Low Distortion Drivers

[goskers]

I am starting a new project and figured I would come here for some wisdom and direction. This may be a little more unique than what is generally discussed here because I am using a DEQX PDC2.6 which will allow me to use steep (up to 300db) linear xo's to really control a drivers passband.

I have a quad set of TC2+ 12's that I will be using in either a u-frame or h-frame arrangement. The mids, tweeter and low pass target are where most of my questions arise. I would like to do something similar to the arvo part but in a fashion that focuses on the lowest distortion drivers made or are in the pipe to date.

So, I want to make a full-range type dipole that has some of the better if not best drivers, for distortion numbers, in the world. The new dayton reference tweeter looks very interesting for me. I would like to stay away from the w18 if possible. Nothing against SEAS but I would like to do something a little bit different. I have a pair of AT C-quence 15H's that I am very interested in using. If that would happen then I would have to get another pair as one 7" just won't cut it in a dipole app. Since the SS8543 did very well in midrange distortion numbers it makes me think that maybe a pp coned mid/woof should not be forgotten.

What do you guys think?

[JonMarsh]

If you already have the C-Quenze 15 H 52 06 13 SD drivers, then my question for you is, have you measured them for midband nonlinear distortion and linear distortion, and what do you think? They're pretty pricey, so the performance would be very good, I'd hope- OTOH, trying to figure the Xmax from the underhung design is tricky- I might guess only 3 mm, given the coil and gap length.

How does it measure and sound in your experience?



I'm very interested myself in the new 7" and 8" Peerless Exclusive Nomex series, but they're not available yet in the US- they appear to be using a similar Faraday shield and phase plug construction as the 850467, which measures quite well for midband distortion, and has good linear distortion performance (ETC plots), too.





















I'm not a big fan of polypropylene cones, which is what would put me off on the C-Quenze drivers and the SS8543 (aside from cost- not that I'm afraid of spending money, but I expect commisurate improvements in performance).

In particular, I've been "burned" in the past on expensive SS drivers which have issues with linear distortion due to cone breakup modes. Examining the FR and impedance curves of the SS8543, I'd have similar concerns using it above 600~800 Hz. But that's just speculation on my part, as I've never measured one. The 8542 looks better behaved in that regard, with the same SD motor. Note that even the 8531G has the characteristic impedance bump at 800Hz, typical of the first cone mode in SS 7" drivers. Still, it's FR is better behaved than most of their models, and it does have the SD1 motor.

Personally, from DST, I don't see the value proposition in most ScanSpeak drivers as being attractive. BTW, I've bought the Kevlar and Carbon fiber versions in the past for systems, in 7" and 8" versions. Just threw out some of the 8" Kevlar.


~Jon

[goskers]

Thanks for the reply Jon,

After some more searching today I actually feel that I have decided on the 830332 with the RS28A-4 as the tweet to try. The new exclusive Peerless offerings are very intriguing to me. I don't know that much about the nomex cone but I am hoping that it is very stiff.

I don't have any way to measure distortion with the AT drivers. I would hate to think that they are inferior either way but they are very impressive to look at. :E

How does the hds exclusive motor design compare to the best offerings of SS and Seas?

What happended to the Saint Saens project? I can't find any info on htguide about it now.

Thanks

[JonMarsh]

The Saint-Saens "dropped off the radar" because I haven't had time to work on it, and because no one is posting to the thread. I'm trying to get the smaller projects wrapped up, which in many cases had large "audiences" of interested builders, before getting into the thick of what is certain to be an intensive and time consuming project.

There's more than one HDS motor, which is why I'm careful to talk about the phase plug version, with Faraday rings, which Eric got me more interested in by forwarding some interesting measurements on mid band distortion made in a German magazine. The phase plug 850467 is quite a bit better in midrange harmonic distortion than the 850439, which is no slouch compared with a lot of drivers. The tip off is the low VC inductive rise.

The SD1 motors are quite good in this regards, too, as our the Excel. It's probably fair to call all of these drivers as being somewhat "benchmark" for fairly widely available parts.

The SS "motors" are quite good, for similar reasons, this shows up as low nonlinear LF distortion and low midrange harmonic distortion. Where I see issues are in the surround and midrange cone modes, which contribute to linear distortion (energy storage); the bit that SS does so that they get a rising resopnse above 800Hz, (to make BSC easier), which seems to be (speculation involved here) partly accomlished by decoupling parts of the cone. Just my 0.02, and remember, I'm not a driver designer, just an amateur speaker builder. But, this is the principle upon which the fundamental Avalon sound is built, and I consider it an important cornerstone in speaker design principles. Minimal energy storage in speaker cabinets, and minimal energy storage in speaker cones.

I'm a major fan and advocate of pistonic cone motion, as efforts to produce gradual decoupling with increasing frequency also seem to always involve energy storage and frequency response irregularities, too.

That's one of the big draws for the HiVI M8a as a low cost woofer- it's not one of the lowest distortion in the LF region, but the cone is damn well behaved up to moderately high frequencies for an 8" midwoofer- much more so than comparable generation products from a lot of folks.

