I'm in the bay area.
Here are 2 pics of the mylar rolled out on a large light table to back light it.
With and without overhead room lights and flash.
The AK site link is full of interesting company details. Since the Model 5 was an economy model and the PCB style was to save production costs could they also have used up less than perfect materials in the production to keep costs down?
Maybe the weak coating is still sufficient for good charge coverage.
And I will now pull my panels and give them the identical back lighting to see how the mylar looks. It is really very difficult to clearly see through the slots.
I also looked at the panels last night with the top plate of the speaker removed and there are little bits of mylar hanging out of the frames in the corners. Small-maybe .125 x .375 or slightly bigger. But I think just enough to work with. So when I pull the panels I can snip them off to measure.
Did you get a chance to listen to them in detail?
One of mine is maybe 1 or 2 dB lower output but otherwise seems fine.
Here are 2 pics of the mylar rolled out on a large light table to back light it.
With and without overhead room lights and flash.
The AK site link is full of interesting company details. Since the Model 5 was an economy model and the PCB style was to save production costs could they also have used up less than perfect materials in the production to keep costs down?
Maybe the weak coating is still sufficient for good charge coverage.
And I will now pull my panels and give them the identical back lighting to see how the mylar looks. It is really very difficult to clearly see through the slots.
I also looked at the panels last night with the top plate of the speaker removed and there are little bits of mylar hanging out of the frames in the corners. Small-maybe .125 x .375 or slightly bigger. But I think just enough to work with. So when I pull the panels I can snip them off to measure.
Did you get a chance to listen to them in detail?
One of mine is maybe 1 or 2 dB lower output but otherwise seems fine.
Attachments
Last edited:
DaveG,
It is possible that Bev used less-than-stellar materials for the Model 5, but I doubt that the Mylar (or PET or whatever) was the place to save money. I'd guess he still had a large spool of it and continued to use it for the newer models.
You are right that it is possible that the panels sound OK in spite of the less-than-stellar diaphragm appearance. One of the capacitance values is fine, and the other is not that far off. And what I see through the stator holes is where the field is not that active. I've not had much of a chance to hear them yet, since my amplifier has been in the shop (Saturday!).
Still, it seems crazy not to replace a 35-year-old diaphragm when it is a known failure point.
I agree that it is difficult to see through the epoxy slots. They are pretty deep.
If you head to North Lake Tahoe, you practically drive by my place.
It is possible that Bev used less-than-stellar materials for the Model 5, but I doubt that the Mylar (or PET or whatever) was the place to save money. I'd guess he still had a large spool of it and continued to use it for the newer models.
You are right that it is possible that the panels sound OK in spite of the less-than-stellar diaphragm appearance. One of the capacitance values is fine, and the other is not that far off. And what I see through the stator holes is where the field is not that active. I've not had much of a chance to hear them yet, since my amplifier has been in the shop (Saturday!).
Still, it seems crazy not to replace a 35-year-old diaphragm when it is a known failure point.
I agree that it is difficult to see through the epoxy slots. They are pretty deep.
If you head to North Lake Tahoe, you practically drive by my place.
I completely agree about the mylar. There are several age and failure problems that can occur and that is why I purchased the mylar myself when I saw it.
But the panels are also fairly easy to remove any time.
Myself...I would get everything else done...get them up and running so there is a reference sound to go by. And if they are raspy or distorted or super weak that can't be explained by dirt stuck in the panel or a failed HV diode then you rebuild.
Maybe you also have some other speakers already for sound comparison.
Prior to me looking at the mylar I bought closely and the photo of your transducer I would have said yours are shot. Now I would get them playing and have a good long listen first.
A note also on the woofers: mine have a non standard DCR. I know some have replaced them, maybe because theirs were fried, but if yours are original maybe just refoam and then go from there.
Of course bi-amping, better separate woofers, etc etc could be better but the originals in mine, while not the best available, do OK for themselves.
And there is mention on the AK thread that all the original crossover capacitors measures OK. I actually did replace mine with cheap polypropylene ones and of course still have all the originals.
And we will keep looking for good mylar and info on all this.
But the panels are also fairly easy to remove any time.
Myself...I would get everything else done...get them up and running so there is a reference sound to go by. And if they are raspy or distorted or super weak that can't be explained by dirt stuck in the panel or a failed HV diode then you rebuild.
