones for the 6640, and measured the N26C-A and 6640 on the Head module or a test baffle of the same configuration.

I dropped the full measurements into the concept crossover, using 30 degree of axis data, and it didn't need much tweaking to get something in the ball park; I still need to look at it some more and would like to raise the HF impedance, it's getting down to around 4 ohms. Still, it doesn't look bad at this point at all...

Why 30 degrees? There's a lot more of the effective power response at that axis ((+/- 5 degrees) than on axis...






Based on my Ardent and Modula experience, I'm planning a slight down tilt for the axial voicing, subject to change with minimal notice. Also, note that I"m playing some games with the woofer to mid overlap, which looks funny plotted here but usually works well in reality; both slopes steepen substantially an octave past the transition region, which is desirable to keep the woofer resonant peak well down in level. Also results in the woofer/tweeter overlap point being in the center of the midrange, where it belongs (by a Duelund design- which this isn't).

Plan to prototype one crossover fairly quickly and investigate and tweak. May use "junk box" parts for that!

Oh, and the Fuji Mini-Mite IV and the Blogs aircap set arrived today, too!

Update: parts ordered for a test build, a little pricier than I'd expected. Should be testing by next weekend, on first pass. Started the order at the junk box level but got a little carried away as I worked my way towards the tweeter. :rofl:

Update 2: Wrong graphic files for schematic and response. Updated Sunday.

[TacoD]

Looks good! Are you planning to listen to these speakers before finishing the cabinets? Chances are you will not take these apart anymore

[JonMarsh]

Looks good! Are you planning to listen to these speakers before finishing the cabinets? Chances are you will not take these apart anymore

That could be a risk!

The plan of record right now is to tack up and test 1 crossover, evaluate and tweak in mono, with the one test cabinet. After that looks OK, then build the two "production" crossovers. Ordered parts last night for the test crossover, including "cheap" stuff like Erse iron lamination inductors. OK for FR tests, but at this time planning to use AWG12 air core. Everything subject to evaluation, of course.

In parallel, get the knock down spray booth built (may be able to start on that next weekend, but GF's daughter is coming out on the 16th, so that weekend won't be productive.) Also need to do finishing process tests, I expect that will take much of one weekend. Not sure which one.

Then there will be the Pore-O-Pac or sanding sealer Plus Silica powder for both cabinets, then sanding, then the spray work. Probably early March by then... best guess, if I can stay this productive, and work doesn't screw me over too much (have some work stuff to do this afternoon).

My general experience is that trying to look and plan this far ahead with any confidence is more ju-ju than science.

[JonMarsh]

LF Data

This is a little more data, just the quasi nearfield woofer response measured between the 2 NS12's, about 8" out; this rolls off the 2.3 kHz peak that occurs because you're off axis of both drivers.



It just confirms the expected sealed box response, down about 4 dB at 38 Hz, jives well with the measured impedance curve and Unibox model for this volume.




A conservative overdamped alignment three plane boundary calculation for room placement at 1.3 meters and 2.1 meters from side and rear walls shows good in room response to 30 Hz; with the actual cabinet performance it will be better than this in most rooms.

More details about these design choices will be in the build thread.

[benchtester]

...I've seen some interesting posts here for how to mate up the more expensive SS tweeters mechanically to a waveguide, but not any measured results. Also, the more manageable smaller waveguides, like the old MCM and the Jantzen which I like, show effective waveguide loading only down to about 2 kHz, dropping off below that. And getting consistent performance in the 10 kHz and above region is tricky- was able to do that with the SS D2608/9130, was NOT able to do that with the Millenium Excel, which had been another candidate for the Modula Xtreme. ... :W

Here's some measurements, my D2608/9130 is mounted to the old MCM 6 1/2" waveguide with a counterbore in the faceplate for closer mounting. In the first graph, the top curve is a raw measurement in a 3/8 c.f. cabinet; and below are two simulation examples using a single cap and resistor crossover. (Graph not calibrated for spl.)