Josh at Madisound has been vague regarding availability of the Exclusive series, but his wording in regards to inquiries in February makes me think May or June is likely.

I'm interested in the 883 and especially in the 884, the latter as a possible "upgrade" to some current projects, including Arvo- I don't intend to finalize that design until I've got some 884's to look at. Might even go into my M8ta's. Also in consideration for the Saint-Saens, for the 75 Hz to 600 Hz area. The 883's are likely for the "Cheap Trick" SLAMM Klone mod. I don't think they quite have the cone area I want for the bottom end of the Saint-Saens- would take too many of them to get down to 75 or 80 Hz. I don't want a higher crossover because I'm leaning towards "Beethoven" style woofer cabs (4 12's per cabinet) on each side, and the W cab gets squirrelly with increasing frequency, which dictates a lowish crossover frequency.

~Jon

[goskers]

Being a former orion owner, it seems as though you and SL have some similar views. I think tone burst, non-linear, linear, and inductance testing can yield some truly great info in which the start to an amazing system can be formed.

I will be decoupling the mids from the front baffle just as SL has done with his current orion mod. the Pioneer paper really shows that this can make a difference in energy transfer.

Since I am using the DEQX processor I stand in a little different group. I am looking for the pinnacle drivers whose passband yields the lowest distortion numbers available. I can easily remove any speaker nasities that would typically need to be dealt with in an analog xo. With all that said, any other drivers I should be looking at for mid and tweet duties keeping in mind that I would still like to use my 4 TC2's?

Thanks

[capslock]

TWhere I see issues are in the surround and midrange cone modes, which contribute to linear distortion (energy storage); the bit that SS does so that they get a rising resopnse above 800Hz, (to make BSC easier), which seems to be (speculation involved here) partly accomlished by decoupling parts of the cone.

I have heard this argument before but never understood it (maybe a block on may side). For BSC you want to attenuate the upper midrange. Why would this step help? It means you have to attenuate even more.


Greetings,

Eric

[denverdoc]

Goskers.

I have one of them new fangled units from the land of Oz as well. I'm in a similar situation, only planning on dipolar line array, and like you have definitely decided on decoupling the midbass drivers. As opposed to dipole woofs, went IB and would recommend same in a heartbeat as you are fighting physics too much IMO. I know you said you ruled out SEAS drivers--I assume you are referring to the mag units. One driver I am seriously considering is the SEAS aluminum driver (8") but the passband is only from 65 to 400 or so HZ in my case whereas you will likely go much, much higher. Keep me posted, as I will likely build more speakers and with switching use the DEQX to run 3 sets of speakers! So interested in simpler projects as well.
John

[JonMarsh]

The argument was that it simplified the crossover design, as a something close to a first order roll off without modification could be used up to the crossover frequency, then the rise in response of the woofer starting aroud 800 Hz (where in typical designs 10-12" wide baffle loading converts to 2pi radiation) provide zero in the pole from the first order network, resulting in "flat" response until the network corner set by the first shunt cap.

It's possible to realize this by other means, by changing the shunt impedance as a function of frequency using an oversized zobel, but the network response looks a bit nasty as regard the corner Q if you have to take this zero all the way out to 2 kHz or beyond as for a conventional crossover point. With my own crossovers at 1200 to 1400 Hz, it's not a big deal to manage it with the first series inductor and some tuning of the zobel.

~Jon

[TacoD]

The Audiotechnology poly cones have early breakups, but the sound realy nice. I have used both the C-quenze 15H, 18H and 23I. How bigger the unit, how sloppier the bass performance. The cone is simply not stiff enough for good bass performance, but if you use the 15H or 18H as mid, it sounds very nice. Good transients and a very dynamic presentation.

[david teltschik]

C-Quenze 15H

I have a DIY three way using the SS9700 tweeter, C-Quenze 15H midrange, and SS25W/8565-00 woofer. I am VERY pleased with all the drivers performance. I had a pair of speakers designed around Vifa PL drivers in the past and these new speakers I build sound infinitely better to me. I think the extra $$ was well worth it to me, but that's my opinion.

These speakers just soke up power and sound just as good at low levels or wall shaking levels.

I would recommend the -00 woofer in a sealed box personally. It sounds so much cleaner than the -01 in a vented box. I'm forever turned off on ported enclosures because they just seem to over do things for me. Besides....my mains only need to go down to around 35Hz anyway. From that point.....with the standard rolloff of a sealed allignment...there is usefull range down to 20 Hz. I am using a 15" Aurasound sub to fill in the low end as well.

Long story short.......don't be affraid to spend the money for the AT C-Quenze or SS drivers.....just get the ones that will work for you....and you'll be happy.

Just be aware.......don't skimp on the crossover! I never did understand why people would spend thousands of dollars on drivers and crossthem over with a cheap crossover. WHY!

David

[David G]

Hi David
That's a similar setup to mine, although I' went 3.5 way - Morel Supreme 110, AT 15h5206sdk, SS 18w8531 & SS 26w8861. The cquenze is excellent.
I agree - sealed box is the best solution, and a subwoofer can be added later for HT if you need more.

https://www.gattiweb.com/delta_design