Acoustat Model 3 Questions

[bolserst]

OK, I removed the additional diode/cap, so we are back to stock.With 100VAC from bias transformer...
With C8 only at 105VDC, is D1 dead?

Your measurements all look pretty darn close to expected values.
I would say your multiplier is in good shape.
Remember, nothing wrong with adding another section or two if your panels can handle that level of HV without crackling or diaphragm collapsing to one of the stators.

The first cap in the multiplier would only charge to 140VDC, not 280VDC like the rest. So, 105VDC is 75% of the maximum possible and is pretty much spot expected results for a 10Mohm load.

It just dawned on me that you could actually use the 500Mohm resistor in the Acoustat HV supply to rig up a temporary HV probe with 510Mohm input impedance. This would get you a lot closer to the actual unloaded HV.

Take a look at the attached image.
Disconnect all the diaphragm connection wires for the ESL panels from the HV supply, and connect the (+) probe from your meter to where the ESL panels had been connected. Then, disconnect the other end of the 500Mohm resistor from the HV supply. This would involve temporarily unsoldering its connection to the PC board. This end of the 500Mohm resistor would now become the (+) input for your HV probe. The (-) input is the (-) probe of your meter.

So, connect the (+) & (-) HV probe inputs to whatever you want to measure, for example, voltage across each capacitor. With 510Mohm input impedance, current draw is a lot less and you will see much closer to the expected values. Remember to multiply your reading on the meter by 51 to get the actual voltage.

You could even crank the Variac up so the output of the HV transformer was 750VAC since the meter will only see 1/51 of the applied voltage. So, if you were measuring the output of the multiplier and it was 5kV, the meter would only read about 98VDC.

 

[john65b]

I am a little confused... with 100VAC input on the voltage multiplier ladder,why do I have to correct for the 10M ohm loading of the DMM? I can see with very HV, but this is 100VAC.

If can properly read voltages on Tube Amp B+ of 500V with this DMM without any correction, why would I need to do so for these readings? Is it due to the extremely low currents we are dealing with?

Shouldn't my end of the bias ladder at end up 1/7.5 (with 100VAC) of the 5000V, which is 666VDC?

AAARRGG!!!!

Please don't take this as a challenge to your help, but I am confused...as I am a mechanical engineer, not a EE (probably a good thing).

Also, I have put one of your sugested Neon Lamps (borrowed from Quad ESL) after the 500M Ohm resistor, and before the diaphraphm. One of the interfaces neon lamps is always lit very very dim, with an occasional flash every two to five seconds...could this mean a leak?

 

[bolserst]

I am a little confused... with 100VAC input on the voltage multiplier ladder,why do I have to correct for the 10M ohm loading of the DMM? I can see with very HV, but this is 100VAC.

If can properly read voltages on Tube Amp B+ of 500V with this DMM without any correction, why would I need to do so for these readings?

Voltage level has nothing to do with it.
It has to do with the ratio of your load impedance to the size of your power supply capacitors.

In the case of your tube amp, the power supply capacitors are probably 50uF to 100uF perhaps?

The HV ladder network capactors are only 3300pF, or about 15,000 to 30,000 times smaller storage capability.
So, the 10Mohm load on them would be like a 300 - 600 ohm load on the 500V tube power supply. That would make the tube amp supply sag just like you are seeing with the HV supply.

 

[bolserst]

Also, I have put one of your sugested Neon Lamps (borrowed from Quad ESL) after the 500M Ohm resistor, and before the diaphraphm. One of the interfaces neon lamps is always lit very very dim, with an occasional flash every two to five seconds...could this mean a leak?

oops, missed your other question.
I actually haven't seen this behavior before, has anybody else?

My first guess would be that one of the diodes is starting to go, allowing considerable AC ripple at the output and causing the dim glow. I don't think it is a leaky panel as that would result in rapid flashing. Will have to think on this a bit....

Is this the supply you measured the voltages on?

 

[tyu]

Remember, nothing wrong with adding another section or two if your panels can handle that level of HV without crackling or diaphragm collapsing to one of the stators.

Yes it eze.....works great....look i have MartinLongan SL3s panles..that run at 3500 when i was working on my new bias that i can set from 4k-6 i would run the logen panels up to about 4500...sound great.... at just hear a little bzz ear right up to the panels...best the sl3 ever sounded..The Acousta are way diff...made to run fulrang....i have ran the Acoustat panels at 6k...no noise...but the topend get a little peky so i back of a little....but my bias is in it own box...it cant get pulled down like in the 121 interface box with the high low mixer 1"away with there on high V.that driving the panels.....Look at this pic....Dose anyone think this is a good way to setup a 5k bias???...I well say it works but come on....5k all over the box it in....It has to sound better out with it tranfourmer.away from the mixer....thanks for the info good post

 

[john65b]

Bolserst, yes, again I stand corrected. I like learning these things by trial and error.

Yes, same panel - the neon lamp is so very dim, but I can make it out. It is the same interface / panel that would send my amps into a pulsing quivering mess - only the Krell KSA-50 Clone and the Icepower 1000ASP amp would work - even the Aleph 5 clone and Acoustat TNT200 would have issues with this speaker.

The only thing I could do to make it play nice was was to change the 1 ohm resistor to a 2 ohm resistor on the LF tranny. But I wonder if its something else with the bias...I was going to return the 1 ohm resistor once I get the real problem nailed down...I think it sounded better with the 1 ohm resistor...

