Acoustat ESLs - What Next?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ok, let me set some background. I study and work in physics, which gives me access to unbelievable toys, and also fuels my addiction to exotic hardware. A few years ago, a collegue (phishead8 here) and I read about Electrostatics and got hooked. After a few test models, phishead eventually produced quite a respectable pair of homebuilt ESLs, and I put together a decent pair of electrostatic headphones. We blew a few amps, and fun was had by all.

Later, I got into homebuilt amplifiers, and managed to assemble one from Randy Slone's book. One of the channels melted from a mistaken bias setting, but while it lasted, it worked well.

4 days ago, I was walking down the street on my way to work, and I saw a pair of large square black objects standing by the dumpster. Getting closer, I spotted the Acoustat logo on the base, and about 5 minutes later these monsters were in my living room. The power cables were shreaded, and they were a bit dusty, but other than that, they are in perfect shape. I replaced the power cables and tested them on a friend's yamaha power amp, and they sound just as good as you would expect.

So here's the first question. What next? I can't even tell what model these speakers are. The box on the back says Mk-141-c, and also has a mention of 'David Hafler' and 'Acoustat Medallion Transformers'. Looking this up, it appears the 141-c is the model of the filter and power supply box, not the speaker itself. I have no way of finding documentation, and don't know where to start. I'd love to have specs on frequency response and the rest of the electrical stats, but any information at all would be useful.

Which leads me to my next question. These puppies need low end help. I was considering a pair of sonotube-based transmission-line woofers. They seem simple enough to build, and I hear good things about the potential sound quality. The way I see it, this would entail an EQ, a passive-crossover, and another power amp for a bi-amping setup. So the question comes up: how do I choose an EQ, and how do I choose/design a crossover? Also, how big of a transmission line driver should I go with? For the amps, I was going to just repair my first Slone amp, and built another one for the Transmission-lines.

And finally, if and when I land on a final setup, does anyone have any experience with properly tuning a bi-amping system? I was imagining something like a metered microphone and a signal analyser, combined with a pink-noise generator, but maybe there's something less physics-lab that I'm not considering.

I know open ended 'help me' questions are out of line in most cases, but if anyone could just drop some off-the-top-of-their-head comments on any of these points, it would at least push me in the right direction. With a jump-start to my system like this dumpster-diving session, I want to make sure I get the rest of it right. Any help is greatly appreciated.
 
This is easy.

Those are the better of the two interface boxes for the Acoustats.
They have the "improved" HF transformers and slightly different values for the xover caps.

Depending upon the number of cells in the panels you have model number : 1 + 1, 2, 3 or 4.

You can find the number of cells with a bright light and looking carefully. They are full range cells that run from top to bottom, arranged in a slight radius left to right. The threes which were the most common version, are about 27" wide, maybe 30" with
the frames.

The 1+1 was stacked double high, one cell above the other.
(there was also a 2+2, 3+3 and the MONSTER 4+4 the latter two were called 6s and 8s...)

The cells are made with stuff that is essentially that plastic grid that is used for fluorescent light baffles... so you can poke at it with your finger and not hurt anything.

The frequency response of the system depends upon the number of cells, which also relates to the width of the dipole. Typically they are solid to 35Hz and extend up past 15kHz pretty flat.

The HF response is slightly limited by the mass of the cells.

The control on the back is a sort of HF level control, although it doesn't exactly work like an "L pad" on a dynamic driver since it is a resistance in series with the transformer that drives the HF, which works out to be a variable roll off filter, not a level control.
That's due to the nature of the load - a capacitor.

The drive system was patented by Jim Strickland of Acoustat.

It works by splitting the matching to the ESL cells at *two* impedances at once - creating an optimized match to the same cell at for two different frequency ranges at the same time.

Usually an ESL cell can be matched at *one* impedance. This gives you a tradoff of bandwidth vs. sensitivity. If you go for a wide bandwidth, you get much lower sensitivity. If you go for higher apparent output for a given input you get much less bandwidth... Strickland figured a way to match to both LF and HF at the same time. He used two separate transformers in a clever arrangement.

The best mod is to change the coupling caps to all Polypropylene film caps. (necessary, imho)

The next best mod is to bypass the speaker fuse entirely.

For best sound, imho, drive these with a big *** tube amp, preferably one with a good square wave response - and that usually means one without feedback (for this sort of load, especially). You can get away with as little as a beefed up Dyna ST-70 as long as you don't try for Rolling Stones levels and live within a reasonable range.

Optimum power for them is ~200watts.

They are a nominal 6 ohm load.

The speakers will reaveal faults quite nicely, and I would expect that the Yamaha will sound a bit thin and maybe not so natural as you might like... and your CD source will play a big role in what you hear too...

