• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Tube amp and electrostatic speakers.

A friend of mine lives by tube amps, I am very interested in
the sound of this type of amplification. My friend lives in Virginia and I in Arizona. I am working on an Electrostatic
Speaker project, and my question is "Is there a project on line for a tube amp that will drive electrostatic speakers"?

I have read pages where people have use the 300b tube, and a few others. And they report 15W of power. I am confounded by this number. But then I read they are using highly sensitive speakers (Unlike Electrostatic).

So are tube amps made for driving Electrostatic speakers?

Or should I be reading more and posting less? (If so
please point me in the right direction).

Tube amps are usually better for electrostatics!

Many people will tell you

"You've gotta use tubes with electrostatics, man".

I tend to agree with them. A tube watt into an electrostatic speaker counts more than a trannie watt.

The problem has several key factors. One of them is that tube amps are inherently high voltage low current devices. This is the same as electrostatic speakers. However, electrostatic speakers use transformers to match to "normal" levels i.e output from transistor amps. Tube amps also use transformers to match to "normal levels" as found in most dynamic speakers.

Still, a good tube amp will usually beat a good transistor amp for electrostatic speakers.

If one were smart, one would drive the electrostatic speakers directly from the tubes -- thereby removing two levels of transformers. I have a pair of headphones that are made that way. The sound is .... oh so good. Check out http://www.headwize.com projects gilmore for more info.

Still, a very low impedance transistor amp will usually get more bass out of electrostatic speakers.

Keep in mind that the efficiency of electrostats can be brought up quite nicely by decreasing the distance between the plates and the membrane. If you're doing your own, you will have complete control over the spacing. The downside is that you will lose some transit capability. The tradeoffs are many, but it's something to think about.
As for driving ELs directly from tubes...it's an obvious match, as a 6550 can easily swing hundreds of volts; the 6CA7 is rated for a plate voltage of something like 800V, if memory serves. Clearly, if you can get rid of the transformers, you're on the way to Nirvana, but getting the iron out isn't as easy as it first seems. I played with this off and on for years before mothballing the project. I intend to go back to it one of these days, but I've got too many other irons in the fire at the moment to fiddle with it now.
Acoustat had a commercial product (the X?) years ago. It was notorious for blowing up on a regular basis, but when it worked, it was glorious. I had two friends who owned a pair each. It was some of the best sound I had heard up to that point.
Note that an EL unit is a purely capacitive load. If you choose to run a tube amp directly into the speaker, you have three options: Leave the EL purely capacitive, add inductance to try to balance the capacitance, or add resistance (possibly with inductance) to try to swamp the capacitive load. The last option is clearly inefficient, as you're losing power in the resistor.
My intention was to use a direct drive EL as the midrange in a triamp system, thereby only having to deal with the capacitance (and phase shift, etc.) over a three or four octave range, which makes life much easier. Further, since I wasn't going to try to reproduce low frequencies, I was also limiting the amount of transit; since I didn't intend to go above 5kHz or so, I would also reduce beaming. I was shooting for a membrane about two or three inches wide and five or six feet tall.
Perhaps some of this will stimulate your imagination.
As for resources, check http://www.tubecad.com for tube driven headphone amps and such. Scaling up one of the amp designs might work for a main speaker.

Voltage required for direct drive

You are likely to need in the region of thousands of volt for an elecrostatic speaker.

Plitron aftermarket transformers for electrostatic speakers come in 1:50 and 1:75

Let's say 1:50

Now, a 40V peak output amp can provide about 100W of audio power. With the small trannie, you are still looking at 2KV. So it looks like you might be able to get by with 1000V per leg since the application is inherently balanced (reads bridged). 300B's anyone?

I ran my experimental direct drive rig with the outputs from a pair of 6BQ5s (I'll double check the tube type, but it was a 9 pin tube, not one of the big octal jobs). I don't remember exactly, but shooting from the hip, I think the audio output was on the order of, maybe, 2-250V. I think I was using a bias voltage of around 2kV, no more than 3kV. The sound output was quite decent, given the wattage limitations.
Yes, the membrane was fairly close to the plates.
Anyway, point being that hundreds (as opposed to thousands) of volts for the audio signal will work quite well for direct drive.
Now, 300Bs...there's an interesting idea. I only played with pentodes. I imagine that triodes might be a worthwhile avenue to explore.
Okay, and while we're at it, how about a tube X circuit (including front end) straight onto the diaphragm?
If you're looking at using a step-up transformer to drive the thing, all you have to do is use a standard AF transformer 'backwards.' Works like a charm. I've even heard of people using power transformers as step-up transformers, but I can't vouch for the frequency response.

Petter, GRollins.

Wow I was looking to build a trannie amp as Petter puts it.
And was allured to the tube amp. (Quite seductive)
I have not built an amp ever.
And yet you both make it seem so easy, to speak of it.
It seems that tube is the way to go here, I did not realise that tube amps worked in hundreds of volts then droped down using a transformer
to accomodate standard speakers. I am currently going over the site http://www.tubecad.com as mentioned by GRollins. The Headwise seems to be down?

Running the EL's right from the tube seems efficient enough.
I will have to research what all goes into a tube amp.

I looked around the net for assembled tube amps (300b) and
about fainted. ($20,000 a pair) mono blocks.

I am interested in clean reproduced sound with no additives, unfortunately I want the world on a kmart budget.

The EL's will be hybrids crossed over at around 250-400hz, the upper end left alone unless it sounds very unfavorable.

I do have step up transformers here if I need to use them. I had bought them from "ESL nfo Exchange".

So as you can see I do not have a definitive plan as of yet.

