[kevin gilmore]

1) except for the esp950 and TOGTE, none of the rest of the electrostatic amplifiers are push pull.
So ultimately the source or sink current in the output stage is the limit before the amplifier has to clip.
Push pull drivers usually can go rail to rail, and may drop out of class A when doing so.
 
2) The output stage power is strictly the voltage across the output stage times the output stage bias.
So for example at +/-400 volts and 18ma, each of the 2 drivers consumes 14.4 watts per channel.
Plus the filament power for tube amps. Which is also significant depending on which tubes and
how many of them.
 
3) The slew rate is hard to calculate, Tube amps have built in compensation (the capacitance of the grids)
which limits how fast they can slew. Grounded grid is a way around this, but requires a fair amount of
drive power.  For almost all solid state amps, some external compensation is required because otherwise
they tend to oscillate. This also limits the slew rate. Miller effect and the resulting zeros in the right half plane
dictate this, whether tubes or solid state. Accurate spice models can predict reasonably well the actual
slew rates.
 
This is why many of the electrostatic circuits begin to look the same.

 

[spritzer]

One other thing to consider is how the output stage handles the given load.  I'm running the B-10 at a little over 11mA (the absolute max the tubes can take) so power roughly on par with the KGSS but the output stage design just can't handle the reactive load so it sounds like crap and measures even worse.  Other points include output impedance and output capacitance though mostly the latter.  If that is too high then it will do just the same damage to the sound as a tube amp with excessive output impedance driving a 8ohm speaker. 

 

[Tilpo]

I'm wondering whether it's true for all electrostats, but my SR-202 seems to play bass in an odd way. Very short low pitched tone bursts that would normally cause a thump do not cause a thump on the SR-202, while it does on my HD650. However, the bass does extend very deep. Even with very low bass tones there really does not seem to be any loss in level.
Is it just me, or is there an actual reason for this phenomenon?

 

[sachu]

Quote:

One other thing to consider is how the output stage handles the given load.  I'm running the B-10 at a little over 11mA (the absolute max the tubes can take) so power roughly on par with the KGSS but the output stage design just can't handle the reactive load so it sounds like crap and measures even worse.  Other points include output impedance and output capacitance though mostly the latter.  If that is too high then it will do just the same damage to the sound as a tube amp with excessive output impedance driving a 8ohm speaker. 

 
what the hell is a B-10?

 

[adktitan]

Quote:

 
what the hell is a B-10?

This could get to be some interesting reading....

 

[Radio_head]

 
what the hell is a B-10?

Link in the shoutbox. 
 
Let's keep this stuff where it already is.  No need for the inevitable future thread locks, post deletion, the usual warnings, etc.

 

[purrin]

huh, what are you guys talking about?
 

 

[arnaud]

Kevin, Birgir, thanks again for the basics, I almost feel like I am getting it (big warning sign). Thanks alex too for confirming the spec of the LL.

What I was trying to get at is a simplified table with the very least amount of information required to compare various amp spec.

From this discussion, it would seem like we need at least:
1. Maximum Voltage swing (PP stator to stator which is close to 4x the rails voltage)
2. Output stage bias current (assuming a class A operation)
3. Slew rate at 20kHz
4. Output impedance / output capacitance (similar to the damping factor into 8 ohms resistive load for an electro-dynamic amp?)
5. Total dissipated power (to get a feel for how much power consumption and heat issues)

It may start to sound like a pissing contest but wouldn't such basic info help anyone get an idea of how an amp can potentially perform?

In regards to the Electra, 1800V (PP/SS) and 40W of "plate power per set of 6L6 tubes"). I assume it's 500V rail (some is lost due to the tubes / circuit so you get less than 2000V swing) and because post mentionned 40mA into each tube. Now does that plate power really translate into 40mA available on the output?

 

[KingStyles]

Just curious, I saw these on ebay. I didnt know sony ever made a electrostatic headphone. Anybody ever try one of these.

 

[milosz]

Yeah I heard a pair of these back in the day (late 1970's.) All I remember is they sounded good. Not good as compared to Stax or compared to other headphones, I heard them at CES, in comparison to nothing.  I listened for about a minute and they sounded good to me.  At the time I owned Koss ESP-9's.  See http://wiki.faust3d.com/wiki/index.php?title=Sony_ECR-500
 
Sony also made some full-range electrostatic speakers, but I've never heard these.  http://www.thevintageknob.org/sony-SS-R10.html
 
Sony has made some interesting products over the years, Sony has made some very good products over the years, and Sony has made some junk over the years.