The second graph is the 2-way system (with 15W mid-woofer) measured off-axis, 0-30-45-60-90 (15 was skipped for clarity). The crossover was a 1.1 uf cap and a 6 ohm resistor. Data below 400 hz is not valid due to short gating.

[JonMarsh]

Wrong image files posted for Schematic and SPL plot Saturday night. Updated this afternoon.

[Eric S]

Jon, these look just beautiful! I've been away for a while and it's a nice welcome back to see a project like this! I thought the Avro design was large at nearly 5ft tall and 120lbs each - you'll need a crane to move these around...

[JonMarsh]

yeah, a crane, two men, and a small boy... or large boy! The weird thing is, the woofers in these weigh almost nothing compared with the Seas LROY woofers - these would be a nightmare if they had woofers with conventional motors!

[JonMarsh]

Small steps

The hand held B&K 879B LCR meter arrived yesterday (along with things unrelated to this project, for GF's sub), and crossover parts should be "hitting the dock" today and tomorrow.

I redrew this for the upcoming build thread, a bit better legibility than the normal LspCAD schematic. Definitely keeping the component count down, if this works out without serious changes. It will require a solid 4 ohm capable amp, though. Drawn in Eagle PCB, which I'm looking at because of interoperability issues with a 3rd party contractor in Europe for our company. They have a good Mac version, which is what this was done on.

[lhwidget]

Hey Jon, I've been following for a while now, and I've enjoyed every minute of this thread. And, your speakers look great!

Your note above about your B&K meter arriving today reminded me of an interesting thread over on PE's forum.

Does your new meter measure resistance using DC, or its typicial 1000 Hz AC signal for L & C measurements?

[JonMarsh]

This is a B&K 879B, it can measure at multiple frequencies, including 1 kHZ and 10 kHz.

From the specs and manual, it appears to measure with AC for all devices. One reason I liked it besides the accuracy and basic features is having a built in USB interface, which is an extra cost option if available at all on most other LCR meters in the $300 price range.

[JonMarsh]

Working test build schematic broken out by components

This is my working test build schematic broken down by individual components, which should make the sizing/specs a little clearer. Believe it or not, this reflects a genuine effort on my part to control costs, and I'm evaluating both Erse iron laminate and Solen AWG 12 inductors on the low end.



Moving to a 3rd order all pass (like the Natalie P and Modula MT used) really allows enough extra roll off that we don't have to play games with LCR shunt equalizers, and through judicious overall system and crossover design, I've also (I think) been able to eliminate all of the Fs conjugate networks. we'll see if it really works out that slick in practice or not.

By the way, SA stands for Clarity SA caps, MR stands for Clarity MR caps; otherwise caps are Audyn PP film, 400V. (made in Germany, relatively inexpensive and compact- find for LF crossover or filling out the larger midrange values, I think.

At this point, resistors are all 20W Mundorf, as used in the Ardent and Modula Xtreme builds.

[sdl2112]

Curious...does L1 really have a DCR of 0.8 ohms? Also what are the internal volumes of the bass and mid cabinets. Thanks for sharing the details of your design/components...looking good :T

[JonMarsh]

Curious...does L1 really have a DCR of 0.8 ohms? Also what are the internal volumes of the bass and mid cabinets. Thanks for sharing the details of your design/components...looking good :T

That's a worst case value for possible air core parts I would consider using, and was used in the LspCAD simulation. Actual Air core DCR for the Hepa Litz Solen is about 0.5 ohms. Laminate core part AWG16, 500W from Erse is about 0.23 ohms. So, insertion loss will be less than modeled in all cases.

I.E., we've got nowhere to go but up, barring a really silly component choice!

With a 0.8 ohm DCR, we'd be looking at about 10% insertion loss, which is about 1 dB.

[Face]

27uf MRs are biggins.

[JonMarsh]

27uf MRs are biggins.