TYU - yes, I think I may change the bias ladder to a simple 15vdc walwart powered DC to HVDC converter, like what I did with the CLS power supply.

The CLS sound so clear, wide (no beaming) and sound perfect to me, but I need to stay above 150HZ or I start to saturate the Toroidal cores. And the Acoustat Model 3 has such impressive bass (crazy for an ESL) . I just don't get that CLS Midrange from the Acoustats. I am thinking of using the two outer Model 3 panels with the CLS in the middle and plugging it all in on the Acoustat interface and connecting from 3 panels to 4 panels on the LF transformer - kind of a Hybrid Acou-Logan or Martin-Stats...the best of both worlds...

 

[bear]

OMG - you have to change out the old multiplier caps if you have not.
Then replace ALL of the HV diodes with new HV diodes.

Then you WILL get the required voltage.

The CLS beams plenty... fyi. It has a rotten interface.
Try the Acoustat interface on the CLS panels for fun. But the CLS does NOT have enough excursion to do bass right in the first place, but if you keep the levels lowish, it should work nicely. Give you more output too, probably. IF you need lower bias voltage, tap down the divider!

Change the caps and HV diodes.

_-_-bear

 

[john65b]

OK, changed out the multiplier caps with 2000pf 10kv ones from Apexjr, but didn't bother with the diodes (they are good to 10kv, so should be ok), and I also added back the single cap diode link, so now should be around 6kv bias. Now it sounds a bit louder and quite a bit cleaner. I think this is it for these speakers. Just perfect now...

I just hauled home a pair of Model 1's with Medalion Trannies and matching sub. One speaker wasn't working, went in and both fuses on the Bias tranny were blown. Now all good. The woofer on sub has nearly all the surronds corroded away. Looks like a nice side project.

Looking to try the Model 1's MK-141 with medalion trannies on my CLS IIa panels (since I am running them at 125hz and above anyway)...this will be interesting...

 

[bolserst]

OK, changed out the multiplier caps with 2000pf 10kv ones from Apexjr, but didn't bother with the diodes

So did changing the multiplier caps fix the dimmly glowing neon bulb issue?
How about the problem you were having with amp stability when driving this particular interface. Fixed?

 

[john65b]

The neon bulb appears fine now, and since changing the 1 ohm resistor with a 2 ohm resistor (on interface input), all amps play nice. I do not know if the 6kv bias would have corrected that issue...kinda liking the way it is now and will leave it be.

 

[tyu]

Yes i went to the 2 ohm last year ...i like the bass better...I Know thay sound way diff with the bias up...
How miny Acoustats are there out there that the owner just think [there old] an just dont have it any more....Hehe...well we can pic them up for a song....... an as you say you can drive the panels ezer...what not to like?...
The BIG SoundLabs on paper look Like there at the top of the ESLs Game... but i have found it like the Apogees any thing biger than the Stage an you give up a lot of the sound because you have to have a amp that over a 100lbs tops make them work....I am driving my Acoustat M3s with a 50watt tube amp... same amp on the Stages....sweet sound .....goodluck

 

[bear]

repeat: replace the diodes.

they die, if they are the stock originals. few interfaces survive to this date without either the caps dying and/or the diodes as well. Not all of the diodes, or all of the caps, but it is best to replace them all and not have to go back and do it again in a short while when the next one dies...

_-_-bear

 

[john65b]

OK, looking at two 1N4007 each leg...good enough??

 

[bear]

no.

you need much higher PIV.

two 4007 per position (in series) is still "iffy".

high voltage diodes only.

_-_-bear

 

[john65b]

Differential across each ladder is only 1000V, correct?

ML CLS interface had a multiplier based on doubled up 1n4007 to 3700V....

 

[tyu]

Funny if we were talkin about caps&coils.. in the crossover .....it would be nothing for someone to spend weeks working puting new caps an coils in there speakers ...An with Maggys people like Me have Re-wired the pr of panels....but when it come to All ESL bias setup....it just to much to ask!...like it dose not matter...I know it HIGH V...but come on were talking about What [CANbe]... the BEST SOUND ON THE PLANT!....there a lot more mods out there...too even get better sound out of the Acoustats....but if the bias is off...you got nothing....Hay there guys out there that sale ajustabel bias setup for esls....$60. ea....from 4-7k.....goodluck

 

[bear]

peak voltages kill the diodes.

use two in series and it may pass, but high voltage diodes give the assurance of greater reliability.

_-_-bear

 

[john65b]

I bought a bag of 50 1N4007 a few months back for like $7.

I could actually string three in series pretty easily and still be quite compact.

 

[bolserst]

I bought a bag of 50 1N4007 a few months back for like $7.
I could actually string three in series pretty easily and still be quite compact.

The PIV (Peak inverse voltage) the diodes have to withstand in a half-wave voltage multipliers is 2 x the peak input voltage. For the Acoustat HV power supply, this is 2 x (750 x 1.414) = 2120 V.

I know that Moray James mentioned he has used 2 of the 1N4007 in series per section and has not had any failures.
http://www.diyaudio.com/forums/plan...-1-high-sound-level-limits-2.html#post2442911

Personally I would use 3 in series, or follow bear's advice and use some higher voltage rated diodes.

 

[AcoustatAnswerMan]

The original Acoustat diodes are rated at 10-kV at 25 mA.

Yes, you can series several lower-voltage diodes to obtain the desired rating, but one can never guarantee that the voltage will divide equally among the string.

I would recommend special high-voltage diodes only.