Regardless, you've scored a set of speakers that will rival speakers costing many $thousands$ of dollars even today.

Congrats!

:D

_-_-bear :Pawprint:
 
I believe the MK141 interface was the one developed for the panels that required/came with a sep. subwoofer (my ancient memory does lapse). Acoustat panels were in widths of 7" and 8"
(the model 3 used 2 - 8" panels and 1 - 7" panel). The model 3 and 6 were each 28" wide, the 2's and 2+2's were 20" wide, I believe. Yours could be a model 1 (about 14" wide I think). I
still have a pair of model 3's (medallion updates) that I no longer use. They are real nice, just too big for the living room. I did
have to replace one of the panels in each many years ago, the signal wire came unglued in a few places on the grid and would vibrate against the membrane. If anyone is interested, they are for sale (local pickup in Atlanta area). Very good condition.
 
Photos and Some More Questions

First of all, a big thanks to everyone who posted. It's really helped out having the communities take on these things, at the very least it assures me that I'm not completely off-base on my thinking.

Taking a closer look at these speakers, I'm thinking they're a single esl panel each. Witness the (slightly out of focus) photos:
http://sam.phys.lsu.edu/People/hanson/acoustat.php

I tried bear's advice, and using a flashlight I could trace what can only be the edges of a single opaque diaphragm. I don't know what Acoustat used for its films, but its evidently not-transparent. I didn't see any breaks in the light in either the top-to-bottom or left-to-right direction. So is this called the Model 1? I'm a little hazy on Acoustat's naming convention.

As far as the box goes, slydwz was right: they look set up to couple to a seperate low-frequency driver. Whether this sub came with the speakers or was supposed to be provided by the customer I don't know, but at any rate they weren't in the garbage heap with the ESLs so I suppose I'll have to improvise.

So this suggests some further questions:

First of all, is this box set up to provide high-pass to the panels, and low-pass to the connection labeled "woofer feed (fused)"? Or is the woofer feed a full-range pass through? It would seem to be counter-productive to allow highs through to the woofers, but maybe there was some motivation here I don't see. From looking at the circuit, it looks like the two pairs of speaker posts are electrically identical, but maybe there's some subtlty of electronics I'm missing here.

If it really is a built in crossover, its presumably set up for a specific woofer set. Assuming I can find out the crossover points and the frequency response of the panels, is there anything special I'm looking for when I build/buy a set of woofers to accompany these speakers? I was considering building a set of the "tuned sonotube with a speaker shoved in the end" transmission lines. Is there anything I should be aware of when it comes to finding partners for these ESLs? How do you deal with the difference in sensitivities between magnetic-coil woofers and electrostatic panels?

And finally, the amplifier. I'm starting to think now that my Randy Sloane amp isn't going to cut it. The rails swing +/- 42 volts on that thing, and considering a nominal 8 ohm load, that's (42^2) / 8 = 220.5 watts of peak draw from the power supply per channel. Factoring in the efficiency of the amp and other people's results from the same design, it appears to be around 80 watts per channel delivered power. This seems a bit light for this application, especially if I intend to power the low-end drivers as well off this one amp.

Combining this with the fact that my Sloane amp is in pieces at the moment, I was considering a pre-built alternative. Am I right in my take that with whatever low-frequency drivers I end up with, I can wire them into the "woofer feed" posts, and then power both the panels and the woofers from one amp? This would appeal to my college-student budget, and would also remove the need for another amp and signal-level active crossover, but would presumably require something toward the high-power end of the available consumer spectrum. Are there any special requirements for these speakers? Our homebuilt panels had funky loads and blew a few amps, but these are presumably "consumer" speakers and are a bit more tolerant. I appreciate the aesthetic engineering of a fine audio amplifier, but unfortunately I'm on a senior student's budget and won't be able to go high-end. I'm not necessarily into bowel-churning volumes of music, but I do occassionally listen above background-music levels, and also use this system to watch movies. Any amplifier recommendations?

Thanks again for all your interest and help, I'm blowing tons of time researching and considering my options here, and any input anyone else can offer would be much obliged.
 
Re: Photos and Some More Questions

G-Daddy said:
First of all, a big thanks to everyone who posted. It's really helped out having the communities take on these things, at the very least it assures me that I'm not completely off-base on my thinking.

Taking a closer look at these speakers, I'm thinking they're a single esl panel each. Witness the (slightly out of focus) photos:
http://sam.phys.lsu.edu/People/hanson/acoustat.php

I tried bear's advice, and using a flashlight I could trace what can only be the edges of a single opaque diaphragm. I don't know what Acoustat used for its films, but its evidently not-transparent. I didn't see any breaks in the light in either the top-to-bottom or left-to-right direction. So is this called the Model 1? I'm a little hazy on Acoustat's naming convention.


