I know this is possible. Also the only way to put the ss/tube debate to rest. I have a pair of Acoustat Model 3 and a pair of Acoustat Monitor 3, the former driven through interfaces, the latter driven directly by Acoustat's own direct drive servo OTL tube monos. I can attest there is no comparison between the two. I'm assuming the culprit is the interface? Resolution is in a different league with the DD OTLs. With interfaces, they are very comparable to anything. It's clearly a question of how ss sounds driving the panels directly. I've compared with some great ss amps. Conversations I've had with potential builders is "the voltages are too dangerous". Don't be a baby! Don't stick your fingers where they don't belong!
So, any takers?
So, any takers?
There is a serious incidental incentive here, no? It's puzzling that nobody here is curious about how an ESL would sound driven by ss without an interface in the way...and anyone who can compare will tell you it's the last frontier. Here's where the debate ends, and/or a new one begins. Honestly, I don't care about that. But I do want to hear the difference. Budget? Talk to me..design and build.
Not a big challenge at all - swinging a few thousand volts with couple of tens of mA is easier than you might think. 10kW output stages of SW transmitters I used to deal with work at 9.5kV smoothly.
Real concern here is user's safety. Construction must not allow a slightest possibility for user to come in touch with such a high voltage (no speaker binding posts and no speaker cables allowed) and that's not a job for occasional solder-slinger or one might end up badly hurt or worse and the other one in court or worse. I would never build such an amp for a numpty but you'll definitely need a pro and I doubt that there is one who will do it just for the thrills of doing it (and no, that project is not a last frontier, not a looong way)
So, if I were you I'd gather more information/knowledge on the subject and give the whole idea some more thought before motivating someone into disaster...
Real concern here is user's safety. Construction must not allow a slightest possibility for user to come in touch with such a high voltage (no speaker binding posts and no speaker cables allowed) and that's not a job for occasional solder-slinger or one might end up badly hurt or worse and the other one in court or worse. I would never build such an amp for a numpty but you'll definitely need a pro and I doubt that there is one who will do it just for the thrills of doing it (and no, that project is not a last frontier, not a looong way)
So, if I were you I'd gather more information/knowledge on the subject and give the whole idea some more thought before motivating someone into disaster...
Bump. I fail to see why you guys are so fearful. Lethal voltages are present regardless. The dd servo charge OTLs are arranged so that the power chord is integral to the cage which serves as a safety cover...no access without unplugging. A 3-way banana plug directly from the panels into the amp such that it is wedged between amp and panels...nothing to touch, similar to how the interface is connected anyway. You really can't prevent self inflicted harm due to deliberate neglect.
I don't think that lethality or similar considerations are a real obstacle: after all, the good 'ol 230V mains in Europe remains the main killer for would-be victims.
Air purifiers, etc. use much higher voltage but are practically never involved in deadly occurences, except when the mains isolation fails.
Domestic high voltage is a kind of paper tiger, which doesn't mean you can allow yourself to be lax, but in general you have more chances to be killed by the more or less neglected 230V input wiring than by the EHV output voltage
(microwave oven transformers are an exception, because the output is high voltage, 50Hz, and high current, a deadly combination)
However, the chances that someone designs you a tailor-made amplifier for free are slim: the best you can hope for is something more or less fitting already in existence.
In practice, if you use semiconductors, not valves, an amplifier output stage will generally resemble this:
Self-cascoding, a tool for creating ultra high voltage building blocks:
The concept can be made to work, even at video frequencies, but the devil is in the detail, and any error generally ends up in molten silicon
It might be an opportunity to apply the ground-centric amplifier concept, as it halves the voltage seen by some active elements:
Ground-centric VAS?
Air purifiers, etc. use much higher voltage but are practically never involved in deadly occurences, except when the mains isolation fails.
Domestic high voltage is a kind of paper tiger, which doesn't mean you can allow yourself to be lax, but in general you have more chances to be killed by the more or less neglected 230V input wiring than by the EHV output voltage
(microwave oven transformers are an exception, because the output is high voltage, 50Hz, and high current, a deadly combination)
However, the chances that someone designs you a tailor-made amplifier for free are slim: the best you can hope for is something more or less fitting already in existence.
