Three Way Evil Design Study

[JonMarsh]

What are the advantages of using phenolic boards (I guess same stuff as pcb boards - Paper/ epoxy resin) for building your x-over (again asuming you will hardwire the components).

Those speakers will make some noise, or does test tones don't count .

Just stronger, and with heavy crossover inductors, just a bit more piece of mind. Finished item will look nicer, but of course only I and you folks will know... pride in workmanship? It's not like phenolic PCB, though, it's more like the stuff used for the separator panels, or HV insulation in power construction.

[gbegland]

What's the planned xover point for the mid/tweeter? I'm curious as to why you did not get the spacing a little closer together...

Greg

[meb46]

Jon, you are making me "pine" for my versions... they occupy my office wrapped in bubble wrap, Cardboard and Cling-wrap... shipping out on Monday

Looking forward to getting the crossover details so I can start procurement again. At least I can build the crossovers in absence of the boxes for a few weeks

[JonP]

OK, just another teaser shot, while I get back to detail work on the second cabinet set and the internal wiring harnesses.

Took a lot of work to get here... I don't know if I would have started these if I'd fully understood what I was getting into! :roll:

Just read a review of the Wilson Alexandria XLF speaker on Stereophile, made me fell good, realizing that these are really just overgrown mini-monitors with a bass bin, and are VERY portable compared with the Wilson's 655 lb per speaker weight! 8O

Heh... there's a local high end store that relocated to OC here a little while back was having dealer event evenings. They had Magico first, ;x( then Nola... :T when I went back the second time, I was able to see the, uh, deeply impressive, marks the Q7's left in the carpet of the big listening room! 8O

Yeah, I think anything over 500lbs is just asking too much commitment...

Looking VERY good... good to see all the progress you've been able to make.
Hope I get to hear them some day...

PS, hope it was a good mail day... I have a new toy or two that are due to show up pretty soon.. yeah, I understand that anticiiiiiiiiipation.....

[dlneubec]

What's the planned xover point for the mid/tweeter? I'm curious as to why you did not get the spacing a little closer together...

Greg

I wondered the same thing. My bet is the spacing is to get the drivers acoustic centers to physically align. I'm not familiar with the Jantzen tweeter or used the Accuton mid, but I would presume as low an XO point as possible will be targeted due to the spacing and also to reduce the directivity difference between mid to tweet at the crossover point.

[5th element]

I thought the original intention was to use a tweeter from transducer labs. At the time these had larger face plates, hence the original cabinetry was done around the increased separation that this would require, then Jon went with the smaller diamond dome.

[JonMarsh]

What's the planned xover point for the mid/tweeter? I'm curious as to why you did not get the spacing a little closer together...

Greg

This is a multi-faceted system design issue, Greg, as you might imagine.

The obvious first order point would be to get the drivers as close together as possible to make the origins more nearly coincident. But there are other considerations here, including the design intent with regards to baffle width/baffle step and how that influences the dispersion of the driver (small baffles having a baffle step transition at a higher frequency, which means a higher frequency before the driver shifts from quasi 4pi to quasi 2pi radiation in the horizontal direction).

The other consideration is approximate alignment of acoustic centers and how that affects the "focus" of the system at a planned minimum listening distance and vector. For an interesting discussion of the challenges with interfacing large speakers in a typical listening room, see the Wilson Audio Alexandria XLF review.

http://www.stereophile.com/content/wilson-audio-specialties-alexandria-xlf-loudspeaker

Starting with this point, consider the image Mike posted summarizing his analysis for this type of design- I'm in substantial agreement with it.



Now, envision a line or plane formed by the acoustic centers of the tweeter (very near the front panel), and the midrange (about 2-1/2" behind the front panel). With a typical all pass crossover, you'd want those planes to be somewhat aligned to a plane 90 degrees perpendicular to the vector to the listener. Because this is a tall speaker, that factor is not going to be parallel to the floor, but pointed down somewhat. If you move the tweeter too close to the midrange, the angle skews quite a bit and you wind up with a perpendicular to that plane pointed downwards too much. That could be compensated by changing the front baffle angle, but then we have other fish to fry, too.

