Servo Amps Adjustment?
Hello!
When my dad passed a few years ago he left two working Acoustat servo amps. After sitting in the shelf for 5 years i gave them a good clean and checked for any obvious problems. Didn't find any, so i hooked them up to my Martin Logan CLS MK1. They do play fine at low volumes or Jazz, but when i give them a little more juice i get bass distortion. And it's a bit weird. It's like bass goes aways for some miliseconds and comes back double as loud. Like you would expect from a microphonic tube. Somewhere in the 70Hz and lower region. Hummm...so checked again, swapped tubes, reseated transistors, opamps etc. The problem remains. Then i found notes my father had written down. Mostly what parts went faulty over the years and stuff.
But one note leaves me clueless: "Adjusted stator voltage to lowest possible value: 2,9kV"
Ummm... My father was a DIY guy, building his own ESL panels and he used the Servos to drive them. He was no noob, doing things like that his whole life... But what the heck? As far as i know only "clipping" can be adjusted via the pots on the PCB. From what i believe that is DC-offset, right? Membrane bias is fixed, anode voltage is fixed. All parts have the correct values like stated in the schematics.
Anyone knows what my dad has adjusted there? Is it possible to lower stator output voltage with the pots on the PCB? If so, could it be the reason for the CLS having distorted bass (given they devices are compatible)? From what i understand the dynamic stator amplitude controls possible volume, similar to the diaphragm voltage.
Thanks for your help!
Hello!
When my dad passed a few years ago he left two working Acoustat servo amps. After sitting in the shelf for 5 years i gave them a good clean and checked for any obvious problems. Didn't find any, so i hooked them up to my Martin Logan CLS MK1. They do play fine at low volumes or Jazz, but when i give them a little more juice i get bass distortion. And it's a bit weird. It's like bass goes aways for some miliseconds and comes back double as loud. Like you would expect from a microphonic tube. Somewhere in the 70Hz and lower region. Hummm...so checked again, swapped tubes, reseated transistors, opamps etc. The problem remains. Then i found notes my father had written down. Mostly what parts went faulty over the years and stuff.
But one note leaves me clueless: "Adjusted stator voltage to lowest possible value: 2,9kV"
Ummm... My father was a DIY guy, building his own ESL panels and he used the Servos to drive them. He was no noob, doing things like that his whole life... But what the heck? As far as i know only "clipping" can be adjusted via the pots on the PCB. From what i believe that is DC-offset, right? Membrane bias is fixed, anode voltage is fixed. All parts have the correct values like stated in the schematics.
Anyone knows what my dad has adjusted there? Is it possible to lower stator output voltage with the pots on the PCB? If so, could it be the reason for the CLS having distorted bass (given they devices are compatible)? From what i understand the dynamic stator amplitude controls possible volume, similar to the diaphragm voltage.
Thanks for your help!
Attachments
From what I remember, the Acoustat Bias was around 5kv, while the CLS were under 5kv. more like 3kv...the original CLS was a higher 5kv, but was reduced later versions..
I think your bass distortions ware due to something other than the Bias, unless you are leaking bias, and never fully charging up...
I think your bass distortions ware due to something other than the Bias, unless you are leaking bias, and never fully charging up...
This is correct. I have the original CLS, which run on 4.5kV bias on the earliest versions and 3.8kV on the later serial numbers. AFAIK ML changed bias for giving the panels a longer lifespan.
5kV should be fine, in my experience that won't make a big difference. A bit more SPL. When bias is way too high, sound will get harsh, or -like you also say- the panel will discharge, maybe even pop, crackle, arc etc... But i don't have such problems. Sound itself is silky smooth. It's only that bass breaks away at higher excursions and comes back double as loud like a resonance.
5kV should be fine, in my experience that won't make a big difference. A bit more SPL. When bias is way too high, sound will get harsh, or -like you also say- the panel will discharge, maybe even pop, crackle, arc etc... But i don't have such problems. Sound itself is silky smooth. It's only that bass breaks away at higher excursions and comes back double as loud like a resonance.
With too high Bias voltage, you will hear random buzz, crackles, etc and may get a diaphragm collapse to stator...none are good. Sometime no noises, but on heavy bass notes (large Diaphragm excursions), these issues will arise...so listen, then have Bias as high as possible without these issues is always best...just my 2 cents...
When I played with my CLS2 panels, I noticed all these issues when I raised Bias voltage over 3kv...
When I played with my CLS2 panels, I noticed all these issues when I raised Bias voltage over 3kv...
It just came to my mind that ML uses a "tone shaping" circuit that equalized the signal before going to the transformer and then to the stators. I'm not sure what it does, but i'm trying to figure it out. Could be that it does limit bass excursion. And i'm running the stators directly from the servo charged amp now, what could be too much power for the bass panels. Just guesswork, but a starting point.
