Wattage required for biamped ESL

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Most of the information that I read about driving ESLs states that at least 100 watts per channel is recommended. Usually the lower frequencies suck down all of the power. If I only drive the speakers at higher frequencies, do I still have to supply all of this wattage? Let's say that the highpass is at 500hz. Can I get away with a much smaller wattaged amplifier?

I haven't got my amplifier building feet wet yet. And I would like to start with a smaller power and hopefully simpler amplifier. Any suggestions for an amp aimed at powering ESLs? I'm open to tube or SS ideas.

Thanks for the input.



Looking for the right amp. Well, your facing a number of decicions.
First of all the power aspect of your amp should be your main concern. All other sonic criteria are far less important.
Second, final voltage for your ESL consists out of two factors.
1) Power output of your amp. Be sure that you amp can deliver enough current. 5 [A] is no exception. If you goal is not using a step-up transformer you definitely need a Tube amp. Voltages between 4 [kV] and 8 [kV] to feed you ESL are normal.

2) The step-up ratio of your transformer. This ratio enables you to find a balance between wide bandwith and max output voltage.
e.g. If you're looking for a 7.5 [kV] output and you amp delivers 50 W max. you need a ratio of 1:150. Now here's the tricky part. The greater the ratio the lesser your bandwith. So you need to find a good step-up transformer.
Your idea of using only the higher frequenties is an option. This cutsdown the wide bandwith which enables you to use a high step-up ratio.

Good luck with your project.

I have driven biamped ESLs crossed over at about 350-400 Hz with a 30 watt, mass market amp with no problems at all. Plenty of spl, too.

I am using two transformers per speaker, each transformer has 4 or 8 ohm input and 8K Ohm out (so I guess if I use the 8 Ohm taps, the turns ratio is about 32:1, and using 4 Ohms the turns ratio is about 45:1).

Forget about tube amps. Transformers in the speakers are a compromise that is made because the alternative of direct drive is practically impossible. If you use a tube amp, you are putting two bandwidth limiting, nonlinear magnetic devices ahead of your beautiful, wide band, low distortion speakers. You may as well go back to conventional speakers. :)

In theory, using two transformers with lower turns ratios has some advantages over using a single transformer with a higher turns ratio. First, the voltage developed across the lower turns ratio transformer is lower, so you're less likely to have arcing occur inside the transformer. Second, the audio power applied is split equally between two magnetic cores rather than being concentrated in one, so distortion due to nonlinear magnetic effects should be lower (but that depends on the geometry of the cores).

I have used single transformers with 20k output impedance and found that the particular transformers I used were easily driven to distortion (the core was probably too small).

Thank you for the replies.

I plan on using my step up transformers. So far I haven't heard a single encouraging word about direct drive amplifiers. I'm not experienced enough to venture out into dark and stormy waters. Currently I am using Barry Waldron's transformers designed for ESLs. I believe that the ratio is 1:40.

You are going against alot of other opinions when you tell me to forget about tube amplifiers. Other people tell me that tubes are the way to go for ESLs. To be honest, I've never even listened to a tube amplifier, so I have to go on hear-say alone. But, I'm still interested in building a tube amplifier, eventually, just for kicks. I think that it would be fun to have. After building it I would then be able to compare for myself.

Do you have any suggestions for a DIY SS amp? Should I start with something from Pass's pen? Everyone here seems to be very happy with his designs. Are they suitable for someone who just finished his first preamplifer?

Some people like the "euphonic" distortion that tubes produce. I prefer wide bandwidth, high power, and inaudible distortion. I can see building a tube amp for fun (I restore old tubed radios), but for serious listening, I'm a solid state fan all the way.

Before you build a solid state amp, look at the circuit carefully. There are a lot of designs out on the web. A lot have no protection for either the amp or the speakers should bias drift or a component fail.

I like Randy Sloan's designs, but haven't built any of his amps yet. I plan to build up a 4 channel Opti-MOS amp some time in the next year or so.

If you like to experiment with speakers and expect to connect a lot of weird stuff to your amp I can safely recommend Soundcraftsmen PM860. You can pick them up on ebay for about $200. They have MOSFET output devices and they are pretty small. About 200W/ch at 8 Ohms. I've had one for about 15 years now and I've connected anything and everything to it and have never had a problem of ANY sort with it. Stable with ESLs, shorts don't hurt it, if it clips it shuts itself down. High-end snobs and golden ear types will sneer at Soundcraftsmen, but they are absolutely bullet proof, unlike a lot of other amps out there. More than enough power for experimental work with any sort of speaker.

ESL amplifier Issues

While we're on this topic, I may as well bring up an idea I heard Nelson Pass mention on another thread here.

If you can include the transformer in the signal feedback loop (assuming your amp is an op-amp design), then you could correct a lot of the distortion that the non-linearities of the transformer might insert. Nelson says he did it, and it worked wonders.

This said, it becomes not at all obvious which is the best way to proceed. From my meager knowledge of amplifier design, the only route i can see is sampling voltage from one side of the output coil, splitting it down with a voltage divider and shoving it into the negative side of the diff-amp input stage.

But this seems kinda kludged together. Perhaps there's some way to measure the voltage across the output coil? Or some more sophisticaed feed-back mechanism I'm not familiar with. I don't know, I'll throw this one open to the forum, ideas and opinions welcome.

On, another note, jgwinner started a few threads here a while back on a similar topic, and there's one design idea that caught my eye. If you build an amp that run's unusually high voltage (say 200 volts or so) then you can get away with a smaller turns ratio transformer (say 1:10), and thereby lessen the transformers' affect on the signal. It sounds like an elegant compromise between the high winding-ratio transformer idea and the direct drive idea. However, I'm not at all sure how transformers work in terms of current draw, and therefore I'm not sure how much current a setup like this would draw through the amp.

That Mark can drive an electrostatic speaker (which, from my understanding almost always have low efficiency) to high SPL with 30 watts suggests that Pass-style class A amps are a possible route to go. The inherent symmetry in an electrostatic panel suggests application of an equally symmetric output-driver stage, say a push-pull class A design. Again, this goes up to and beyond my understanding, so discussion would be great at this point.

So, are there any pitfalls in driving a transformer from a class A amplifier? What about DC offset? Can feedback around the transfomer still be implemented? Will the power demand be uneconomical / impractical? Is it possible to implement a high voltage Class A push-pull amp (say 100 volts or so)? Are there any other routes toward this problem?
<b>G Daddy sez

Are there any other routes toward this problem?</b>

Yes. this application is just begging for a high power, say 845 or GM70 PP tube amp, with a custom transformer. You get the advantage of low turns ratio because the tube are already swinging hundreds of volts, and the reliability of having <i>linear</i> devices which are designed to run at these voltages.

Many of the high power tubes were designed for and acheived long service lives when operated conservatively. Heck, one of my domestic receivers is still using almost all of the original tubes, and it's been run for several hours a day for nearly 40 years. It's out of service at this moment because the psu caps finally failed.

I wish you success with your project.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.