Monsters, in fact. It's weird, the 27uF MR's and the 50 uF SA's are WAY bigger than the 100uF Audyn's. The midrange crossover will be pretty big because of the caps. Someone can always chose to change that of course, but I'd not personally recommend that; my experience is that as you move up in the frequency range, capacitance physical resonance and transparency becomes more and more important. No Duelunds here, can't afford that; heck, if I was really flush, I'd be using 27uF MR's in the midrange crossover! Might still retrofit that at some point, at least a partial. But we're talking some serious Pound Sterling here. Just adding one 27uF MR as part of each midrange cap would add $400 to the overall crossover cost. With some other irons I have in the fire, I don't feel I can do that right now, but as the crossovers are external (at least for now), I could do this later on this year.

I just paid for a pricey DAC to be delivered in April (actually, picked up in Europe), so I'm feeling decidedly poor right now! :roll:

[Hdale85]

Well we have different definitions of poor I think haha

[JonMarsh]

Well we have different definitions of poor I think haha

Been there, done that... it's all relative.

In some ways, I don't think I've changed much... my basic lifestyle and where I live is no different from 20 years ago, literally, but I can buy more power tools and parts now! A lot of saving over the years has definitely paid off.

Then there's the difference between what the bank account looked like before Christmas and after Christmas! This is generally a "poor" feeling time of year for everyone, regardless of circumstances. Still, I have a decent job and a lot to be thankful for.

[Hdale85]

Yeah after Christmas always sucks.

[cjd]

I come here to put what I otherwise consider to be a very healthy paycheck into perspective.

[dar47]

Considering this is in the $80,000 commercial range, You go Jon :T
It is much more relative then many other thing guys go nuts on.

I was talking to some guys at work and they were telling me how they are spending $5000. on upgrades and mods to their snow machines that cost $13,000. and my nasty little Engineering brain could only think your putting all that money into a machine that can not possible carry a 250+ lbs of man over rough snow covered terrain at 100 miles per hour and last. These machines should have $20,000+ in real parts. When I told them I will be at $3000 after mods on my newly purchased TT, they thought I was nuts. :P

[5th element]

Am I the only one that finds it somewhat ironic that the loudspeaker where Jon tries to be economic in the crossover, is probably the most expensive, by a mile, in all other areas. :B

What DAC did you go for though Jon? I'm sure you've got more than just a few of us curious, especially as you appeared to have only just bought the discrete R2R DAC, or was that just window shopping?

[JonMarsh]

Considering this is in the $80,000 commercial range, You go Jon :T
It is much more relative then many other thing guys go nuts on.

I was talking to some guys at work and they were telling me how they are spending $5000. on upgrades and mods to their snow machines that cost $13,000. and my nasty little Engineering brain could only think your putting all that money into a machine that can not possible carry a 250+ lbs of man over rough snow covered terrain at 100 miles per hour and last. These machines should have $20,000+ in real parts. When I told them I will be at $3000 after mods on my newly purchased TT, they thought I was nuts. :P

That is so typical, but you know, my thought is that it comes down to emotional resonance. They relate to their own desires and upgrades emotionally, and so their decision isn't made on an engineering cost/effectiveness level, but on the "Will it get me closer to my emotional nirvana for this topic/project?" Of course, they relate emotionally to what THEY'RE into, but because they don't have a "connection" to your TT and why you are willing to invest in it, for them it seems a non-starter. (purely on an intellectual level, I'm with you though... OTOH snow machine guys are more like motorcyclists; we can be a crazy bunch, and cost effectiveness has little to do with decisions of that type in many cases (in my own defense, I ended up buying a Honda ST1300 sport tourer in late 2010 because I found a dealer in Arkansas with them in stock AND with a $4K honda rebate program, so the bike was half of the price of the other top candidate, a base model BMW K1600).

Upgrading stuff is a slippery slope- I've got 95% of the pieces to do a 2002 Type R conversion on my '96 NSX, but I bought 75% of the stuff from a guy up in Portland Oregon who was going to do just a 2002 conversion, until his wife read him the riot act and told him to sell the parts and buy her a new car. Those parts were up on NSX Prime for months, while I watch the price he was offering them at go slowly down; at 60% of what he'd paid, I bit.

