AR 2ax Woofer replacement - Peerless 1727

[Pete B]

I'm new here, been doing audio and speaker work for 30 years (got an early start):

http://members.aol.com/basconsultants/spkrs.htm

Came across a pair of AR 2ax's used that were all beat up with foam rot. Figured the other drivers would be dead also, but thought I'd rescue and use them as test boxes since they were cheap. They've been sitting for many years and I recently needed another pair of speakers to upgrade the computer system and wanted to move out these speakers that were taking up space.

Turned out the mid and tweeters were all good, woofers are original Alnico with foam rot as I mentioned. Inspected and cleaned pots with cramolin, and left everything stock original dual cap, everything.

I also had a pair of Peerless 1727 10" woofers in my unused pile and decided to see if they fit, they bolt right in:

http://www.partsexpress.com/pe/pshowdetl.c...r=297-616&DID=7

I expected them to work well since mid and tweeter level are adjustable, and that the woofer response would be dominated by the crossover. Forgot that this has a high woofer to mid crossover.

Still, they sound very, very good, amazing for such an old design.

After the first impression, some colorations can be heard and the phasy tweeter to mid crossover (1st order) when moving around the room.

I tried them out in the room in an audiophile position and added a passive line level circuit for baffle loss compensation.

Running 200W/ch to handle peaks and just hit the clip lights on peaks, holding up just fine. I did not like the 1727 in other designs but they seem to be a good match here. I'd expect the Peerless 850146 CSX version to be even better.

I'm completely aware that drivers usually have to match the crossover and box but this seems to work well, probably even better than the original. I do plan to have the original 10" Alnico drivers refoamed and to drop them back in, or maybe I'll compare them and keep the best.

Anyone have T&S parameters for the original foam Alnicos? How about Fc, Qtc and passband sensitivity for the original Alnico drivers in the 2ax box?

Noticed a loud buzz coming from the back on heavy bass passages, and found that it was the input terminal/crossover board. Found the cap box has a soft window caulk around the bottom but it was not pushed in between the bottom and crossover board, pushed it in and the buzz is nearly gone.

[ninohernes]

You will like the sound of the system much better, with the original woofers. Replacing the driver with a different one kills the AR sound. The original driver was designed to be in that cabinet, the new driver was not.

[Pete B]

Well ... Joe do you have a crystal ball? You sound very confident.

Your young, loyal to AR, impressed with acoustic suspension, keep up the enthusiasm. I'll wait to do the actual comparison, my bet is on the 1727 not going to argue with you.

I probably should have offered this:

I'm an engineer, digital, analog, systems, computers, design, synthesis, simulation, emulation, chip design, board design, test, SPICE, proto development, production, quality control, value engineering, DSP, etc. I do this professionally and have done some audio work in the business world as it crossed my path over the years. 3 patents. Referrals for consulting work are always appreciated.

I've done audio and speaker design for over 30 years for the enjoyment. I'm not that old, got an early start. Did first order systems with TLs for bass in my younger days, 2 way, 3 way, 4 way by ear then by a simple measurement method that I came up with. Do higher order now, CALSOD, and various box programs, LAUD for measurements. Have done my own programs for box and line arrays which are now obsolete. Did senior project work (1980) on TL speakers where I developed a Thiel and Small type model.

Never charged for speaker design, help, and advice, starting to think I should since people tend to value information when they pay for it.

Former AES, and IEEE member for over 20 years.

[soundminded]

Looks like you hit the jackpot. How lucky or was it just intuition?

I'm always amazed at how accute some people's hearing is.

"After the first impression, some colorations can be heard and the phasy tweeter to mid crossover (1st order) when moving around the room."

Do you think a Linkwitz Riley mid/tweeter crossover would help or just a second order Butterworth filter for the tweeter? Some people here revere the original sound of AR speakers so much that anything that deviates from them is automatically suspicious. I'm not one of those and frankly, I never particularly liked the AR2 series speakers. They didn't sound all that accurate to me even though the AR2ax had the same mid/tweeter as the AR3a (I think.) Looking forward to your report comparing the Peerless and the refoamed AR woofers. BTW, as someone, I think Tom pointed out, we do not have any original AR speakers in pristine condition to know what they actually sounded like when they were new 35 years ago.