As I said, I was wrong about the "interface box". The 121 was/is for the two part speaker... I actually have the schematic somewhere, and if you email me I can try to find it sometime before you forget that you emailed me! ;)

Yes, the Acoustat cells used an opaque coating.

The jpegs make it look like a 1+1 variant - two cells stacked...
the cells are about 40" long each.

As far as the box goes, slydwz was right: they look set up to couple to a seperate low-frequency driver. Whether this sub came with the speakers or was supposed to be provided by the customer I don't know, but at any rate they weren't in the garbage heap with the ESLs so I suppose I'll have to improvise.

So this suggests some further questions:

First of all, is this box set up to provide high-pass to the panels, and low-pass to the connection labeled "woofer feed (fused)"? Or is the woofer feed a full-range pass through? It would seem to be counter-productive to allow highs through to the woofers, but maybe there was some motivation here I don't see. From looking at the circuit, it looks like the two pairs of speaker posts are electrically identical, but maybe there's some subtlty of electronics I'm missing here.

If I can fish out the xover we'll know for sure - but if you put an ohmmeter on the two sets of binding posts and the answer is DC, zero ohms, and you don't see any inductor inbetween, then its just a pass through...

But have a listen to the panels, since the 1+1 can be rather full range... although the interface may not be set up for full range...
but it will likely be set to go down ~125-250 Hz range IF it is set up to take an external sub...

heck you can tell pretty quickly, just put a random speaker on the "woofer" speaker connections and see what it does!! :D

If it really is a built in crossover, its presumably set up for a specific woofer set. Assuming I can find out the crossover points and the frequency response of the panels, is there anything special I'm looking for when I build/buy a set of woofers to accompany these speakers? I was considering building a set of the "tuned sonotube with a speaker shoved in the end" transmission lines. Is there anything I should be aware of when it comes to finding partners for these ESLs? How do you deal with the difference in sensitivities between magnetic-coil woofers and electrostatic panels?

Adjust the amplifier level??


And finally, the amplifier. I'm starting to think now that my Randy Sloane amp isn't going to cut it. The rails swing +/- 42 volts on that thing, and considering a nominal 8 ohm load, that's (42^2) / 8 = 220.5 watts of peak draw from the power supply per channel. Factoring in the efficiency of the amp and other people's results from the same design, it appears to be around 80 watts per channel delivered power. This seems a bit light for this application, especially if I intend to power the low-end drivers as well off this one amp.

I'd build up a nice toobe amp for the top end... you'd likely find that it would produce the best sound per $$ spent in the end...

Triodes please, not pentodes. If you must, triode strapped pentodes.

But ur amp will run them pretty well, just don't try to crank them up until it is "LOUD"!!

Combining this with the fact that my Sloane amp is in pieces at the moment, I was considering a pre-built alternative. Am I right in my take that with whatever low-frequency drivers I end up with, I can wire them into the "woofer feed" posts, and then power both the panels and the woofers from one amp? This would appeal to my college-student budget, and would also remove the need for another amp and signal-level active crossover, but would presumably require something toward the high-power end of the available consumer spectrum. Are there any special requirements for these speakers? Our homebuilt panels had funky loads and blew a few amps, but these are presumably "consumer" speakers and are a bit more tolerant. I appreciate the aesthetic engineering of a fine audio amplifier, but unfortunately I'm on a senior student's budget and won't be able to go high-end. I'm not necessarily into bowel-churning volumes of music, but I do occassionally listen above background-music levels, and also use this system to watch movies. Any amplifier recommendations?

Tubes for the top. Anything you can find in soylent state for the woofer, big, cheap is the criteria. Use a simple speaker level xover to start to save $$ or a simple line level passive or active (op amp) LP filter to drive the amp for the woofer. Beherenger sells used super cheap for active
xovers - just DO NOT use the HP function to drive the ESLs!! (will sound like death)

Crown has some (new on the market) ungodly powerful and inexpensive amps that bridged make HUGE power... probably horrendous sounding for full range, but jim dandy for bass!! Anything like that will work... even an old Carver (gasp!) will produce the requisite watts.

My preference is for a PR or sealed box sub to match - keeping the box size down and reducing the delay in the low bass compared to a TL... ymmv.


Thanks again for all your interest and help, I'm blowing tons of time researching and considering my options here, and any input anyone else can offer would be much obliged.


OH - see that little BLUE electrolytic cap in front of the big power resistor?? LOSE IT! Replace immediately with a film cap(s) or equal value, best with a polyproylene...

;)

_-_-bear :Pawprint:
 
Changed My Mind!

Ok, I changed my mind a little bit...