Now, to my older middle aged friends (almost all of whom are younger than I am), that sort of conversion project seems extravagant and is something that wouldn't pass muster with their significant other. That's understandable to me...

FORTUNATELY, my GF/significant other has what is her own project car, owned for many years, a 356 Porsche that she's restored and repaired from few expensive accidents, so she understands or agrees on a gut level about going certain places for an interesting car. Yeah, to me that 356 is a VW Karmen Ghia on steroids (I owned one of those in the mid 70's), but for her, it's an iconic ground breaking Porsche model. And it did come way before the Karmen Ghia.

I'm hoping later this year, after finishing up this speaker project, to get back to a little car oriented DIY- I have the local paint shop that does Acura work on speed dial (much good that does me!), and all the factory manuals for my car, including the paint and body manual.

What did you do to your TT?

[JonMarsh]

Am I the only one that finds it somewhat ironic that the loudspeaker where Jon tries to be economic in the crossover, is probably the most expensive, by a mile, in all other areas. :B

What DAC did you go for though Jon? I'm sure you've got more than just a few of us curious, especially as you appeared to have only just bought the discrete R2R DAC, or was that just window shopping?

Well, I've been window shopping the TotalDAC D1, and I just finally ordered the configuration I want, a full D1 dual, which is a full balanced DAC. I won't be taking delivery until I'm in Europe next in early April; this will allow another upgrade to the design in the channel to be implemented in mine- I could have had it by direct shipment in about 3 weeks. But there's a reason; we'll be doing some shoot outs and also checking out a TotalDAC Digital Pre in series with it and with the customized TotalDAC's my colleague in Munich has.

Vincent is offering a new product which is basically a digital preamp, all the front end options including his customized XMOS USB implementation, the 10 msec FIFO buffer for reclocking the incoming digital signal (no sample rate conversion, just a high isolation FIFO, with the full DSP processor with 69 bit volume control, but instead of the foil resistor R2/R ladder DAC, it's just a digital output, AES/EBU and S/PDIF. One might ask why, and I don't know the story behind why he even tried this, but his finding and that off those who have bought them is that having the double isolation and double FIFO in the system really works some magic. Magic. I hate that word.

Anyway, he can't keep up with demand for them right now (even his own personal unit is out on customer evaluation), but we'll have one for evaluation in Munich when I'm there, to try with my D1-Dual and with my colleague's setup.

Here's a forum link of one owner's description- key part is the fourth post in the thread by cstien.

http://www.homecinema-fr.com/forum/source-dematerialisee-haute-fidelite-et-dac/dac-totaldac-t29963819-1755.html

Google translate is your friend...


One other thing expensive about this current speaker project, is how much I've spent on tools- the main necessary items being the DWS780 compound dual bevel miter saw (critical), a limited edition Woodpeck bevel gauge, the DeWalt D27905 dust extractor, a Fuji Miti-Mite IV HVLP sprayer, another DeWalt router, and don't forget the explosion proof variable speed exhaust fan for the knock down spray booth yet to be built!

It's a good thing I didn't try doing an ROI prior to starting the project, and adding in tools and other expenses... :roll:

But one cost saving measure has worked perfectly so far, and is inherent to my version of this design- no money spent on veneer or veneer glue or veneer tools! :B


Last, my current thinking on the midrange crossover, based on listening tests with the Modula Xtremes, is to convert 25 uF of each of the midrange capacitors to Jantzen Z Superior- PE's got some of those still, though not that many values- otherwise I'll have to go to Jantzen or one of the German distributors. They have a very anti-microphonic structure and low ESR, but aren't nearly as expensive as the MR caps. If I had cubic money, I'd do the whole midrange crossover with them, but that's a bit out of my league.

There's also a slight cavity resonance in the baffle mount behind the midrange driver I can measure with nearfield measurements compared with a thinner baffle IB measurement; I'm debating whether to address that with an LCR notch or not- that will probably come in the final tweak stages.

We'll see how this works out.