To the best of my recollection, the bias voltage in the servocharge ampilier is derived from the output tube plate supply and is therefore not adjustable on the amp. You might be able to rig up some sort of high-impedance voltage divider using the existing 500 Mohm outfeed resistor.
As far as i understand the schematics that is correct. Maybe a separate (adjustable) bias supply is my best option.
Do maybe you have an idea why my father noted that he adjusted stator voltage (what actually is nothing else than B+, right?) to lowest possible value? I mean, i just guess he didn't understand the meaning of the PCB pots. But he was most often smarter than i thought... in retrospect
Do maybe you have an idea why my father noted that he adjusted stator voltage (what actually is nothing else than B+, right?) to lowest possible value? I mean, i just guess he didn't understand the meaning of the PCB pots. But he was most often smarter than i thought... in retrospect
Sorry to bother again. I made some measurements with my UMIK1 on the CLS Martin Logan interface vs the servo charged Acoustat amps.
It's just a theory, but i think that ML's tone shaper in the electronic box is working around an resonance issue the CLS bass panels have. CLS owners know about the lean bass... However, it seems that Acoustat did the exact opposite with their EQ section in the servo amps.
With the help of a DSP i could mimik the frequency response of the ML interface while running the servo amps. And now it gets a bit ridiculous. I needed a first order filter from 90Hz downwards to roll off the bass plus an additional notch filter with quite high Q at 50Hz. Then the servo amp sounds like the original ML interface.
So i assume that Acoustat added a bit of bass fun on purpose here. Something the ML Kindergarten panel just can't cope with.
I will investigate further tomorrow and have a look at that EQ input circuit. I may be wrong with my assumption. But maybe i can bridge it and everything's will be as it should be.
It's just a theory, but i think that ML's tone shaper in the electronic box is working around an resonance issue the CLS bass panels have. CLS owners know about the lean bass... However, it seems that Acoustat did the exact opposite with their EQ section in the servo amps.
With the help of a DSP i could mimik the frequency response of the ML interface while running the servo amps. And now it gets a bit ridiculous. I needed a first order filter from 90Hz downwards to roll off the bass plus an additional notch filter with quite high Q at 50Hz. Then the servo amp sounds like the original ML interface.
So i assume that Acoustat added a bit of bass fun on purpose here. Something the ML Kindergarten panel just can't cope with.
I will investigate further tomorrow and have a look at that EQ input circuit. I may be wrong with my assumption. But maybe i can bridge it and everything's will be as it should be.
For the tiny bit of current keeping Bias voltage on diaphragm, I would leave on 24/7. Collects a bit more dust when charged, so if your worried, put a dust cover - a garbage bag or pillow case - over them when not in use...
If you were to try to drive panels with low, or zero charge on diaphragm, you would burn up your amp and / or the step up transformer in speaker...why risk it?
I remember selling a pair of nice Martin Logan CLS speakers to a guy a while back...he sent an email months later saying he did not need his CLS powered, they sound great disconnected from mains...i again reminded him of the concequences...never heard from him again...hopefuly he finally understood the importance...
If you were to try to drive panels with low, or zero charge on diaphragm, you would burn up your amp and / or the step up transformer in speaker...why risk it?
I remember selling a pair of nice Martin Logan CLS speakers to a guy a while back...he sent an email months later saying he did not need his CLS powered, they sound great disconnected from mains...i again reminded him of the concequences...never heard from him again...hopefuly he finally understood the importance...
Is the C-Mod necessary? No, the speaker works fine without it. But it's cheap to execute, and most people have reported very positive results, with lower distortion in the upper frequencies. I say go for it.
In regards to leaving the speakers powered at all times, the answer is an absolute YES! As noted above, damage can occur if the speaker has no bias and is driven hard (as in you forgot to turn on the AC power and you turn up the volume in an attempt to compensate for the lack of bias.) Power consumption is minimal. The only time I unplug my speakers is if I am away for a week or more.
In regards to leaving the speakers powered at all times, the answer is an absolute YES! As noted above, damage can occur if the speaker has no bias and is driven hard (as in you forgot to turn on the AC power and you turn up the volume in an attempt to compensate for the lack of bias.) Power consumption is minimal. The only time I unplug my speakers is if I am away for a week or more.
That's certainly your choice to make, either now or in the future. But keep in mind that ALL speakers produce distortion, typically way more than the distortion produced by our electronics. Although the speaker may be "close to perfect" now, it can always be improved. Another thing to consider is that the C-Mod gets rid of that nasty 220-uF electrolytic in the signal path and replaces it with a 47-uF. This smaller value is then practical to upgrade to a film capacitor (like polypropylene), which is not practical for the 220-uF due to size and cost. It's hard to go wrong with the C-Mod, but if you don't like the results, it can always be reversed.
Do you think there a significant sonic difference between the acoustat models, when paired with multiple dipole servo subwoofers? I've been waiting for a pair of 6 or 8's to become available, but if the only real difference is base quantity, do you all think a pair of 1+1's work just as well (or even better)? Thanks!