Maybe you also have some other speakers already for sound comparison.
Prior to me looking at the mylar I bought closely and the photo of your transducer I would have said yours are shot. Now I would get them playing and have a good long listen first.
A note also on the woofers: mine have a non standard DCR. I know some have replaced them, maybe because theirs were fried, but if yours are original maybe just refoam and then go from there.
Of course bi-amping, better separate woofers, etc etc could be better but the originals in mine, while not the best available, do OK for themselves.
And there is mention on the AK thread that all the original crossover capacitors measures OK. I actually did replace mine with cheap polypropylene ones and of course still have all the originals.
And we will keep looking for good mylar and info on all this.
DaveG,
I've not yet checked out the HV supply, but that is the other obvious failure point. I do have a HV probe, so it is no problem to check. I also have a couple of power supplies that deliver about the right voltage that I can use for comparison if I want.
The woofers on mine look new, and sounded pretty weak when I auditioned the pair. That is on the list to look at too.
On the Model 5 being an economy model, it is only because it is half of a Model 6, which is very similar to the Model 3, which in turn is similar to the model 2 SWx but with transformers instead of the direct-drive amplifier. All of the third Beveridge company incarnation (Cal Audio, I think) models used the circuit board stators; he also diddled with the radiation pattern and getting the costs down. Rick claims that they lost money on all of the old models, so cost reduction and increasing SPLs were an emphasis of Cal Audio.
I've not yet checked out the HV supply, but that is the other obvious failure point. I do have a HV probe, so it is no problem to check. I also have a couple of power supplies that deliver about the right voltage that I can use for comparison if I want.
The woofers on mine look new, and sounded pretty weak when I auditioned the pair. That is on the list to look at too.
On the Model 5 being an economy model, it is only because it is half of a Model 6, which is very similar to the Model 3, which in turn is similar to the model 2 SWx but with transformers instead of the direct-drive amplifier. All of the third Beveridge company incarnation (Cal Audio, I think) models used the circuit board stators; he also diddled with the radiation pattern and getting the costs down. Rick claims that they lost money on all of the old models, so cost reduction and increasing SPLs were an emphasis of Cal Audio.
No, I got it from Flagship converters up here in New England. You can't get anything less than 0.5mil
Well, you can get thinner, but it is costly. Here are two links from Goodfellow for 6 um metallized Mylar, the first for single-sided, and the second for double-sided:
Single-Sided
Double-Sided
Their markup must be huge.
Hmmm. That didn't work right. Try following it from here:
Polyethylene terephthalate Metallized Film - online catalog source - supplier of research materials in small quantities - Goodfellow
Single-Sided
Double-Sided
Their markup must be huge.
Hmmm. That didn't work right. Try following it from here:
Polyethylene terephthalate Metallized Film - online catalog source - supplier of research materials in small quantities - Goodfellow
Last edited:
I just had a very interesting visit with a fellow who has a lot of Beveridge "guts" from either the Model 2 or 2 SW. I talked with him about the Mylar, and he claimed Rick said that 2 mil (or 50 um) Mylar was used. Rick further said that they had tried 1 mil (25 um) Mylar, but that they couldn't tension it properly.
Then he remembered that he had Mylar from one of the electrostatic cells that he had disassembled. I measured it, and it was 25 um. I have a sample here, and will do the conductivity measurement.
The roll that Modjeski has allegedly is from the same batch that Bev used.
How these all fit together isn't clear. But the main point is that the Mylar is a LOT thicker than is normally used in electrostatics.
Then he remembered that he had Mylar from one of the electrostatic cells that he had disassembled. I measured it, and it was 25 um. I have a sample here, and will do the conductivity measurement.
The roll that Modjeski has allegedly is from the same batch that Bev used.
How these all fit together isn't clear. But the main point is that the Mylar is a LOT thicker than is normally used in electrostatics.
I'm not saying this is what is happening, but ESL transformers tend to peak up in the upper treble somewhere from the inductance of the transformer and the capacitance of the cells. So if you designed an ESL carefully, you could use thicker mylar that rolls off say at 15KHz and also use somewhat crappy transformers than start peaking at 15KHz and get effectively better treble extension than just looking at the mylar thickness would imply.
Sheldon
Sheldon
I managed to snip 4 small slivers from the corners of the 1981 Model 5 epoxy transducers.
