Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
Firstly the voltage is the highest of the lot.
Secondly, it is a one panel ELS.
Please remember why we have so much bass when the others struggle. The bass transformer has more turns than the High Frequency transformer. So the two transformers actually acts sort of like a RIAA section.
One transformer, one panel when driven only have highs and no bass.
If we go direct drive, we need some form of mixer, or crossover network.
The biggest problem is that a direct drive amp is almost illegal to sell. The only way to do it is a DIY Kit with very limited liability - maybe.
Unlike ordinary amps, we can touch the speaker cables and not get electrocuted. With direct drive, touching the speaker wires will kill.
Ralph has given a lot of thought to this problem, and I have not. So I leave it to Ralph to comment definitively, but the SL panel is no different from any other full range ESL in that it behaves like a giant capacitor. Thus the impedance is very high at low frequencies. Parameters like mylar to stator distance, bias voltage, size of the panel, etc, all affect the "size" of the capacitative load, but this is why we need the two transformers in parallel or the SL bass transformer alone to get adequate bass response from an A1PX or the like. Direct-drive amplifiers can be designed so as to produce more "drive" or power into bass frequencies and to gradually reduce the drive as frequency increases, so as to deliver power as an inverse complement of the load presented by the speaker. I am sure this is what Acoustat did back in the day; I KNOW that is how the Beveridge dd amplifiers work. So a "crossover network" as we know it should not be necessary. However probably there is some sort of filter that must be built into the input stage. Not as damaging as a conventional speaker crossover, because very low values of capacitance could be used, therefore of very high quality. Ralph?
Sheesh. You could do a filter, or do something in a feedback loop, although I imagine that latter would be eschewed by a number of the crowd here :)
Was my thesis above basically correct?
If so, I guess it is a trade-off between small value caps in the signal path (and the dangers of very high voltages) vs transformers in the signal path. This is a familiar dilemma, encountered at many junctions in the audio chain.
You don't need coupling caps if you are doing a direct drive amp- you can obtain the bias from the amp itself. But- what a scary thing to ever have to go near!
Perhaps I did not make myself clear; I was speculating that one might use some sort of RC network at the input, to contour the response curve of the direct-drive amp so that it would complement the response curve of the ESL. I was not thinking about coupling caps in the conventional sense of the word.
The question is really twofold- do you *really* need to equalize, and if so, what's the best way?
But before that you have to sort out how to do the amp in the first place and not get killed. IMO that's a lot more important :)
so lets assume that the amplifier is safe, and we will not get killed. It has been done; I am unaware that any end-user of Beveridge or Acoustat speakers died during a listening session, at least not by electrocution (perhaps someone died in one form or another of ecstasy while listening to one of those speakers, which is not a bad way to go). I am also aware that we now have a surfeit of lawyers in the world, whioh accounts for the dearth of direct-drive ESLs, moreso than other factors.
Those amps were built into the speaker and caged to prevent shock. I imagine that stuff was a nightmare to ship...
Then the issue is, do you need EQ, and if so, passive or active? Active implies feedback; and I'm thinking that you might be able to tailor the amplifier to have a response that complements the actual curve of the speaker.
I guess you have not been reading my last few posts; that is precisely what I was asking, whether one would need to equalize the response curve of the direct-drive OTL to complement the response curve of the speaker. From the little bit of reading I have done, that would seem to be needed, but I assume it could be done passively. I think there is such a circuit in my Beveridge amplifiers. Well, it doesn't matter much one way or the other; it was just something to talk about.
I was certainly reading them! The thing that really has had my attention since I first considered something like this is the high voltage, and how that relates to different speakers. Seems to me the SL needs more than the Beveridge, and its different enough that I am not convinced that EQ is actually needed either. I don't see it anyway in the SL as it is now, and direct-drive or not it would be a constant.
Its true that the SL uses 2 transformers of different ratios but I don't have enough data to be able to say that that is actually EQ. If so, it would be rather unpredictable as a lot would depend on the source impedance (the amplifier) and we already know that Dr. West does most of his work with a transistor amplifier. And it is true that the SL does act differently with an amp of higher output impedance (better, IMO) so it does not seem as if EQ is what the different ratio thing is about. But in lieu of another explanation, EQ is a good place-keeper for now.
I was also looking at schematics of the old Beveridge amp. I remember doing some mods on one of those years ago. I never did like how that op-amp thing is set up.
The two transformers are definitely some form of EQ.
The bass transformer is wound with very thick wires and when they are played full range to drive the whole Soundlab panel, the natural roll of is around the 500 Hz (my ears tell me). There is no highs what so ever.
The mixer does the rest of the EQ work by bring in another transformer to perform the high frequency work.
The above is quite smartly done by obviously varying the turn ratio to give a balance sound.
Without EQ, direct drive amps straight onto the panels would only yield sound without much bass. Similar to vinyl without RIAA.
Well i have had the Acoustat ESL direct-drive OTL amps ....an sold them..for the tranfourmer type setup....now you all are just talking as if this well be the best thing for your soundLabs....this is where i get off...