[Pete B]

>Looks like you hit the jackpot. How lucky or was it just

>intuition?

Probably a bit of both. I chose those 1727 years ago for their long Xmax of 9mm most 10" of that style are 3 to 6mm so it fits well in an acoustic suspension system. Forgot that it has a smooth response and rolloff around 2 kHz that probably matches the mid just about right.

>Do you think a Linkwitz Riley mid/tweeter crossover would help

>or just a second order Butterworth filter for the tweeter?

I'd go LR acoustic probably 2nd order electrical, but I'm not that familiar with the mid and tweeter, never saw any response charts or anything along those lines. I do know that both the mid and tweeter would not allow for improved XO frequencies, neither can go much lower without distortion or thermal problems.

>Some people here revere the original sound of AR speakers so

>much that anything that deviates from them is automatically

>suspicious. I'm not one of those and frankly, I never

>particularly liked the AR2 series speakers. They didn't sound

>all that accurate to me even though the AR2ax had the same

>mid/tweeter as the AR3a (I think.)

I never liked any of the ARs of that day very much. I believe the 3a had a better dome mid. The 2ax started out as a 2way with a fairly high crossover to the tweeter (mid in the 2ax) then they added the dome tweeter as more of a super tweeter. This is why the XO caps are so close in value of 4 and 6 uF. Most mids would have a 10 to 20 uF. The high crossover means that program material at the top end of the woofer's range will have intermodulation distortion when there is heavy bass content which modulates the voice coil inductance. Methods (shorting rings, copper caps) are used today to reduce the VC inductance and minimize this effect as is seen in the Peerless CSX 10":

http://www.partsexpress.com/pe/pshowdetl.c...r=297-636&DID=7

and even more so in the well known Scan Speak drivers.

The AR2 was in my opinion the worst and most expensive of the competing 2 ways of the day, the Dyna A-25, the EPI 100, and the large Advent. The mid and tweeter should be located closer to the woofer to reduce path length differences, the woofer to mid crossover should be lower. The tweeter suspension looks like globs of rubber cement or something - seems to have no suspension. Still, as the 3-way 2ax version it sounds better than the Advent as I remember it, which also had some midrange coloration but was even more expensive. Piano and vocals are very close with the Advents but just not quite right. I'd really like to redo the Advent XO someday because I think it can be fixed. They hesitated back then to add even one additional component in the XOs because it was so cost competitive.

>Looking forward to your

>report comparing the Peerless and the refoamed AR woofers.

>BTW, as someone, I think Tom pointed out, we do not have any

>original AR speakers in pristine condition to know what they

>actually sounded like when they were new 35 years ago.

I remember hearing, the 3ax, the B*se 901s, and a Macintosh bookshelf speaker in the high end room of a store back around 1969-70 and the Advents vs KLH 6 I think it was in the main room. The KLHs had no high end at all and I heard that some stores disconnected the tweeters to push the brand that they wanted to sell - that's how it sounded anyway. I was not impressed at all with the 3ax that day, but I never trust listening in a show room, need to set them up and hear them at home.

I'd bet that a well preserved 3ax sounds today pretty much as it did back then, don't know if the suspensions and cones stiffen up over the years, especially that strange tweeter suspension. This could be checked if we know the Fs of the drivers, linearity could be checked through distortion tests.

I see two very tempting challenges, one to modify an AR3a to get the best possible sound out of the existing drivers, and then to simply keep the box format and produce a 3way system of the same dimensions to get the absolute best sound, just to show what can be done with modern technology. Thinking LAB12, SAE1204, or NHT1259 woofer, SEAS or Vifa 5" mids and dome tweeters, LR4th order XO:

http://www.partsexpress.com/pe/pshowdetl.c...r=290-570&DID=7

I guess I rambled on here a bit.