In practice, if you use semiconductors, not valves, an amplifier output stage will generally resemble this:
Self-cascoding, a tool for creating ultra high voltage building blocks:
The concept can be made to work, even at video frequencies, but the devil is in the detail, and any error generally ends up in molten silicon
It might be an opportunity to apply the ground-centric amplifier concept, as it halves the voltage seen by some active elements:
Ground-centric VAS?
I'm not looking for a free lunch but there is an implied incentive. I think it's a matter of ignorance due to the rarity of the dd servo charge amps and hence commensurate experience/lack thereof. One listen is all it takes.
The dd servo charge amps would work with any ESL, as would it's ss counterpart. I think Beveridge made a dd servo OTL for ESLs?
The dd servo charge amps would work with any ESL, as would it's ss counterpart. I think Beveridge made a dd servo OTL for ESLs?
Last edited:
Or you can Circlo this guy:
http://ixapps.ixys.com/DataSheet/DS100458B(IXTL2N450).pdf
or some of his buddies:
4500 V Power MOSFETs - IXYS | DigiKey
Hi,
I have several Acoustat models and also the servo amps. The servo amps do sound better.
But there is also quite a difference between the DD amps and transformers and it is probably not the high voltage tube output stage making all the difference.
The DD amps have 2 adjustable EQ's, HF balance and the wall effect and I can't believe that the signal from 2 overlapping range transformers blended together would have exactly the same EQ curve as the amp. In the upper bass/lower mid range just a slight hump or dip could easily appear to add or mask some detail.
I never measured the signal at the speaker terminals but that might give some insight to the sound differences.
Dave
I have several Acoustat models and also the servo amps. The servo amps do sound better.
But there is also quite a difference between the DD amps and transformers and it is probably not the high voltage tube output stage making all the difference.
The DD amps have 2 adjustable EQ's, HF balance and the wall effect and I can't believe that the signal from 2 overlapping range transformers blended together would have exactly the same EQ curve as the amp. In the upper bass/lower mid range just a slight hump or dip could easily appear to add or mask some detail.
I never measured the signal at the speaker terminals but that might give some insight to the sound differences.
Dave
What you are asking for is not trivial. A commercial linear 5kV amp for electrostatic drive of instruments etc is quite pricey.
TREK Model PD06087 High-Voltage Power Amplifier
Since current is low, you might consider a simple
SE Class A device with as few transistors as possible and capacitive output coupling. I am thinking a HV mosfet linked earlier might work very well. Of course it needs a HV driver circuit.
TREK Model PD06087 High-Voltage Power Amplifier
Since current is low, you might consider a simple
SE Class A device with as few transistors as possible and capacitive output coupling. I am thinking a HV mosfet linked earlier might work very well. Of course it needs a HV driver circuit.
Hi,
don't believe that You can get away with just a couple of mA to drive a big panel.
Rather think in several hundreds of mAs.
The more so, the more full power bandwidth You want to achieve.
Typically the so far known candidates don't achieve more than ~5kHz.
The esl is also a almost pure complex load, demanding mostly reactive current.
Then account for the low efficiency of max. 12.5% or 25% (depending on anode load) of a SE class A.
Modulated current sources as anode load (making it a SEPP circuit) could achieve higher efficiency (50% in theory) but can be difficult to stabilize with a complex load.
I don't think, that it is as easy as Juma claims to build a esl dd amp.
jauu
Calvin
don't believe that You can get away with just a couple of mA to drive a big panel.
Rather think in several hundreds of mAs.
The more so, the more full power bandwidth You want to achieve.
Typically the so far known candidates don't achieve more than ~5kHz.
The esl is also a almost pure complex load, demanding mostly reactive current.
Then account for the low efficiency of max. 12.5% or 25% (depending on anode load) of a SE class A.
Modulated current sources as anode load (making it a SEPP circuit) could achieve higher efficiency (50% in theory) but can be difficult to stabilize with a complex load.
I don't think, that it is as easy as Juma claims to build a esl dd amp.
jauu
Calvin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.