The front panel baffle tapers to a very small area on top for a reason- if you look at all the drivers, they're mounted on baffles that are not much wider than the drivers themselves. Hmmm, something about that sounds familiar, with some of the newer generations of dipole systems!

This opens up the radiation pattern for a box monopole speaker, and the effect on the power response is significant. This ALSO makes life a bit more "difficult" for the midrange and tweeter, as this lowers the axial response below the baffle step frequency at the same time that it raises the power response. So, some careful choice of drivers and crossover characteristics is required, IMO.

Based on some back of the napkin calculations, (I have very special napkins around Casa de Juan, with direct electronic interface via an iOS driver to Evernote), I expect that I'll be using a 3rd order All Pass configuration, like the NatalieP's, in order to have at least 2nd order electrical attenuation in the crossover transition region. I expect the midrange to tweeter crossover to be in the range of 2.2 to 2.5 kHz, and the woofer to midrange in the 250-300 Hz area. This should keep electrical drive and acoustic behavior confined to a fairly favorable range for all drivers in the transition region as well as the nominal passband.

My past experience and that of many, many DIY builds is that the 3rd order all pass (as implemented in a quasi series network) in the Modula MT and NatalieP has very nice characteristics in the transition region and stop band, so that is my first pass choice. I haven't eliminated the possibility of a Duelund transfer function crossover, but I'm concerned that the drivers may not quite be up to it. I'll probably evaluate that as well as the all pass configuration with LSPCAD simulation before making a final decision.

Of course, I could be totally delusional and this is all self rationalizing BS, but hopefully, NOT! :W

[JonMarsh]

I thought the original intention was to use a tweeter from transducer labs. At the time these had larger face plates, hence the original cabinetry was done around the increased separation that this would require, then Jon went with the smaller diamond dome.

Mike is using the Transducer labs N26R; I'm measuring the N26C versions I have to get data for his crossovers. My plan has been Lucy in the Sky with Diamonds all along...

[JonMarsh]

Heh... there's a local high end store that relocated to OC here a little while back was having dealer event evenings. They had Magico first, ;x( then Nola... :T when I went back the second time, I was able to see the, uh, deeply impressive, marks the Q7's left in the carpet of the big listening room! 8O

Yeah, I think anything over 500lbs is just asking too much commitment...

Looking VERY good... good to see all the progress you've been able to make.
Hope I get to hear them some day...

PS, hope it was a good mail day... I have a new toy or two that are due to show up pretty soon.. yeah, I understand that anticiiiiiiiiipation.....

Yes, we had a good UPS day yesterday, nice shiny AWG 12 Solen inductors are in my clutches, er, that is, storage! yes, my pretties! You will work very well in this new project... very shiny!

GF is inquiring about me refinishing a couple of pieces of furniture again, which has led to more discussions about setting up a temporary lacquer spray booth, so we may be back on for that route. Powerpoints, Keynotes, ROI plans, and full analysis will be ready by the time she gets back from Cuba the end of the month.

Believe it or not, a friend/associate was considering Magico's in Munich, one of the two big ones- for some reason, he wasn't impressed with the demo of the Q5, but I suspect associated system gear may have been a factor. On the other hand, he's getting in some Eggleston Savoy's to listen to on Jan 17, and at 355 lb each, they're well past my cut off! I figure anything over 200 lb just really needs to be reconsidered! Or seal them really well and pump them full of Helium... :rofl:

[Hank]

PowerPoints and ROI plan - I love it!! ;b> :dothewave:
My PowerPoint dipstick says that I'm a skosh above the full line :blink:

Relief this weekend:
Annual 3M Half Marathon Sunday morning (have to be there at 5:00 am - volunteer at the 1.5 mile waterstop).
Austin Cars & Coffee at 10:00 am at The Oasis.
Austin Traditional Jazz Society annual Jam Session at Capitol City Comedy Club at 2 pm.
Best part: not ONE PowerPoint slide!! :dancenana:

[gbegland]

On the other hand, he's getting in some Eggleston Savoy's to listen to on Jan 17, and at 355 lb each, they're well past my cut off! I figure anything over 200 lb just really needs to be reconsidered! Or seal them really well and pump them full of Helium... :rofl:

Just looked at the Ivy Sigs. What is with a double isobaric? Each mid has two more behind it? Whew, and talk about CTC spacing...