[ninohernes]

I am not really crazy about the AR-2 series three way speakers. The AR-2 was a great two way speaker, but the 2a and 2ax three way models are not that great. If I wanted a three way with a 10 inch woofer, I would get the AR-5.

[Pete B]

whoops that should have been AR3a where I said 3ax, getting the 2s and 3s mixed up, getting tired!

[soundminded]

We've discussed this a lot on this message board over the last years or so. At the time, the early 1960s it was pointed out and agreed by many that the acoustic suspension speakers of those days had a high end rolloff. Personally I chose the KLH Model 6. If the pair you heard had no high end, it may have been damaged. The tweeters and crossovers were extremely prone to failure. Mine had to go back to the factory several times, the early versions were not field servicable. I've posted about it before but the gist of it is that the woofer was epoxied into the baffleboard so if the crossover had to be repaired, they had to remove the woofer cone to make the repair and then recone it. The grillcloth wasn't removable either. Later versions corrected this obvious design limitation. This lack of audible high end of many acoustic suspension speakers (even though they may have measured flat in an anechoic chamber) may have been one side of the east coast/west coast sound which I thought was a kind of myth until I read it in Sams Audio Engineering Handbook. The other side of course was the 2 way horn systems from JBL, Altec, and EV which had a harsh bright and uneven high end and although large and efficient didn't have quite the extended bass of the east coast acoustic suspension speakers unless they were huge. When stereophonic sound required two in a home, they tried to scale them down but they sort of died out except for their followers who are as avid about them as those of us here are about the East Coast speakers.(If you are relatively young, this must seem like ancient history to you.)

While the Peerless 831727 looks like a very good driver for the money, the 830452 and its 12 inch counterpart 830500 look like real winners to me. But even more so the Dayton Titanic MKIII series. Of course these are considerably more expensive. The trick to these acoustic suspension woofers IMO is tuning the enclosure optimally. It occurred to me about 15 years ago that you could understand the entire problem to a gnats eyelash if you viewed it as a mass-spring-dashpot in Newton's second law as applied to forced resonance. I thought I was pretty clever until I looked at Sams and saw that everybody else was way ahead of me. But it got me to realize that the role of the internal damping provided by the stuffing which controls the velocity related constant ( is critical. I think that the designers at AR, KLH, and the others determined their final design of this through trial and error.

AR tinkered with the mid and high end of their speaker endlessly from 1954 through the late 1970s while they kept the woofer/enclosure design almost unchanged. They finally gave up on the problem of integrating the dome midrange with the woofer by realizing that there was no happy meeting ground in the middle and so added the 8 inch lower midrange in the AR9. If you read Tim Holl's wonderful explanation of the AR9 design, you'll see the bass is really two AR 11s stacked which was the lineal decendent of the time of the AR1. Even so, I wasn't happy with the rest of the system balance myself and resorted to additional indirect firing tweeters and re-equalizing the system using the crossover controls and a 10 band equalizer. It took me two years of tinkering but finally, it now sounds accurate to my ears.

If there's one speaker which IMO had no high end, it's Bose 901. Towards solving this problem, I've added four tweeters per channel (one direct at -6db and 3 indirect at 0db all with first order 9 khz crossovers, biamplified them, and re-equalized the system using an external equalizer. They now sound much better (more accurate) to my ears. I have a couple of threads about them on the "other" board and if you PM me, I'll send you some pics.

[Pete B]

Hi Joe,

I agree the AR5 was much better, my friends and I said well they finally got it right, because they did seem to just tinker with the mid and high end of the AR2 over the years. However, it was so close to the AR3a seems why bother, even the enclosures were almost the same size. But I believe it was 8 ohms and gave up some bass extension and maxspl for efficiency, really seems like they just offered it for marketing reasons.

[Pete B]

Yes like I say I think the KLH 6s that I heard were tampered with or damaged as you say, didn't know that about repairs. You remind me that the AR2 was a much older design than the competition that I mentioned, and I agree it seemed like they couldn't make up their mind about what to do with the mid and tweeter.

The XLS Peerless drivers look very nice at a glance, however their T&S parameters don't seem to fit vented or sealed very well, even with the passive radiators that they suggest. Yet many are choosing them, I'd have to hear them to decide. They are true sub woofers with severe break up in the midrange where they should not be used as can be seen in the FR plots. There are a lot of nice new drivers out there.

Yes sealed systems do boil down to a simple parallel mass spring system. Fc, Qtc, and passband efficiency define the low frequency, small signal performance. They're simple second order high pass filters from the electrical analogy perspective and this is what Thiel and Small theory use to leverage their analysis. I've been working on a tutorial discussion for my web pages but no one seems interested so I've stopped work.

Interesting that you mention the stuffing, because I notice that most older acoustic suspension systems have so much stuffing that it's pushed right up against the back of the driver. Stuffing deep in the box acts to dampen standing waves or box modes, and to also take the system from isothermal to a nearly adiabatic condition causing an increase in effective volume. However, damping right up against the back of the driver adds an acoustical damping component essentially in parallel with the others in the driver. There are two damping components the mechanical resistance in the spider and outer edge as Rms or Qms, and the electrical damping component due to motor braking Res or Qes, these in combination define the total system Q or Qts in free air, or Qtc in box. It's interesting that the power radiated should provide some damping, but remember these systems are less than one percent efficient thus the power radiated can be ignored with little error being introduced into the model, isn't that ironic? However, damping right up against the back adds additional acoustical resistive loss that can be modelled as a reduction in Qms and can be significant thus lowering Qtc. It's cheating, since if the target is a lower Qtc, a system with higher efficiency results when it is obtained through a stronger motor (more magnet)or lower Qes. There's one positive aspect of using acoustical damping which is to make the system less sensitive to source impedance. If the majority of the damping comes from the motor then the system Qtc is very sensitive to source impedance or damping factor. The Peerless 1727 has a fairly strong motor so I pushed the damping material further away from the driver. Sealed systems are really pretty simple, moving mass sets Fc once a box size is chosen and provided Vas is much larger. Motor strength establishes Qtc if Qms is high and acoustical losses are low as they should be. This is what establishes the volume/bandwidth/efficiency tradeoff. Make the motor stronger and passband efficiency goes up, but Qtc goes down and the output at Fc is reduced (bandwidth loss). Make the moving mass higher to lower Fc and efficiency also goes down as a result. In quality control and value engineering we like to look at how sensitive a final performance characteristic is to a particular component in the system. A properly done acoustic suspension system is dominated by the air spring, so system performance is nearly independent of Vas as long as it is much higher than the box volume, yet many worry about Vas error in drivers that they purchase. The important value is moving mass which unfortunately is only indirectly indicated through Fs and Vas, but it can be computed. It's also unfortunate that most of the other Thiel and Small driver parameters will be off simply if there is some Vas error. It's possible to recompute them by applying a correction to Vas and then see if they're correct. Motor strength is important since it primarily determines Qtc when Qms is very high as it should be. Motor strength also effects passband efficiency which is obviously important.

I believe that there's a simple reason why home systems do not sound real, compression, and the inability to play at the same level as the original performance. Loudness compensation really is needed when playing at lower levels, and some build it into their speakers. Your additional EQ makes sense from this perspective.

I always thought that the 901s sounded better turned around. B*se seemed to ignore the time delay required to make the reflected sound from the 901 mimic hall reverberation. He was really altering the power response of the system and I'm not sure it was in a good way. Many high end systems are using rear firing tweeters in an attempt to flatten the power response so this might support your efforts.

You can send the pictures to my email, I'd like to see them.

[ninohernes]

They said that they offered it for the people that like the sound of the 10 inch woofer. It is indeed tighter sounding than the 12 inch, but it does lack extereme low bass response.

[ar_pro]

Not sure what "tighter sounding" is supposed to mean, but the 3a was superior to the 5 in every measurable way.