For you the cheapest way to get going is to get *any* cheap amp with power you can find... used. Scout ur local pro sound companies, they usually have bins full of trashed stuff that they will sell you a few chassis of and you can fix... or even stuff they no longer take on the road having switched brands, etc...

Ebay of course... too...

Your ideal is in the 200watt + class for solid state... but NOT too big... 500 watts is pushing it a bit for these...

Anyhow, I think the xover point for the panels is likely to be ~500Hz.

This changes the equation on the woofer side... since now you need really cracking good midbass...

In my mind that means a line source of multiple drivers is your only good choice here... what you want are really flat little woofers... preferably with the lowest FS you can find... the other atlernative is to put up a line source of midbass drivers and add in a nice 12" long throw subwoofer(s), like the Titanic or the Adire offerings, etc...

You could try to use one of those subwoofer drivers up to the panels, but that wouldn't really be terribly satisfactory...

You could use a set of up to 8" woofers, like 4 or more in a vertical row... although 6" or 5.25" might be better... the key is low cost per driver and flat response in the passband... even lowly "computer monitor" wide range drivers will work fine, assuming a flat response in the pass band (you'll just roll off everything above 500Hz. ! The Tang Band drivers are super in this sort of application, but maybe pricier than you might wish...
there are others and since you don't need to be terribly concerned about the response above 500Hz. you get a free ride in terms of selecting drivers just for below 500Hz.! :D

Option 2:

Modify the interface box to be a 141! That only requires relatively minor changes and the acquisition of a second transformer, some caps (high voltage) and finding a place to put it.... that would give you response down to ~40Hz. very flat...

Option 3: Look for someone selling some 141 interface boxes, they do get sold from time to time...

Option 4: there was a guy in Florida who was servicing and selling parts for Acoustat, you could buy the parts from him... although it is just a high quality output transformer, run backwards...

Option 5: scrap the interface box and build a direct drive amp, run it full range.

Hope this helps focus the project... but once done you'll have a speaker system that is pretty fine indeed!

_-_-bear :Pawprint:
 
A-ha!

Ahhh, so Model 1's, its nice to know what I'm dealing with, then. 12" sub you say? Interesting!

So whats the story on this attached subwoofer? Presumably, there were two of them. What were their frequency range? What was their wattage rating? What was their construction, how were they configured? It would help me immensely to know what Acoustat considered appropriate to accompany these speakers.

Specifically, I'd really like to know what sort of electronics went between the woofer drivers and the "woofer feed" terminals on the back of the ESLs. I haven't hooked up a regular speaker there yet to see, but I suspect I'll need something to low-pass the signal that the woofers see.

If anyone has any information on the specific nature of the low-frequency speakers that Acoustat provided with these Model 1's, I would love to hear it.

For what its worth, I thought through Bear's suggestions, and my plan now is to find a decent modern consumer amplifier that can generate respectable power-per-channel, say in the upper 100's to lower 200 watts per channel. I'll use my current speakers in conjunction with my ESLs by going from amp to ESL-in, and then from ESL "woofer feed" out to the standard speakers.

This will be the "for the time being" configuration, while I research and build a vertical array line source pair. I'm pretty good with woodworking and have access to a limited supply of tools at work, so it'll just be a matter of doing my reading and scratching together the parts.

If that's not enough then perhaps I'll track down a powered sub-woofer to cover the very low end, but I suspect I can live with the line source and ESL combo.

So, I'm in the market for a new amp. I've bought a few before, and helped out some friends on their acquisitions, but I've never bought an amplifier with the intent to saddle it with electrostatic loudspeakers. I know its probably not the best choice in terms of sound quality, but I'm considering a high-end integrated receiver, something that can swing dolby digital decoding, and that has enough surround channels to cover any future home-theatre expansion I go with. Am I going to run into any troubles with stability or overheating? Are there any amps of this sort that are considered more reliable than others in dealing with unusual loads? Thanks everyone for the help so far, I really appreciate it.
 
Specifications- Acoustat Model One:

Frequency Response: 30-18kHz +/-3dB
Sound Pressure Level 108dB measured @15' in a 14' X 18' room (music peaks)
Minimium Power Requirements 100 watts/channel.
Nominal Impedance 4 ohms
Power Consumption 5 Watts (120V, 50/60Hz)
Dimensions 72" X 11" X 3-1/2"
Bass Module (only one is used ) 18.5" X 18.5" X 18.5"
Weight 65lbs (Speaker and Interface) 37lbs (Bass Module)

Woofer was a single 10" driver.

There was a passive XO in the Bass Module. 2nd order filter consisting of a 4.5 mH inductor series inductor, parallel shunt cap 330mfd, with a 10 ohm power resistor in parallel.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.