[JonMarsh]

Best part: not ONE PowerPoint slide!! :dancenana:

Yup! That's the way to spend the weekend! GF understands PPT's and ROI plans, being a 28 year IBM'er. So we joke about that for any big decision to make...

Having gone through my book again on "Spraying for Dummies", I think we're going to go ahead with that plan... besides, I love the smell of lacquer in the morning! :B



Enjoy your weekend, even the early in the AM stuff! :T

[JonMarsh]

Just looked at the Ivy Sigs. What is with a double isobaric? Each mid has two more behind it? Whew, and talk about CTC spacing...

Do you notice they don't even give the weight on the Ivy Signatures? If you have to ask, you won't be able to handle it.

I sort of get the ISOBARIC on the bass benefits, but I'm really scratching my head about doing that for the midranges. Guess I'm must not aiming high enough, huh? :roll:

[5th element]

It's not just Isobaric on the mids, there are TWO drivers in isobaric configuration behind each mid, so three in some odd compound arrangement.

I can understand going iso on the mids but only if they are expected to play down low. The mmtmm arrangement makes me think that the outer pair are for BSC, so perhaps they do play down to around 100Hz. Going iso would reduce the cabinet volume necessary for the 4 mids, which appear to be morel supreme carbon/rohacel based, but from where I'm standing it would be more of a marketing gimick rather than anything truly worthwhile. After all you lose efficiency when going iso, so their spiel about creating an effortless sound is rather tenuous at best.

Still if you need to ask how heavy or how much then these are not the speakers for you!

[JonP]

Believe it or not, a friend/associate was considering Magico's in Munich, one of the two big ones- for some reason, he wasn't impressed with the demo of the Q5, but I suspect associated system gear may have been a factor. On the other hand, he's getting in some Eggleston Savoy's to listen to on Jan 17, and at 355 lb each, they're well past my cut off! I figure anything over 200 lb just really needs to be reconsidered! Or seal them really well and pump them full of Helium... :rofl:

Practically speaking, it's so difficult with that approach... putting in dual skylights over the speaker positions, the long cables to the 30' diameter Helium balloons for the weight reduction, the sealing problem where outside air leaks into your house thru the cable ports. (especially this time of year!) I won't even get into what happens when the wind comes up... :W

Here's to a productive weekend!

[Paulie]

Just stronger, and with heavy crossover inductors, just a bit more piece of mind. Finished item will look nicer, but of course only I and you folks will know... pride in workmanship? It's not like phenolic PCB, though, it's more like the stuff used for the separator panels, or HV insulation in power construction.

Jon, could you post a link for the phenolic boards you used? Thanks.

[JonMarsh]

It was purchased on Amazon, Leecraft BK-2, 11" by 15" by 1/4", and it looks like I cleaned them out, as they just have a notice to sign up for when it's available again.

http://www.amazon.com/gp/product/B0002898CQ/ref=oh_details_o00_s00_i00

[CADman_ks]

It was purchased on Amazon, Leecraft BK-2, 11" by 15" by 1/4", ...

Will you actually have to cut that "sheet" down to size?

If so, how do you cut that stuff without it chipping? I personally have never used it, but I'm familiar with phenolic, and I would think that it would tend to chip pretty bad when you're cutting it.

Not that this is the same, but when I built my Stents, I used some HIGH grade 3/16" Lexan in there that I had laying around, and it was NOT any fun to cut on the table saw. I was going slow, and it still chipped on the edge...