For a time in the '70s, the only AR speaker that was politically correct enough for a "golden ear" to not bad-mouth was the AR-5...again, the claim was that AR had gotten it wrong with the 12" woofer, and that the smaller 10" was superior. This was/is a disingenuous argument, and not supportable in any way, other than saying "it's just a matter of taste".

Check with the old-timers here who've actually *heard* a new 3a and a 5 side-by-side, and see which speaker they all chose for their own use.

[soundminded]

"Yes sealed systems do boil down to a simple parallel mass spring system."

Actually you can analyze other speaker types the same way. In fact, IMO, the Newntonian analysis shows the hopelessness of the vented box system. K in a true air suspension system is due strictly to the difference in pressure between the inside and the outside of the box and is predicted as completely linear because of the Boyle's and Charles' ideal gas laws. (P1*V1=P2*V2) On the other hand, the restoring force K in a vented system is controlled by at least the highly non linear tight suspension, especially typical of the corrugated accordian type suspensions that were practically universal in the old days and at least as important, the fact that the resistance of air pressure as a function of frequency will oscillate as a series of resonant and anti resonant nodes. This means that No matter what frequency you tune the box to, as soon as you move away from it, the internal pressure resistance and K will increase sharply to a peak and then drop down again to another low pressure node at the harmonic of the principal resonance frequency. Theil and Small may mitigate that to a degree but I really don't see that they can get around it. Too bad you didn't publish your tutorial. I'd like to see if and how they got around that. OTOH, an acoustic suspension speaker of any size should be able to be tuned to any arbitrary Fc and damping factor just by adjusting the parameters K,M, and B. In other words, at least in theory, you should be able to get as low and flat a frequency response as you want from even the smallest enclosure.

I've done a lot of thinking about spatial radiating patterns of loudspeakers and how they interact with room acoustics. My home experiments have lead me to conclude that the two way or three way direct firing loudspeaker is not a very good paradyme for home high fidelity loudspeakers. Naturally I've changed all of mine. I've written a couple of threads on my little Bose 901 project on the "other" message board here. I've owned an original pair for 35 years but this is the first time I really like them. What is your e-mail address? I'll send you some pix.

[Pete B]

Your way off here, most of my systems are vented and trust me they're not hopeless. Sorry air is not completely linear, far from it. Grossly nonlinear suspensions are a thing of the past, motors today are generally not very linear and an air spring doesn't fix that. Take a look at some Bl curves.

A vented system is fourth order, and is complex, I think most analyze it by dismissing it due to their favorite flaw real or imagined and don't really understand it at all. But AR groups tend to be brain washed into the sealed is the only way mentality so this is not the place to discuss vented.

>"Yes sealed systems do boil down to a simple parallel mass

>spring system."

>

>Actually you can analyze other speaker types the same way. In

>fact, IMO, the Newntonian analysis shows the hopelessness of

>the vented box system. K in a true air suspension system is

>due strictly to the difference in pressure between the inside

>and the outside of the box and is predicted as completely

>linear because of the Boyle's and Charles' ideal gas laws.

>(P1*V1=P2*V2) On the other hand, the restoring force K in a

>vented system is controlled by at least the highly non linear

>tight suspension, especially typical of the corrugated

>accordian type suspensions that were practically universal in

>the old days and at least as important, the fact that the

>resistance of air pressure as a function of frequency will

>oscillate as a series of resonant and anti resonant nodes.

>This means that No matter what frequency you tune the box to,

>as soon as you move away from it, the internal pressure

>resistance and K will increase sharply to a peak and then drop

>down again to another low pressure node at the harmonic of the

>principal resonance frequency. Theil and Small may mitigate

>that to a degree but I really don't see that they can get

>around it. Too bad you didn't publish your tutorial. I'd

>like to see if and how they got around that. OTOH, an

>acoustic suspension speaker of any size should be able to be

>tuned to any arbitrary Fc and damping factor just by adjusting

>the parameters K,M, and B. In other words, at least in

>theory, you should be able to get as low and flat a frequency

>response as you want from even the smallest enclosure.