The 4 stacked measured around .0043 to maybe .0045 and were a bit rough and wrinkled. So pretty close to the .0009 samples from the other roll.
I did see some scratches and small clear holes but as a quick first inspection appears much better looking than the mylar in the photo I posted.
The 4 stacked measured around .0043 to maybe .0045 and were a bit rough and wrinkled. So pretty close to the .0009 samples from the other roll.
I did see some scratches and small clear holes but as a quick first inspection appears much better looking than the mylar in the photo I posted.
It has been a while since I updated this thread, so here it is!
Some progress has been made, but not as much as I'd hoped. Real life got in the way. I had my amplifier refurbished -- an old Mac MC2105, which is not great, but stable and good enough for now. And before I started, I wanted to measure what I have now, and see if any changes made could be measured.
I've not done audio measurements before, and with Sheldon's advice, I now have a calibrated Dayton EMM-6, a Focusrite Solo, a microphone stand, and yards of cable. I had to get a few additional adapters to get everything to work together. For software I started with REW, which is easy to use and free. It seems to be limited to swept sine waves (is that right?), but it is very powerful from those basic data. It really is a pretty slick program.
The measurements only partially explain what I hear. The bass indeed is awful. Mid/Upper bass (say, 80 to about 300 Hz) is down about 10 dB in level from the 300 Hz and up levels. Low bass is flatulent (by ear), and is lumpy. I'd bet that the replacement woofer used was not the right one, since this sort of response never would have been released. The bass of Bevs always has been bad, but not *this* bad.
From 300 Hz and up, the measurements are not too bad, giving about +- 3 dB from there to about 10 kHz, where response falls rapidly (the thick membrane?). Impulse response is pretty clean, with some ringing that dies down after about 1.5 msec to background levels. They invert phase.
Waterfall plots are interesting. They are pretty uniform with frequency over the entire frequency range, though they wobble a bit. But the decay time is long, at about 50 msec.
Harmonic distortion is below 1%, which is the limit of my (noisy!) room. REW seems to have no good way to measure intermodular distortion.
So what's wrong? Bevs always have had an ability to float a stereo image that is unique and wonderful. These don't do that -- there's no image at all. They also sound painfully bright, and other than the bass and maybe the ringing problems, there is not really anything else to suggest why that is so.
My best guess is that the diaphragm is not uniform, and that there are some measurement challenges. The most common way to measure is to do so in the very nearfield, so that the response one measures is unencumbered by the room. Here it is very difficult to get a true nearfield measurement, since the electrostatic element is buried behind an acoustic lens. The room is excluded, but what exactly one is measuring isn't clear. I'm reluctant to pull out the elements and measure those independently because of domestic concerns. I found the expected difference in nearfield and farfield frequency response slopes. (I note that "farfield" here is not far for a line source -- it was about 8 feet.)
Also, I only checked one portion of the lens, not along its radius or height. It certainly is possible (or likely!) that other portions would measure differently.
I guess the best path forward is to replace the bias/crossover elements that degrade with age (electrolytic capacitors, resistors and diodes), find a better woofer, and replace the electrostatic element diaphragm and repair any other damage.
Stay tuned!
Some progress has been made, but not as much as I'd hoped. Real life got in the way. I had my amplifier refurbished -- an old Mac MC2105, which is not great, but stable and good enough for now. And before I started, I wanted to measure what I have now, and see if any changes made could be measured.
I've not done audio measurements before, and with Sheldon's advice, I now have a calibrated Dayton EMM-6, a Focusrite Solo, a microphone stand, and yards of cable. I had to get a few additional adapters to get everything to work together. For software I started with REW, which is easy to use and free. It seems to be limited to swept sine waves (is that right?), but it is very powerful from those basic data. It really is a pretty slick program.
The measurements only partially explain what I hear. The bass indeed is awful. Mid/Upper bass (say, 80 to about 300 Hz) is down about 10 dB in level from the 300 Hz and up levels. Low bass is flatulent (by ear), and is lumpy. I'd bet that the replacement woofer used was not the right one, since this sort of response never would have been released. The bass of Bevs always has been bad, but not *this* bad.