 

[spritzer]

Quote:

Kevin, Birgir, thanks again for the basics, I almost feel like I am getting it (big warning sign). Thanks alex too for confirming the spec of the LL.
What I was trying to get at is a simplified table with the very least amount of information required to compare various amp spec.
From this discussion, it would seem like we need at least:
1. Maximum Voltage swing (PP stator to stator which is close to 4x the rails voltage)
2. Output stage bias current (assuming a class A operation)
3. Slew rate at 20kHz
4. Output impedance / output capacitance (similar to the damping factor into 8 ohms resistive load for an electro-dynamic amp?)
5. Total dissipated power (to get a feel for how much power consumption and heat issues)
It may start to sound like a pissing contest but wouldn't such basic info help anyone get an idea of how an amp can potentially perform?
In regards to the Electra, 1800V (PP/SS) and 40W of "plate power per set of 6L6 tubes"). I assume it's 500V rail (some is lost due to the tubes / circuit so you get less than 2000V swing) and because post mentionned 40mA into each tube. Now does that plate power really translate into 40mA available on the output?

 
It would certainly give an indication but we'd also need distortion specs into a fixed load at full output to get a clearer picture.  The bigger issue though is that manufactures do not want the buyers to make an informed decision but much rather rely on forum hype, mythical BS and other such things. 

 

[Solude]

^
Truth

 

[milosz]

I think that some of the smaller "manufacturers" aren't equipped to make these measurements.  Might not have anything beyond a DVM and a 'scope, though they SHOULD.  I bought an HP distortion bridge for $85 on eBay.  Got a low distortion sine generator there too.
 
Stax does list output THD, frequency response and "[size=small]Maximum output voltage[/size]" - I assume this means voltage swing.  Slew rate is not listed.
 
WOO says the WES is capable of 1300 volts p-p swing; no THD or slew rates given.   WOO says their GES has less than 0.1 % THD, but doesn't list voltage swing or slew rate. 
 
Cavalli lists no specs for their Liquid Fire
 
Ray Samuels Audio does not list any specs for their A-10
 
HeadAmp lists the Blue Hawaii as capable of 1600 volts p-p swing; no distortion specs or slew rate numbers given.
 
No specs for the Koss amp output, distortion, slew rate etc.

 

[seacard]

I am generally a numbers guy, but even if we had all these numbers, would we be able to tell much about the amp's sound? For example, on mwilson's Liquid Lightning, things like voltage swing, output stage bias current, slew rate, etc. wouldn't change from 100 hours of burn-in to 130 hours of burn-in, and yet whatever component is responsible for soundstage and instrument separation has apparently changed significantly in performance somewhere in that 30-hour range so that "soundstage has started to widen, instrument separation is getting better." If an amp can indeed make such a noticable transformation late in the burn-in process, then there must be something else going on here.

 

[spritzer]

I take any remarks like that with a healthy dose of skepticism.  I've built and rebuilt enough equipment to to know there is no such thing as burn in and nobody has been able to prove its real.  If component is changing its value after a few hundred hours then it is by definition broken and should be replaced.  Let's look at real possibilities then:
 
One cause can be due to thermal drift.  Going by the numbers Alex has posted then the Liquid Lightning must be burning hot even at low ambient temp and that can cause all sorts of issues.  80W is a lot of heat to dissipate on those tiny sinks and if the amp was setup with the top off then it can drift even further with it on. 
 
Another factor which ties into the heat issue is the actual mains voltage in the wall.  Up here in the north it never changes more than 1V but I guess it's getting quite warm in Florida and the load on the AC line must fluctuate quite a bit.  Even though the amp has a regulated PSU the wall voltage can still change how it sounds and even knock it out of regulation in extreme cases. 
 
Then we have the whole problem with space charge in the drivers and issues related to the bias supplies.  These can change the sound from hour to hour, day to day depending on humidity and other environmental factors.  I know enough people had issues with the Exstata bias supplies so if the LL uses anything similar then that could apply.