Hi.
I want to drive some homemade ESLs with my Fisher KX-200 which has about 475V on the plates. The ESLs only play from 250Hz up (I use an active crossover). Questions:
- If I connect a dummy load on the output can I just connect the stators to the plates?
- Should I connect the bias supply negative to the amp ground?
Thanks.
Elias
I want to drive some homemade ESLs with my Fisher KX-200 which has about 475V on the plates. The ESLs only play from 250Hz up (I use an active crossover). Questions:
- If I connect a dummy load on the output can I just connect the stators to the plates?
- Should I connect the bias supply negative to the amp ground?
Thanks.
Elias
Well, if you never worked with HV staff, do not do it, period.
As a safety measure you can not connect stator plates directly to the tubes' plates. Use good quality film capacitors of the value 10..20 times of the panel's one, preferably snubber film foil (Wima, Epcos) in series between plates and plates
1000..2000 VDC rated.
In worst case you will be shocked by capacitor discharge but nor killed by dc power supply of the amplifier of yours. Install capacitors inside the safety enclosure i.e. do not fan out wires directly from the plates. Some interlock measures, like blocking power if HV connectors are not engaged would help.
Do not use ceramic HV caps: they are prone to fail short.
You need to organize dc path from plates to the ground 1..2 meagaohms.
Be aware of high AC voltage across dc path resistors. In your case two 2..3 watt rated resistors connected in series will suffice. Bias supply has to be connected to the same ground.
And yes you have to load amplifier's output for stability reasons.
As a safety measure you can not connect stator plates directly to the tubes' plates. Use good quality film capacitors of the value 10..20 times of the panel's one, preferably snubber film foil (Wima, Epcos) in series between plates and plates
1000..2000 VDC rated.In worst case you will be shocked by capacitor discharge but nor killed by dc power supply of the amplifier of yours. Install capacitors inside the safety enclosure i.e. do not fan out wires directly from the plates. Some interlock measures, like blocking power if HV connectors are not engaged would help.
Do not use ceramic HV caps: they are prone to fail short.
You need to organize dc path from plates to the ground 1..2 meagaohms.
Be aware of high AC voltage across dc path resistors. In your case two 2..3 watt rated resistors connected in series will suffice. Bias supply has to be connected to the same ground.
And yes you have to load amplifier's output for stability reasons.
Adding DC blocking will indeed make the circuit somewhat less dangerous, but you can always get killed by the AC voltage driving an ESL when it is playing loud, no matter whether it is driven from an amplifier with or without DC blocking or by a transformer. The only way to make it safe is to put everything into an enclosure that makes it impossible to get too close to the high-voltage stuff.
Anyway, are you sure the output voltage swing is sufficient?
About twelve years ago I built a lethally dangerous amplifier with 2660 V supply and 1100 V output common-mode voltage to directly drive the home-made ESL of a colleague of mine, see
Elektrostatic Loudspeakers
When we tried it, the sound was very good, but the maximum sound pressure level was just sufficient for soft background music. Of course his ESL was full-range.
Anyway, are you sure the output voltage swing is sufficient?
About twelve years ago I built a lethally dangerous amplifier with 2660 V supply and 1100 V output common-mode voltage to directly drive the home-made ESL of a colleague of mine, see
Elektrostatic Loudspeakers
When we tried it, the sound was very good, but the maximum sound pressure level was just sufficient for soft background music. Of course his ESL was full-range.
The Roger Sander's article might be of help. See page 23. Please be safe. http://sanderssoundsystems.com/down...r_1976_the_Sanders_electrostratic_speaker.pdf
oneway................good info............I had not seen this for some time...I still have
Dynaco mk3s...............I have done this setup with the mk3.........Acoustat.panels
but these panels... only had med SPL...............sound was good....
got have high SLP panels......not sure this will work less the panles are. crossover....higher like 250hz. up.....................good luck
Dynaco mk3s...............I have done this setup with the mk3.........Acoustat.panels
but these panels... only had med SPL...............sound was good....
got have high SLP panels......not sure this will work less the panles are. crossover....higher like 250hz. up.....................good luck
Isn't the bias DC? Bypass the caps with 20-megOhm?
I built a Sanders amp - parameters like Marcel's - used it for a few decades. Direct drive is the best sound I ever made. Drove Dayton-Wright cells 140-20kHz fairly loud but not Mahler-symphony loud. 250Hz and up far less demanding of power.
Ben
I built a Sanders amp - parameters like Marcel's - used it for a few decades. Direct drive is the best sound I ever made. Drove Dayton-Wright cells 140-20kHz fairly loud but not Mahler-symphony loud. 250Hz and up far less demanding of power.
Ben
Bypassing the caps with high-value resistors would be slightly safer than a direct connection to the plates.
But, in either case, the effective bias voltage would only be the quiescent plate voltage which the OP stated was 475V.
For reasonable sensitivity, most ESLs need a bias voltage of 5x or more this value.
Thank goodness we are not talking about ESL headphones.
In first post, OP mentions "Should I connect the bias supply negative to the amp ground?". I took that to mean he planned on an additional source of bias. And the answer is the opposite way (so that the two voltages would add together; I don't think bias polarity matters.)
20mOhm - even if I never actually did the math - should be sublethal. But can be tested on a neighbour's pet you don't like (kidding mostly). Unless panels are leaky or you live in Florida, more resistance using HV resistors might work too.
Ben
Oh, ok. I thought your comment meant that you didn't think an additional source of bias was needed at all.
Either bias polarity could be used, but using negative polarity relative to amp ground would be the most efficient use of supply voltages.
(as shown on page 23 of Sanders article linked above)
Don't forget to connect a dummy load resistor to the output of the step-up transformer.
Well, if you never worked with HV staff, do not do it, period.
As a safety measure you can not connect stator plates directly to the tubes' plates. Use good quality film capacitors of the value 10..20 times of the panel's one, preferably snubber film foil (Wima, Epcos) in series between plates and plates
But if the Fisher plates are at 475 VDC, I should be able to use a 600V rated .047 uF cap ( which I have laying around). Correct?
Theoretically you are right, but there are two things to consider:
-There will be some signal voltage drop across the capacitor. The sum of DC and peak signal voltage has to remain below the capacitor's rating, even when the amplifier happens to be playing some subsonic signal.
-If you want good reliability, it is good practice to stay well below the component's ratings.
-There will be some signal voltage drop across the capacitor. The sum of DC and peak signal voltage has to remain below the capacitor's rating, even when the amplifier happens to be playing some subsonic signal.
-If you want good reliability, it is good practice to stay well below the component's ratings.
Well, yesterday I opened the Fisher KX-200, soldered 2 wires to the plates of one of the channels and connected directly to the stators, no capacitors. Fired up and it did sound pretty good, to my ears better than with the transformer I was using to test. But after a few minutes one of the tubes started to glow orange pretty strong. The other was fine. Any ideas why one tube is trying to dissipate more power than the other?
Elias
Elias
Why not just pick up a Shackman amp PCB? I have a populated Shackman ESL Tube amp PCB (for use on >250hz) that uses stacked EL84, with bias at 1800v. I still need to house it and finish, but should be perfect for my CLS.
I have been playing around with my remylared CLS panels with modified Acoustat MK-121 interfaces (just the HF Transformer and different bias supply) and have a bias voltage pot that can go from 650vdc to 5000vdc. The CLS panel starts to sizzle at 2200V so it should be perfect for the Shackman. Just need to clear my projects up so I can get back to that project.
My re-mylared CLS panels require much less bias voltage that the Acoustat panels I have, so perfect for the Shackman amp, and just running at 250hz on greater should allow the panels to be driven to high-ish (hopefully) levels....
I have been playing around with my remylared CLS panels with modified Acoustat MK-121 interfaces (just the HF Transformer and different bias supply) and have a bias voltage pot that can go from 650vdc to 5000vdc. The CLS panel starts to sizzle at 2200V so it should be perfect for the Shackman. Just need to clear my projects up so I can get back to that project.
My re-mylared CLS panels require much less bias voltage that the Acoustat panels I have, so perfect for the Shackman amp, and just running at 250hz on greater should allow the panels to be driven to high-ish (hopefully) levels....
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.