[JonMarsh]

Will you actually have to cut that "sheet" down to size?

If so, how do you cut that stuff without it chipping? I personally have never used it, but I'm familiar with phenolic, and I would think that it would tend to chip pretty bad when you're cutting it.

Not that this is the same, but when I built my Stents, I used some HIGH grade 3/16" Lexan in there that I had laying around, and it was NOT any fun to cut on the table saw. I was going slow, and it still chipped on the edge...

Thing is, the term "phenolic" covers a host of materials of various properties!

Many types of base material (paper, cloth) and many different formulations of binders.

I often just use one of my 80 tooth count veneer saw blades, but I also have a special blade with TCG ground teeth and a negative hook angle for non ferrous materials like plastics and aluminum, which works very well for those. That's sort of a "non-woodworking" tip!

[JonMarsh]

Work is back to

normal

with a vengeance...

Well, being employed is a good thing, but there can be too much of a good thing! :roll:

As a result I still haven't had time to get setup for the new cabinet measurements. But as frustrating as that is, I took some earlier measured data on somewhat different baffles and have taken a full pass through on a prototype crossover, which looks pretty promising on several levels. I stayed up late last night working on this... can't do woodworking or testing at those hours, but can prototype crossover on LspCAD.

• Sealed system means Fs impedance resonance is at a low enough frequency that I don't need an LF zobel to kill the interaction with the crossover- for a ported version, this will still be required, I suspect
• Driver data is fudged for tweeter- using D3004/6640 Be tweeter data for modeling the top end for now. Using older Aurasound data measured in different cabinet but with identical width.
• 3rd order three way all pass concept seems to be panning out nicely, and have deep notches with out of phase driver connection
• Using 3rd order all pass eases the design issues for the upper transition band for the midrange, due to steeper roll off, and it looks like the LCR networks for driver equalization at the upper end will be uncessary. (yes, that means lower cost! Who would believe? )
• The midrange seems to need enough attenuation due to it's high sensitivity that the L-Pad shunt resistor will swamp out impedance variations to a degree, and the LF mid zobel may not be necessary.
• Mostly ditto for the tweeter, it's looking fairly simple at this point (subject as always to future change).
• There is some voicing built into this network, but not as previously discussed. Too early to get into this, pending final measurements and evaluations with external components.
• Oh, and crossover points are 300 Hz and 2500 Hz. Just for the record...
• Once again, EVERYTHING subject to change, with notice. :B

[5th element]

Considering some of your past designs, the component count here is quite low! That must be a bit of a bonus too

[JonMarsh]

Considering some of your past designs, the component count here is quite low! That must be a bit of a bonus too

It wasn't just a bonus, it was a Design Goal!

With a low slope like 2nd order L-R or Duelund on the bandpass, you really have to manage the driver behavior at the outside limits of it's working passband. With a 3rd order All Pass, like the Modula MT and NatalieP, (which were realized with quasi-series crossovers, not practical here, and not wanting the HF signal to go through too many parts), we're attenuating the transition band faster, and I think this may be a good tradeoff as regards signal path, driver attenuation outside the pass band, AND as you note, total parts count.

Of course, the proof is in the pudding, many a slip twixt the cup and the lip, but I have a good feeling about this...

[Evil Twin]

It wasn't just a bonus, it was a Design Goal!

With a low slope like 2nd order L-R or Duelund on the bandpass, you really have to manage the driver behavior at the outside limits of it's working passband. With a 3rd order All Pass, like the Modula MT and NatalieP, (which were realized with quasi-series crossovers, not practical here, and not wanting the HF signal to go through too many parts), we're attenuating the transition band faster, and I think this may be a good tradeoff as regards signal path, driver attenuation outside the pass band, AND as you note, total parts count.

Of course, the proof is in the pudding, many a slip twixt the cup and the lip, but I have a good feeling about this...