From 300 Hz and up, the measurements are not too bad, giving about +- 3 dB from there to about 10 kHz, where response falls rapidly (the thick membrane?). Impulse response is pretty clean, with some ringing that dies down after about 1.5 msec to background levels. They invert phase.
Waterfall plots are interesting. They are pretty uniform with frequency over the entire frequency range, though they wobble a bit. But the decay time is long, at about 50 msec.
Harmonic distortion is below 1%, which is the limit of my (noisy!) room. REW seems to have no good way to measure intermodular distortion.
So what's wrong? Bevs always have had an ability to float a stereo image that is unique and wonderful. These don't do that -- there's no image at all. They also sound painfully bright, and other than the bass and maybe the ringing problems, there is not really anything else to suggest why that is so.
My best guess is that the diaphragm is not uniform, and that there are some measurement challenges. The most common way to measure is to do so in the very nearfield, so that the response one measures is unencumbered by the room. Here it is very difficult to get a true nearfield measurement, since the electrostatic element is buried behind an acoustic lens. The room is excluded, but what exactly one is measuring isn't clear. I'm reluctant to pull out the elements and measure those independently because of domestic concerns. I found the expected difference in nearfield and farfield frequency response slopes. (I note that "farfield" here is not far for a line source -- it was about 8 feet.)
Also, I only checked one portion of the lens, not along its radius or height. It certainly is possible (or likely!) that other portions would measure differently.
I guess the best path forward is to replace the bias/crossover elements that degrade with age (electrolytic capacitors, resistors and diodes), find a better woofer, and replace the electrostatic element diaphragm and repair any other damage.
Stay tuned!
I don’t have direct experience with Beveridge units, but a few random thoughts:
1) You may be able to get more information about coating condition by measuring capacitance from diaphragm to stator for each panel.
2) Your idea about measuring up/down and side-to-side on the lens seems like a good one. In theory the path through the lens isn’t going to affect things much, so measuring the lens output is likely similar to measuring at the panel. Near the panel edges, you’d expect some drop-off in output/bass, but if you have any other major oddities in your measurements, it might point you toward panel problems. Checking the left versus right speaker might point you to something also. Overall, you’re looking for things that don’t fit the trend; you wouldn’t expect exactly the same nearfield measurements everywhere though.
3) You can also look at impulse response to check polarity of the individual panels and left/right speakers.
4) I’d try to measure bias before assuming it’s the source of your troubles, unless you have a symptom (like left or right speaker having significantly more output). BK Precision’s PR 28A is a 1000x attenuator probe that comes in handy when working with ESL bias supplies.
1) You may be able to get more information about coating condition by measuring capacitance from diaphragm to stator for each panel.
2) Your idea about measuring up/down and side-to-side on the lens seems like a good one. In theory the path through the lens isn’t going to affect things much, so measuring the lens output is likely similar to measuring at the panel. Near the panel edges, you’d expect some drop-off in output/bass, but if you have any other major oddities in your measurements, it might point you toward panel problems. Checking the left versus right speaker might point you to something also. Overall, you’re looking for things that don’t fit the trend; you wouldn’t expect exactly the same nearfield measurements everywhere though.
3) You can also look at impulse response to check polarity of the individual panels and left/right speakers.
4) I’d try to measure bias before assuming it’s the source of your troubles, unless you have a symptom (like left or right speaker having significantly more output). BK Precision’s PR 28A is a 1000x attenuator probe that comes in handy when working with ESL bias supplies.
Thanks, Matt!
I've done so, and the capacitance is pretty close to what is written on the panels.
I did, and the polarity is the same (and negative) for both speakers. Still, it is easy enough to swap leads on one speaker and try it. It sure sounds like it is running out-of-phase.
It is on the list to do. I have a high-voltage probe and decent meters (an old BK and an Agilent).
I've done so, and the capacitance is pretty close to what is written on the panels.
I did, and the polarity is the same (and negative) for both speakers. Still, it is easy enough to swap leads on one speaker and try it. It sure sounds like it is running out-of-phase.
It is on the list to do. I have a high-voltage probe and decent meters (an old BK and an Agilent).
Hi DrJJ,
If I have understood this correctly the worst initial problem is the level difference between the woofer and panel.
If yours are similar to mine there are no level adjustments.
I guess mine are the original woofers since they blend nicely.
I would guess it will be mostly impossible to find a woofer that will exactly match without padding down the panel and bi-amping will be the only practical method.