>

>I've done a lot of thinking about spatial radiating patterns

>of loudspeakers and how they interact with room acoustics. My

>home experiments have lead me to conclude that the two way or

>three way direct firing loudspeaker is not a very good

>paradyme for home high fidelity loudspeakers. Naturally I've

>changed all of mine. I've written a couple of threads on my

>little Bose 901 project on the "other" message board here.

>I've owned an original pair for 35 years but this is the first

>time I really like them. What is your e-mail address? I'll

>send you some pix.

[tysontom]

>Came across a pair of AR 2ax's used that were all beat up with

>foam rot.

>

>Turned out the mid and tweeters were all good, woofers are

>original Alnico with foam rot as I mentioned. Inspected and

>cleaned pots with cramolin, and left everything stock original

>dual cap, everything.

>I also had a pair of Peerless 1727 10" woofers in my unused

>pile and decided to see if they fit, they bolt right in:

>

>I expected them to work well since mid and tweeter level are

>adjustable, and that the woofer response would be dominated by

>the crossover. Forgot that this has a high woofer to mid

>crossover.

>Still, they sound very, very good, amazing for such an old

>design.

>After the first impression, some colorations can be heard and

>the phasy tweeter to mid crossover (1st order) when moving

>around the room.

From what I understand, the Peerless 1727 free-air resonance is around 22 Hz, and the original AR 10-inch AR-2ax woofer has a resonance of 26 Hz, +/- 10% or so. Mounted, the system resonance is set at 56-58 Hz for the AR-2ax, so it would be slightly lower with the Peerless if it is flat down to resonance in the AR enclosure, an unknown quantity unless you have measured it. The dc resistance of the AR woofer is 6.5 ohms; the efficiency is 87 dB SPL (referencing .0002 microbar), about 1-2 dB less efficient than the Peerless woofer (which might indeed have *too* much magnetic strength for this enclosure). A bigger problem might therefore be damping, and whether or not the Peerless would be overdamped in the AR enclosure. Adjustment of the amount of fiberglass in the enclosure could help alleviate damping problems, but you need to measure (close-mike or 2-Pi anechoic) and plot the output of the Peerless/AR-2ax system to be sure. Have you done any unequalized frequency-response plots with the Peerless woofer in the AR cabinet? Share them with us on the forum. The AR frequency-response spec calls for +/- 2 dB variation from 1000 Hz down to resonance. There is less than a 1 dB peak at resonance in the AR-2ax. You could always equalize the Peerless, but that’s a pain. This is the age-old problem associated with retrofitting another woofer in a system specifically designed for a woofer with specific parameters.

Incidentally, the AR-2ax woofer’s response is not dominated by the crossover. You can play the AR-2ax in a “woofer-only” mode, and the woofer sounds pretty much the same as being in the circuit except for the very highest frequencies it has to reproduce. The coloration you refer to regarding the crossover is not really coloration, and is completely inaudible when you get back into the reverberant field. Just look at the acoustic-power response of the AR-2ax in a reverberant chamber; it is very extended and flat, with no dips and peaks of any consequence. It’s also not that the drivers are “phasy,” as you say, but that there is cabinet edge-molding diffraction and normal driver interference effects. This is only audible when you try to listen to this type speaker as though it were a “near-field” design, in which it was not intended to be used. The AR-2ax is a reverberant-sound design. One needs to listen in the far field to appreciate the qualities of an AR-2ax.

>...I do plan to have the original 10"

>Alnico drivers refoamed and to drop them back in, or maybe

>I'll compare them and keep the best.

When you talk of the “Alnico” woofers being refoamed, I assume you are talking about the 4-hole stamped-basket, pot-magnet version of the 10-inch woofer. This woofer was changed later on to include a ceramic-ferrite magnet in place of the pot magnet. The “Alnico” woofer is usually referred to as the 6-bolt version with the cast basket.

