Electrostatic Headphone Amplifier Review & Comparison

[LearningToSmile]

I almost bought the Koss ESP/95X at one point when it was on large discount due to closing of Drop's EU warehouse - I'm glad I didn't. Electrostatic headphones are still something I'd like to try in the future, but I don't think I'll be buying any blind.

 

[Helicopter]

How old are these? Looks like mains leakage in the 006 (old capacitors?) is adding a lot of noise. Have you replaced the tubes recently? Wonder if new tubes and capacitors or whatever is needed to make the power supply perform better would improve these measurements?

He stated early 2000s.

 

[Vini darko]

Eeek fur gonna fly in electrostat land today. Thanks for the measurments. I agree with previous posts about input sensitivity and load resistance. Be good to check for thoroughness.

 

[o7_brother]

Shame that the analyzer doesn't go higher than 100 000 ohm. A better number would be something like 1.7 million ohm, because that load would more closely represent the midrange frequencies of a Stax headphone.

Then again, interesting results nonetheless. I knew those damned Stax tube amps were terrible.

Would be cool to see how those "mafia" amps perform, like the BHSE and KGSSHV Carbon. Having compared one of the better Stax amps (SRM-717) and Carbon in a pretty flawed single-blind test... you can guess what I thought of the claims of increased soundstage, dynamics, black background, detail, etc.

 

[the_brunx]

WOW. I always kinda knew electrostatic amps are a bottle neck of the electrostatic headphones. We need to find one which measures Great with lots of power. ASAP. Thanks for this revolutionary post.

 

[kevin gilmore]

So much fail here.

100kohm load is way to low to measure these amplifiers. And in fact without 120pf across the output, the results are meaningless.

You need a keithely 2015 high voltage thd meter that does more than 500vrms with a 10 meg load resistance. And you still need the
120pf caps to get a real idea of what the high frequency response looks like. hint: its down more than 10db at 20khz at 200vrms

The koss is lighting the current error light because of the 100k load. With a 10 meg scope probe its more than happy to do 2400 volts peak to
peak balanced, or 1200 volts peak to peak measured with respect to ground. At 1khz with 120pf load.

srm-006t and srm-007t are identical schematics. srm-007t uses twice as many tubes with 2 triode sections in parallel in the output stage.
These and all the other stax tube amps are solid state front ends. Only exception are the t8000 and T2 with use 6dj8 input stages.

 

[phoenixsong]

T8000 when?

 

[kevin gilmore]

not paying attention?? more than 1000 x T8000 have been sold. way more. 6dj8 front end, srm727/srm700s solid state output stage.

 

[phoenixsong]

not paying attention?? more than 1000 x T8000 have been sold. way more. 6dj8 front end, srm727/srm700s solid state output stage.

Sorry, I don't understand

 

[kevin gilmore]

warning to those who think its simple to just add a high resistance to the input of an audio-precision or dscope etc.
bad things will happen (meaning expensive to repair) especially the audio-precision which has active transformer based inputs.
you need a pair of 10x tektronix active fet probes and associated power supplies to do this and then you still have to be careful
because those probes are only good to 400vrms.

All stax amplifiers whether solid state or tube have current limited output stages. Around 6 to 7 ma, 15ma T2.
The koss amplifier is actually push pull and the only current limit is that of the power supply.

 

[phoenixsong]

warning to those who think its simple to just add a high resistance to the input of an audio-precision or dscope etc.
bad things will happen (meaning expensive to repair) especially the audio-precision which has active transformer based inputs.
you need a pair of 10x tektronix active fet probes and associated power supplies to do this and then you still have to be careful
because those probes are only good to 400vrms.

I see, so that's what you were talking about! I was referring to measurements of the T8000, but if it entails risking the destruction of Amirm's Audio Precision then maybe I'd pass on it LOL

What would you suggest to safely measure these devices then?

 

[kevin gilmore]

t8000 measures virtually identical to srm700S or feedback modified srm727

you want to measure safely, then one of these
https://www.ebay.com/itm/Keithley-2...874448?hash=item3b258a0bd0:g:MQ8AAOSwXc9cGvoJ
and one of these
https://www.ebay.com/itm/Keysight-A...277024?hash=item2cbf0415e0:g:xAQAAOSwRAVfMPdS
and the james m king software and a reasonably recent pc with windows 10 and the java development kit and the gpib libraries etc.

a little less safely a pair of these and audio-precision or dscope etc.
https://www.ebay.com/itm/Tektronix-...253759&hash=item4465a38d6f:g:HmAAAOSwjGxfmr0j
and associated power supplies.

 

[DuxServit]

Welcome Mr Gilmore, great to see you’ve joined the discussion. Your Stax expertise is needed here.

 

[phoenixsong]

These look really interesting. it's a pity I have near-zero knowledge of electrostatics, having only heard the SR009s and T8000 combo once in an audio store as my sole electrostatic experience ^^"

 

[kevin gilmore]

picture

 

[Veri]

picture

It's interesting having you chime in, Kevin. All we want is good measurements of these electrostat amps

 

[gvl]

Do you have an electric blanket?

I don't, because you never know when enuresis might develop. Even if I had it, it's 120V here, just tickles

 

[whero]

@amirm couldn't you just add a ~ 2m ohm resistor in series to your measuring rig? also I'm wondering how it'd stack up against a trafo setup (srd7mk2 / woo wee) / aftermarket amp (maybe spritzer from mjolnir could lend one of his creations to be measured?)
would stack against? (and in case of a trafo setup how much the measurements would be affected by the transformer and how much by the amplifier)

 

[o7_brother]

couldn't you just add a ~ 2m ohm resistor in series to your measuring rig?

Apparently not:

warning to those who think its simple to just add a high resistance to the input of an audio-precision or dscope etc.
bad things will happen (meaning expensive to repair) especially the audio-precision which has active transformer based inputs.
you need a pair of 10x tektronix active fet probes and associated power supplies to do this and then you still have to be careful
because those probes are only good to 400vrms.

 

[Xulonn]

As many of you know, I do not use headphones. Aside from intellectual curiosity or a desire to be different and non-mainstream, what are the purported sonic advantages of electrostatic headphones vs. conventional headphones? Are they more accurate? Or are they superbly euphonic?

I have been a fan of ribbon and AMT tweeters in the past, but I know enough about human and consumer psychology to be aware that my preference for them is likely not strictly based on sonic qualities. I certainly do not promote or defend ribbons or AMTs as providing sonic nirvana.