Do yours have the 8 banana jacks for the speaker cables?
If I have understood this correctly the worst initial problem is the level difference between the woofer and panel.
If yours are similar to mine there are no level adjustments.
I guess mine are the original woofers since they blend nicely.
I would guess it will be mostly impossible to find a woofer that will exactly match without padding down the panel and bi-amping will be the only practical method.
Do yours have the 8 banana jacks for the speaker cables?
DaveG,
While the woofer level is annoying, it is the lack of imaging and brightness that are the primary problems. The imaging is what makes the Bev special, and without that they are not very interesting.
I figured I'd have to biamplify anyway, since the Bev woofers always are an issue, and they set the entire "case" into resonance. So yes, I'll make a separate woofer arrangement, biamplify, and not use the one in the Bev cabinet.
If by your banana jack question you whether it has separate connections for biamplifying or for using the crossover, the answer is "yes." One thing to try is to use my amp with one channel and a digital crossover to see if it helps.
While the woofer level is annoying, it is the lack of imaging and brightness that are the primary problems. The imaging is what makes the Bev special, and without that they are not very interesting.
I figured I'd have to biamplify anyway, since the Bev woofers always are an issue, and they set the entire "case" into resonance. So yes, I'll make a separate woofer arrangement, biamplify, and not use the one in the Bev cabinet.
If by your banana jack question you whether it has separate connections for biamplifying or for using the crossover, the answer is "yes." One thing to try is to use my amp with one channel and a digital crossover to see if it helps.
My guess would be at the moment the woofer is the problem with the imaging and brightness. Mine is a 250 hz xover.
If you have a spare amp to use or can borrow you can rig up a simple level control and bi amp with the internal crossover woofer xover. You could even do it for one channel with your existing amp.
I use an ALPS 250K in a little plastic box with some old cheap RCA cables that came with a dvd or tv or something as a simple stereo level contol.
I use it as a safety valve when playing with stuff, sound cards and testing, so one click of the mouse does not go to ear splitting full output.
Use it to drop the panel level the 10 dB
Parts Express has them for about $15
Then you are using the original crossover points so when you get the levels correct hopefully you will hear the correct original fullness.
If you have a spare amp to use or can borrow you can rig up a simple level control and bi amp with the internal crossover woofer xover. You could even do it for one channel with your existing amp.
I use an ALPS 250K in a little plastic box with some old cheap RCA cables that came with a dvd or tv or something as a simple stereo level contol.
I use it as a safety valve when playing with stuff, sound cards and testing, so one click of the mouse does not go to ear splitting full output.
Use it to drop the panel level the 10 dB
Parts Express has them for about $15
Then you are using the original crossover points so when you get the levels correct hopefully you will hear the correct original fullness.
If I read you right, you are suggesting bi-wiring so that the existing crossover can be used. I'll have to check that this will work with the connections. If that is possible, then I can do one side with my existing amp -- it has output potentiometers.
I don't have a spare amp at the moment, but an inexpensive plate amp may be good enough for the woofer(s). I suspect that quality here is not really limiting.
To touch back, a lot of this diagnostic work isn't really critical. The speaker is 35 years old, and most parts need to be replaced or rebuilt. The costs are modest, and other than the stators, it is not much work once you get into the thing. Also, I only want to do this once, so it makes sense to do as much as is reasonable.
I was curious to see if there were major anomalies that could be measured that correlated with what I hear. It is no surprise that the woofer level is way down. But that nothing else seemed terribly amiss was a surprise.
I have to admit that the measuring is very cool! Remarkable powerful, though it is hard to do unambiguously.
I don't have a spare amp at the moment, but an inexpensive plate amp may be good enough for the woofer(s). I suspect that quality here is not really limiting.
To touch back, a lot of this diagnostic work isn't really critical. The speaker is 35 years old, and most parts need to be replaced or rebuilt. The costs are modest, and other than the stators, it is not much work once you get into the thing. Also, I only want to do this once, so it makes sense to do as much as is reasonable.
I was curious to see if there were major anomalies that could be measured that correlated with what I hear. It is no surprise that the woofer level is way down. But that nothing else seemed terribly amiss was a surprise.
I have to admit that the measuring is very cool! Remarkable powerful, though it is hard to do unambiguously.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.