Use your intuition... trust your feelings, Jon...

But do not become complacent with the progress you have made!

The Chancellor has allowed you unfettered freedom to pursue your own direction on the Three Way Design Study, based on your initiative early last year. However, that could change in a heartbeat should progress stall, or the ultimate results not be satisfactory...

I suggest you redouble your efforts and quit whining about a cushy job in the Terran sector, lest I decide I need to find new ways to motivate you...

[cjd]

I feel odd to be thinking about just walking away from my job, which seems to be at about the ONLY place in town I really would have much interest in working... Yes, that means I'd just.. not have a job, and not really want to work anywhere doing what I do now, unless I wandered into a completely different city.

I am also feeling some pressure to work on the Nebbiolo again! Who can get a lower parts count and still be swimmingly happy with the result??

C

[JonMarsh]

I feel odd to be thinking about just walking away from my job, which seems to be at about the ONLY place in town I really would have much interest in working... Yes, that means I'd just.. not have a job, and not really want to work anywhere doing what I do now, unless I wandered into a completely different city.

I am also feeling some pressure to work on the Nebbiolo again! Who can get a lower parts count and still be swimmingly happy with the result??

C

That sounds like a potentially uncomfortable spot to be in- something changed in your work situation, or changed in you?

As to the Nebbiolo, let's pile the pressure on! (is that internally developed or are there external inputs for finishing the construction and development?) If you can't build bamboo boxes, maybe a bamboo cone is the next best thing? :B

Either way I applaud, and hope you find good things musically as a result.

[cjd]

Either way I applaud, and hope you find good things musically as a result.

Oh, should I put the violin I really should be playing more next to the project? It's even assembled. And I walked by the place without knowing it when I was in Florence!

It's actually been a bit of a clean-up binge beyond what post-holiday mess usually requires, which will leave me with a little more project room. This weekend I think, though!

As for bamboo, no one here has wanted to sell me any yet. Haven't looked in a while.

Here's to hoping you discover you had too many parts in this tentative crossover.

[JonMarsh]

Here's to hoping you discover you had too many parts in this tentative crossover.

Well, never say never, but this IS a minimalist crossover for me! :B I'm not one of those guys that goes for 4 octaves of slow overlap within the 10 dB transition band for the midrange to tweeter crossover, or the 3 octaves for the woofer to midrange (you know who you are, Curt and Jim... :W )

That's why we call them "cross-over filters", not "blend-together filters". :rofl:

Many paths to the Buddha, and then we're supposed to kill him on the road if we meet him? Seems strange to me... If I was the Buddha, I'd hide. 8O

[JonMarsh]

Fiddly bits, wiring department

Some of you might be surprised at how much time it takes just to make an internal woofer cable creating a series connection for two Aurasound NS12-513-4A. Others won't be, because you've been down this road, too.

This will be discussed more in the build thread, which won't be posted until something is really built! That is a relatively jumbo Euro connector; spec'd for up to AWG8, which we're using on the input side; outputs are AWG12.

For all the flat earther's and ABX blind comp guys, just use some AWG18 wire and wire nuts from Home Depot. :roll: :W

[JonMarsh]

Ready to Measure...

Speaker: Ready to measure tomorrow.

Me: Ready to rest today!





The perspective on this shot looks a bit weird- makes the woofers and bass cabinet look smaller than reality.

I think when it's time for real pics I'll get out the medium format Pentax and a tilt/shift lens, maybe have a little fun with making the perspective really strange, and maybe more natural, too. We'll see.... that's probably a month down the road. The other cabinet is quite close behind this one, and I'm also about ready to do finishing tests on scraps soon.

[CADman_ks]

WOW! Just WOW!

Really, really nice....

[Hdale85]

Agreed

[Steve Manning]

Very nice Jon, those are looking awesome :T Love that wiring harness as well.

[sdl2112]

Jon, I think you're going to make the guys at Avalon Acoustics jealous

[Face]

Jon, I hope you plan on replacing the woofer screws...