You still have a lot of work YOu can do... too get the better sound out of your ESLs....but i gess it ezer to look away from the work to....well dreams
It just one thing to the next with most..It just like doing all the MoDs to a OTL.. till it sofar away from where it was in the stock setup no one knows where you are...or what you have.. not even you ....but it GOT to be better...i have done this also but i STOPet....it means nothing...Well it means you always Lost...not for me..goodluck
Direct-drive amplifiers have never been my cup of tea either! Making the kind of power you need for something like a Sound Lab sounds really scary...
OTLs....ESLs....on paper..look just right..in real life not so much...Amp get Hot as Hell an EAT AC power..ESLs can be Big...there not for eveyone....love to see a 6SN7 criltron preamp setup i could DIy would look like on paper....hint,hint...look it your work so...
ESL are all about there bias...bias bias bias, the way this is setup has a big afect on the sound..One thing i have said over an over is Dr West lets you set you own bias an this is a goodthing... but what dose he have in the speakers.. high V..or Low V tranfourmer ...like 1000v on the secdary side... thin the diodes.. or 300v an how about the Amps of the tranfourmes... i have seen ESL bias Xfourmers that were 1VA....a better one like 1000-1500..Then the doides.. 5-7VA xfourmer sounds way better...An better caps in the bias setup...but most just think well it just bias...better bias can give more output.. an way better base...An sweeter highs i have found...An it a eze fix...goodluck
The Bev is different also from any other present day ESL in the way it uses charge; there is no need for a separate bias supply. The diaphragm is low resistance type. Looking at the amplifier schematic (which I do have), I do agree it looks weird and do wish I could get someone to look at it who could update the circuit and possibly convert it to a balanced input. The equalization on the Bevs was needed to compensate for that acoustic lens through which the output of the speaker has to pass.
The input stage is very complex, to my eyes. But I am not comfortable with SS schematics. I have attached a copy. I have a better one here but it is not digitized. For sure I see there a complex equalizer network.
Hope this is better, so you can see the input, I presume. I have no idea why it comes out so small; the file should be identical to the other in all respects. Part of it includes, I think, a crossover and amplifier for the woofer, all of which I will throw away.
I think you could apply the inverted signal to the base of Q2 using a coupling cap and a series resistor of about 15K. Sure is a lot of EQ in this thing!
Why, I wondered, do they use that series resistor? I had a hunch it is to create a hi-pass filter in combination with a parallel capacitance, but why choose such a hi-value R, when it might be more efficient to use a lo-value R and a correspondingly higher value of C? If you selected a 15K resistor, would you then also halve the value of the R in series with the input of the pos half of the drive signal? (It's 33K, I think. Did not consult the schematic.) Also, wouldn't you have to mess with the ground connections?
Yes, one of those things that the designer is not around to ask about. You shouldn't have to mess with grounding.
Rumors of his death are exaggerated, to paraphrase Mark Twain.
However, I have heard that he is reluctant to comment on this schematic. He sells upgrades to these amplifiers, or so it is said. In an interview with respect to the history of Beveridge speakers and the writing of a Harold Beveridge bio, Mr. Modjeski was reported to say that this was one of his very first designs and that he and Beveridge parted company very soon thereafter.
Perhaps I can draw up what I think you mean, and you can comment.
Lew an others
Maybe we can move back here....for other so if thay run into the SL mod in tiime... thay can get some info.... or what ever ...
To Me it looks like Jim got it right with Acoustat "befor c mod" as for geting the best output out of the high frc. tranfourmers
But i gess Dr. west an jim were facet with people needing to driver these ESL at full output...so this led to the C mod an the setup on the soundLabs.... like thay gave up the best sound for the most sound output..
look can any one give just a drawing of the SL setup
Some of the best sound so far with the Acoustats.. with this setup..the M60s do sound good..
the 60years old pr of Dynaco MK3s...with the Stock 60years old tubes 6550 GEs...an a 6AN8...this is sweet as honey...an it well drive you out of the room...it louad...best highs i have ever got out of any ESL to date...
So were looking at 6 days i had this mod in an i still have to get up in the morning an see if the sound is still there...may be the output tranfourmer on the MK3 match the tranfourmers in the Acoustats interfaces but it work great...Thanks for any input on geting good sound out of OTLs an ESL....goodluck
Lew i think your looking at about 10kv...to get a Amp for the soundLabes.
It not going to happen...Balanced of not...goodluck
Tyu, I don't want a direct-drive amp for my Sound Lab. I am happy with what I've got. I've tried to make that clear a few times.
I posted the schematic for the Beveridge amplifier (+/-3200V) for Ralph's benefit. And yours if you are intellectually curious to look at it. The Beveridge amplifier topology is not at all applicable to driving any other commercially available ESL, because they put signal voltage on the diaphragm, not the stators (or maybe both). The Bev speakers have no separate bias supply per se.
The lens on the Bev is also an EQ feature to boost the bass I believe.
The Bev is like a horn loaded ELS.
There were some scary voltages in there :) Made you numb for an hour or two...
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: