Electrostatic amplifiers - power requirements

[John_M]

Amir says he's going to test some electrostatic amps at some point. Will be interesting to see what he tests and what comes out it (and, Kevin, you might have some comments on whatever methodology he uses...)

https://www.audiosciencereview.com/...o-equipment-for-free.1995/page-12#post-435216

 

[kevin gilmore]

keithely 2015 high voltage thd meter and tektronix high voltage differential probes available on ebay.

the keithely is a great place to start because it has a balanced signal source and is measures up to 1kv

 

[T3RIAD]

to drive the headphones at 0db reference 1khz at 20khz and maintaining 6db of feedback requires...

at 450v power supplies, 18ma
at 400v power supplies, 16ma
at 350v power supplies, 14ma (most stax amps run at this voltage, but none of them come close at 5 to 6ma typical)

Two questions:

1. Does "0 dB reference" here mean the maximum SPL rating of the headphones (118 dB for the SR-009S)?

2. What would be the current requirement for lower volumes, like 100 dB peak?

I think the main question posed in this thread is whether upgrading from a Stax amp to a Spritzer amp can improve sound quality at low volume levels, in addition to increasing the amount of power available. No one disputes that a Carbon can play louder, but if you don't listen loud, is it still worth it?

 

[kevin gilmore]

1) 0 db reference is the peak to peak voltage the amplifier delivers at 1khz before clipping. has nothing to do with spl

2) this would require knowing the efficiency of a particular headphone. which varies with temperature and humidity among other factors.

 

[T3RIAD]

Thanks for the responses.

Would it be correct to extend your chart and say that 6 mA can swing around 150 V?

 

[kevin gilmore]

So the Stax D10 has +/-175v power supplys and a 4ma current limit. Every person i know of that has one say the same thing,
no where near enough volume.

Its more than that. A Carbon is much more like the T2 sound than the 727 sound. Even at the same low volume level they do not
sound the same.

 

[T3RIAD]

So the Stax D10 has +/-175v power supplys and a 4ma current limit. Every person i know of that has one say the same thing,
no where near enough volume.

Its more than that. A Carbon is much more like the T2 sound than the 727 sound. Even at the same low volume level they do not
sound the same.

Not everyone. This person says a D10 sounds the same as an SRM-353x.

 

[T3RIAD]

So I did some more research and, as it turns out, we cannot hear distortion in frequencies higher than 5 -10 kHz, because the harmonics are outside our range of hearing. Try playing an 8 kHz sine wave, then an 8 kHz square wave on this site. They will sound identical. The only difference between them are distortion products that you cannot hear.

Thus, I believe the only result of limited slew rate would be a slight roll-off in the higher frequencies when the amp is asked to play them at high volume. The distortion products added would be inaudible, but the volume would be reduced by the limited slew rate.

That would make the Stax amps less bright, which is the opposite of what people report when the hear massive difference upgrading to a KG design. In fact, many reviews of the Carbon report a huge improvement in bass frequencies, which would require the least current and thus be unaffected by slew rate limitations.

Thus, if the differences reported are real, there must be another technical reason other than current and slew rate for the improvements in bass and warmth that are typically described.

A Carbon is much more like the T2 sound than the 727 sound. Even at the same low volume level they do not
sound the same.

What makes "the T2 sound?" Why do they sound different at low volumes despite not running in to current limitations?

 

[waltzingbear]

one of the factors not mentioned so far is that the new KG DIY amps are built with well regulated supplies and generally improved current sources and loads in the build. ie morrrrreee stable. Which in my experience is morrreeee important than raw power. (beyond simply having enough) The stock Stax units are non regulated.

for me, my SRM1/MKII (modified) is just fine.

 

[strom]

Hi all. This is my first post - I've come over from Head Fi where there has been a slightly frustrating discussion of the above (it quickly descends into a flame war). I was hoping some of the technical experts on this forum would be able to offer some insight (I don't have any real technical knowledge myself!)

In essence there is a view that in order to drive electrostatic headphones to their full potential, it is necessary to purchase a huge, super-powered amplifier. There appears to be no upper limit on this - electrostatic fans will have heard of the huge Stax T2 amp which both busted the company and was also prone to busting itself. Stax no longer make amps this powerful but enthusiasts have been making huge and enormously expensive DIY T2s.

Generally speaking the electrostat community appears to take the view that Stax's own amps are under-powered and can be bettered by expensive and powerful third party amps (Mjolnir Audio amps are the main example) but even these amps are bettered by the enormous and super-powerful/ expensive DIY T2 amps.

At the same time, there are counter-claims being made to the effect that sub-£1,000 Stax amps with (i) next to no distortion, and (ii) far less power than the Mjolnir amps, let alone the T2, are perfectly capable of driving the Stax SR-009 (my headphones) and would therefore be indistinguidable from the T2 save at loud listening volumes. Of course, if you search Head Fi you will easily find posts saying that the T2 will give the SR-009 "superpowers" on another level (and the word "superpowers" has been used in some of the more colourful audiophile descriptions).

A summary of what appears to be the technical basis for these claims can be found here:

https://www.head-fi.org/threads/the-stax-thread-iii.677809/page-1204

See in particular:
- Post #18,049
- Post #18,054
- Post #18,059
- Post #18,060

Can any of you good people help to resolve this issue? Some of the third party amps are large and expensive and I would be reluctant to spend money on them without clear evidence that they do actually make a difference.

Please assume that no owner of any of the expensive amps will be prepared to participate in a properly controlled/ double-blind test of their amps.

Hi John

Please see my most recent post. I was unaware of your post but have read it now. My setup is similar to yours: Stax 009 and Stax 323s. I was also interested in your post as it paralleled my post. Have you made any decisions regarding your setup?

 

[Craylock]

The conclusion in this thread is simply horribly wrong. Closest thing to a 323 I have had is the 717. Also had 007TII. I have SRD7 w promod, Mjølnir Kgsshv, and a Blue Hawaii, and there is NO WAY the sonic differences aren't audible to 99.X% of anyone within normal hearing range. (Just passing by looking for some other impressions on the 009 today, and wish you all the very best of sound). Cheers

 

[egellings]

Big thing to worry about with electrostatic drive amplifiers is that the load (the ESL) is viciously capacitive, an amplifier's worst nightmare.

 

[Craylock]

Big thing to worry about with electrostatic drive amplifiers is that the load (the ESL) is viciously capacitive, an amplifier's worst nightmare.

I think that is one of the issues Kevin Gilmore has addressed, but I am awfully unskilled technically. Yes, probably a main difference between estat amps and conventional amps eh?

 

[Dialectic]

A more mundane question I've been wondering is whether a Class D amplifier (such as NCore or Purifi) could be used to drive Sanders electrostatic loudspeakers.

 

[egellings]

Class D could do that, but since class D amplifiers also use negative feedback to control the output signal, they could be prone to similar problems. Likely, a DSP approach using class D could handily solve the problem. I have no expertise at all in that field.

 

[DonH56]

Big thing to worry about with electrostatic drive amplifiers is that the load (the ESL) is viciously capacitive, an amplifier's worst nightmare.

I don't know about headphones, but for loudspeakers the panels themselves are capacitive, but the audio signal is inductively coupled via a transformer that also isolates the high-voltage DC panel bias from the input signal. As a result, the impedance is indeed very low at high frequency, but the phase angle actually tends to be positive (inductive) IME. There is usually a capacitive dip somewhere in the upper midrange but then the transformer "takes over" and the load moves toward inductive despite the very low impedance magnitude. The load as seen by the audio amplifier driving them, at any rate.

Modern class D amplifiers should have no problems driving ESLs. There are threads here and elsewhere (other fora) documenting folk using Hypex and PuriFi amps, among others, with ESLs based upon a quick search (plus I happened to be reading one recently wherein a Quad owner switched to class-D amps and is quite happy).

IME - Don

 

[egellings]

Electrostatic headphones have maybe 100pF or so of capacitance each. That is not hard to handle for an amplifier.