Otherwise, simply beautiful!

[JonMarsh]

Jon, I hope you plan on replacing the woofer screws...

Otherwise, simply beautiful!

Oh yeah- have some #10 black square drive on order with Woodcraft!

[Juhazi]

Jon's " Ultimate Serial Connector" strangely makes me think of brides' and beauty pageants' secret tape trick... you could not guess

[CADman_ks]

... Love that wiring harness as well.

Me, not being an EE, I had to study that thing for a while to get it thru my head exactly how it worked. ops:

[5th element]

Those are certainly quite imposing and due to the angular look, if done in black, would no doubt suit the emperor quite well. They should be quite formidable on the sonic front too, I'd have picked a wave-guide to go with the tweeter, but you've already tried that with the c90 before anyway

[JonMarsh]

Those are certainly quite imposing and due to the angular look, if done in black, would no doubt suit the emperor quite well. They should be quite formidable on the sonic front too, I'd have picked a wave-guide to go with the tweeter, but you've already tried that with the c90 before anyway

For various reasons I decided to "stay true" to the original implementation concept/inspiration. Shortly I'll get to find out how good an idea that was... I expect the off axis sweeps today will tell me a fair amount about what I want to know, and if this choice paid off the way I expect.

The turntable is ready... (but it's too early hear in CA for making noise- I'll get my Starbucks, then take some more pictures, first!)

[JonMarsh]

I love the sound of camera shutters in the morning...

Real shutters, not the electronics sound effect from my Sony NEX!

This looks better to me, even downsampled heavily like this (the other one was too, but wasn't on a tripod, and the Zeiss lens has visible chromatic aberration when opened up- this SMC67 55 mm Pentax doesn't. But it's an old school pro lens, not a consumer lens- let's hear it for eBay!)

[cjd]

Eh - it looks like you're shooting part tungsten, part natural light, which makes white-balance a pain. That, and the second has much more depth of field (tripod enabled longer exposure I assume?) - color on the first could probably be tweaked if you were so inclined.

[Hank]

First thought: Beautimous, Jon!
Second thought: send him some square drive black screws - but I see you've got some on order.
Wish I could hear those beauties!

[ColoradoTom]

Wow, Jon those really look good! I'm imagining a pair in figured cherry in my family room!! You are planning on packing them up and bringing them out the RMAF next year.... aren't you!!

Tom

[JonMarsh]

Eh - it looks like you're shooting part tungsten, part natural light, which makes white-balance a pain. That, and the second has much more depth of field (tripod enabled longer exposure I assume?) - color on the first could probably be tweaked if you were so inclined.

Lighting was ad hoc and mixed in color temp. Photo lighting is in one of this storage units!

First was an f2.2 to try to get a decent exposure time for the NEX7 with Zeiss 24mm lens; 35mm ASP-C size sensor.

Second was f4 on an SMC 67 55-100mm zoom, at the wide end, which on medium format gives a similar field of view as 30mm on 35mm SLR. This is a new for me lens, acquired as most have been through eBay. If I only had one Pentax 67 lens it would be this one, but I also really like the 75mm f2.8 prime (late model, few made but sharp and contrasty), the 55mm prime wide angle from 86 and on(very sharp even wide open), the 200mm (post '86) prime, and the 100 mm macro. The only one I don't have and wish I did is the 400mm F4 ED. Big and heavy, but much better glass than the other long telies; another '86 design.

[cjd]

I don't even haul out the 30D or 5DII for my quick snaps. Just the little Olympus XZ-1 (f/1.8 and pretty sharp, but the sensor goes noisy pretty fast > ISO400). Not really sure why I don't, to be honest - I have a couple nice f/1.4 50mm and 80mm - I shot my sister's wedding with them so they're not slouches) and the rest are 2.8 including the 14mm. Nothing bigger than 200 though.

That aside, this thread is about the build, not the photography (yet)! And that is looking superb. I love that bamboo ply.