Currently I have a Sony integrated at 85watts, the power is going into a step up transformer 50:1 and it is barely driving the speaker. The speaker is 6" x 6'. There is no crossover point. I am not trying to reproduce the low end. Even Violin is barely driven.

well enough rambling for now. Any tube help is greatly appreciated.

If you are crossing that high, and don't need to play very loud, you should definitely consider going with tubes.

Have you considered getting into som Quad ESL57 or 63? depending on your location, you can get 57's cheaply.

Search the web for Electrostatic web ring/circle or something like that -- there are special forums/sites for this sort of thing.

My Stax OTL's require 600V peak to peak, and they are positioned next to my ear with good interface matching ....

Word of warning: if building your own speakers steer away from kits if it seems to be too good to be true. At the very least do a search on Usenet. Most (all?) commercial kit offerings for speakers are crap.

300B: Russian units are very competitively priced these days.

With tubes you have the advantage that instead of an output transformer that is made for 4-8-16Ohm loads you can convert the role of the transformer and use a step up transformer to drive the esl directly. The problem is you need to design the amp specificaly for an esl since it won´t be usefull for any other load. And you should consider that esls are pure capacitance loads so you will have some losses from the equalisation circuit. You can find I think more info in the esl circuit.

I think having a look at the following valve amp kits may provide some ideas for a reasonable price.


These are very good designs at a very good price. While reading how good the amplifiers sound the review is provided by hi-fiworld magazine, one of the least biased british mags.

As the technicians who design these amps have some very good knowledge an email to them may identify what kit would work well with electrostatics if you say alter the design. You will also find .pdf circuit diagrames on the designs.

Hope this helps
Mr Mac,

Hifi World may be unbiased when reviewing other peoples kit (they tend to go for the irritatingly vacuous flowerery nonsense instead of bias, but do have a Kessler like anachrophile/label worship edge as well) but when they review their own kits? I gave up reading them because partly because I got sick of that pretence - so strange that they never reviewed one of their kits badly- and when they started reviewing 'upgrade kits' and blahing on about how drastic a single resistor change was I gave up on them totally. Least biased? Most biased, I'd say, as I don't know any other magazine that makes money from punting its own equipment on the back of its own reviews.

Just my 2-pence worth

Thanks for youre thoughts,

Have you heared one of their kits???

I think not! For the peanuts they sell the kits for the reviews are spot on. You can't say they are biased just because they gave a good review to the equipment.

They do not promote label worship. From every review I have read and susequently demoed the equipment I have to say the reviews are spot on. They also describe the equipment well.

The magazine is also one of the last independent mainstream left in UK the rest are owned by companies and some employ people who have intrests in a manufacturing company they have even reviewed their own kit. At least where hi-fiworld is concerned it is in the open and as such it would be stupid of them to review it incorrectly as everyone would return the kits. It would be common knowledge

However as you say that is my opinion and if everyone agreed then the world would be a dull place.

In relation to the upgrade kits and "them" going on about one resistor change etc.. etc.. Well with the low number of components small changes do make a diffrence. and most upgrades include more than just one resistor

The main messege was they have alot of expertise and may be able to reccomend the best way to drive electrostatics and alterations to make etc.

Cheers :eek:)


p.s. If you get the chance listen to one of their kits and I am sure you will agree the review was not that rose tinted for the price point.

Of course a company will smile as they tell you there product is better for the buck. I do not deny that for one second. But then if you cannot hear it or others you know have not heard it I would not buy nor build it.

I do not know a company yet that will gladly try to sell you the compitition because they know it is a better product.



2001-10-13 2:25 pm
The transformer at the output of a tube amplifier is rather a good thing than a bad one. Of course, this will be a high-quality transformer.
Even two transformers in series are not bad. Some electrostat speakers use the leakage inductances of their internal step-up transformers as the components of their crossover networks. Also these 'speakers become less reactive loads at highs.

From other hand not any tube amplifier can withstand the torture of a pure capacitative load. This is especially concerning those single-ended triode amps, which are so popular now. It is a truth that a well-designed pentode push-pull amp will do this job much better.
I am also curious about building a tube amplifier for my electrostatics speakers. I have heard that it is possible to drive the speakers directly from the output of the tubes, as the step-down and step-up transformers can negate each other.
I am wondering if a small-mid sized amp could be made for less than $200. Let me first give some specifics about the type of amp that I want. It must be able to drive the ESLs, of course, at a moderate volume. I plan on having an active cross-over before the amp, so it will not have to worry about the bass (xover at ~300hz). I also have access to some junked scientific equipment (some with tubes of their own) for parts such as a chassis and some other components. Is there hope? Should I bother? Does anyone have a suggestion for a book on Direct Drive Tube amps?


2001-10-13 2:25 pm
Active crossover

Active crossovers are the worst things of audio electronics.
Any high-pass filter significantly increases the crest factor of a musical signal(this is apparently not the sine wave). The sharp spikes enter the input of the power amplifier and overload it. Active crossovers often result in harsh and dirty sound. Their use in professional sound reinforcement systems is often dictated by the complex driver matching problems includind time alignment. For the latter purpose and the simplicity of adjustments they often operate in digital domain.
For QUAD ESLs it is essential not to omit their own passive crossover and not to split it for biamping! The best amplifier will be a classical pentode push-pull one rated at least at 10-15 watts. You may look for a circuit like QUAD II.
THe wideband output transformer is not a problem if properly designed. I think any attempt to drive the electrostats directly from the tubes will result in complex and far more expensive circuit, and the sonic benefits will be obscure.
I also do not recommend SET amplifiers with any kind of capacitative load.
Sorry, but you will spend at least $200 for just a pair of acceptable output transformers.