--Tom Tyson

[kkantor]

Here's a piece of total trivia:

I co-designed the 1727 with Knud Thorborg (sp?) at Peerless, working as a consultant to one of their larger Japanese customers. The 1727 was a specific attempt to strike a blow in favor of low Fs, which had really become unfashionable at that point in time. The driver later was added to their standard catalog.

It's nice that it is still in use.

Ken Kantor

kkantor@gmail.com

www.tymphany.com

www.aural.org

[kkantor]

In a complex world, most people tend to take over-simplified stances. A such, I agree with you that vented systems have their place. In fact, the contribution of the box to system compliance can NOT be determined simply by looking to see if there is a vent. There are sealed boxes that rely mostly on mechanical compliance, and vented boxes where the air spring is very significant. Villchur, himself, suggested to me that most small, vented speakers should be analysed as acoustic suspension systems over most of their range.

Having said that... I do think air is pretty damn linear for the purposes at hand. More linear than soft parts, for sure. However, here's a cheap shot self-promotion, for those going to AES next week:

http://www.aes.org/events/117/papers/C.cfm

Ken Kantor

kkantor@gmail.com

www.tymphany.com

www.aural.org

[Pete B]

My point, stated for Soundminded, is simply that air is not perfectly linear. Indeed nonlinearity of the air in the throat of horns is a major contributor to their high levels of distortion at high SPL as is discussed in the AES papers you reference:

"A significant part of distortion is generated in the phasing plug and the horn due to the nonlinear propagation of the high pressure sound waves."

I think we agree and I was just pointing this out to soundminded. We discussed this many years ago and several commented that the in box SPL would be so small for a typical subwoofer that air nonlinearity would not be an issue. I claimed otherwise but am not going to argue with people who are set on a conclusion. Here Linkwitz calculates an in box SPL of 159 dB for a typical 12" woofer:

http://www.linkwitzlab.com/images/graphics/enclosure-spl.gif

I'm really not interested in a back and forth. I'd say that at 159 dB air is going to contribute some significant nonlinearity. Linkwitz also shows the well known nonlinear P vs. V curve for air that Soundminded should take a look at.

[Pete B]

K. Kantor wrote:

"Villchur, himself, suggested to me that most small, vented speakers should be analysed as acoustic suspension systems over most of their range."

Certainly in the lower passband this is true, but it is the trivial case, in the low frequency region where there are performance differences that are of primary interest a vented design obviously cannot be analyzed as acoustic suspension, how would you account for the vent output? It is true that the general model for passive radiator systems can be used to model any of the common box types by setting the PR compliance very high for vented, and the mass very high for sealed. But acoustic suspension analysis will never work for the more complex systems, it is simply wrong. I don't expect this to go over well here in AR ville, but that's life. I really don't want to get into a back and forth but I'm going to offer my perspective since you post in response to my previous statements.

[johnieo]

Correct. Confusion may arise because there is more than one "Ideal Gas law." P*V = constant applies to an isothermal gas, whereas P*V sup k = constant applies to an adiabatic gas. An isothermal gas is one that can exchange heat with its surroundings to maintain its temperature constant. An adiabatic gas is one that is insulated from its surroundings. The gas near the driver cannot get rid of its heat energy and maintain constant temperature, because pressure waves travel at the speed of sound and heat flows much more slowly. Thus for high SPL in a small box, we use the adiabatic and not the constant temperature ideal gas law.

[kkantor]

1- I have designed many different types of bass enclosure, including sealed, vented, IB, TL, PR, horn, etc. I don't feel any one type holds >all< the advantages, and so I hope you don't dismiss my opinions as overly biased.

2- I am confused about your comments on the differences between AS and Vented at LF. Yes, below resonance, one is a 2nd-order and the other is a 4th-order. One is controlled below resonance, one is decontrolled. Clearly, they are different. However, through out much of the bass, some types of vented systems exhibit an alpha well over 1, indicating an upward shift in primary resonance that arises from "back pressure" or acoustic compliance. This acoustic compliance is in parallel with the suspension compliance. In a sealed system, it is always so. In a vented system, it is so above the port resonance. That's all I was asserting.

3- Back pressures in modern, compact vented systems are not at all negligible over much of the operating range of the woofer.

4- Most woofers designed for modern vented systems have much lower Fs than their earlier counterparts, to allow for the effect of acoustic compliance.

5- Air linearity? It's not perfect. It is a limitation in high power horns. It is rarely a limitation in a sealed speaker box.

I don't really thing these points are particularly controversial.

[Pete B]

Been meaning to get back to this,

I wrote:

>

>Stuffing deep in the box acts to dampen standing

>waves or box modes, and to also take the system from

>isothermal to a nearly adiabatic condition causing an increase

>in effective volume.

I inadvertently reversed adiabatic and isothermal in the above discussion, it should be:

and to also take the system from adiabatic to a nearly isothermal condition causing an increase in effective volume.

The stuffing or damping material being long and very thin strands has a large surface area which promotes heat transfer from the gas to the strands. The stuffing material offers "thermal inertia" and tends to keep the gas closer to constant temperature. A few here have offered the well known isothermal gas law relationship:

pV = K where K = nRT

Note that this is a reciprocal relationship when solved for p, obviouly V is in the denominator and it is non-linear:

p = K/V

Here's a discussion of the theory:

http://farside.ph.utexas.edu/teaching/sm1/...res/node52.html

[Pete B]

Hi John,

Thanks for your input, I understand the two cases of adiabatic and isothermal and I think your suggesting that the air near the driver, roughly that between the back of the driver frame and cone will follow the adiabatic case since there is no stuffing material.

I think we agree that the rest where there is stuffing material is under a nearly isothermal condition.

I wanted to point out that both cases are non-linear since even for the isothermal equation it is a reciprocal relationship which is non-linear:

p*V = K

solving for p:

p = K/V

It is approximately linear for small displacements or small delta V. This is why acoustic suspension systems had better linearity than most systems of the day with poor mechanical linearity. My point in bringing this up from the start is that many believe that the air is *perfectly* linear when it is not. Even Vilcher in his AES article states: "The substitution of acoustical for mechanical elasticity virtually replaces a nonlinear element by an almost ideally linear one, within the range of compressions and rarefactions involved."

This was probably very true at that time (1957), however today we have much better mechanical parts. His use of the words virtually, and "almost ideally" is interesting.

>Correct. Confusion may arise because there is more than one

>"Ideal Gas law." P*V = constant applies to an isothermal gas,

>whereas P*V sup k = constant applies to an adiabatic gas. An

>isothermal gas is one that can exchange heat with its

>surroundings to maintain its temperature constant. An

>adiabatic gas is one that is insulated from its surroundings.

>The gas near the driver cannot get rid of its heat energy and

>maintain constant temperature, because pressure waves travel

>at the speed of sound and heat flows much more slowly. Thus

>for high SPL in a small box, we use the adiabatic and not the

>constant temperature ideal gas law.

[Pete B]

>This was probably very true at that time (1957), however today

>we have much better mechanical parts.

Here's an example of a modern driver where care has been taken to provide a linear motor and mechanical system, no air spring is involved here. It is a plot of motor strength (Bl) versus cone displacement in millimeters. Most Bl curves do not even come close to this level of linearity. I'd say that driver motors and "soft parts" have improved when care is taken in selecting and engineering them:

http://www.adireaudio.com/images/eX6BL.gif

[soundminded]

If delta V/V is very small as I expect it would be, wouldn't the response be nearly linear?

I'm curious about why you chose the 1727 woofer instead of the 0452 XLS woofer. It costs of course 2 1/2 times as much and has much less high end so I guess it might require adding a low midrange driver but this one seems superb. OTOH, its lower Vas would probably be a better match for the small box, wouldn't it??

How did your project work out? Were you able to get the 1727 to match fairly well? I'd think the sensitivity is about the same as the AR and could easily be compensated for with the tweeter and midrange level controls and the damping could also be adjusted for optimal performance relatively easily by simply adjusting the amount of stuffing. High end response extends out to 2800 hz so that shouldn't be a problem either.