• 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.

The perfect load for an OTL

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

I've been extra busy recently--putting in a new kitchen at home really eats up the hobby time.

A math prof. and AES member, whose opinion I respect, recently evaluated my driver concept and pronounced it sound, so that's exciting. I'm looking for more input, but I'm trying to be circumspect until I file patent applications. I'm still researching patentability, but it's a green light so far.

It's frustrating trying to keep my mouth shut and gather info when I'd rather be floating my ideas publicly, but I'd like the option of getting paid for this someday.

I've more fully developed the bass version of my driver concept. The mid-fullrange (high-impedence?) version I've been dialoging with you about still needs some fleshing out.

On the whole, I'm trying to dream up ways to leverage the strengths of my motor design. As I see it, these strengths include:

1. Easily scalable magnetic gap height and length while maintaining high field strength and uniformity
2. Very large Xmax values and extreme motor strength
3. Fully pistonic operation into high frequencies while maintaining reasonable transducer area.
4. Underhung Air-core voice coil with extremely low self-inductance and virtually no hysteresis distortion or eddy currents
5. Enhanced heat sinking
6. High magnetic efficiency and inherent magnetic shielding
7. Magnetic braking controls over-excursion beyond Xmax
8. Design eliminates pole piece and dust cap, both sources of distortion

But, until I file, that's about all I can say about the driver/motor topology.

Have you been giving high-impedence drivers any more thought?

parallel tubes

I know your proposed topology suggests use of a single valve as an SET circuit....

Clearly, however, you could get much closer to the ideal load impedance for the valve you are using if you use several valves in parallel. You may think that it detracts from the layout, but I think it would work at least as well. If you ran say 4 2A3s in parallel you could get substantially more power and reduce the required load impedance by a factor of 4. For push-pull amps this is also a very desirable approach. Only drawbacks I can see are increased cost, extra chassis space and potential matching problems if you share the bias. For push-pull amps, it reduces the effect of unmatched valves. i have done this before, mainly to drop the OPT turns ratio and therefore improve the OPT's performance & coupling.

Also, with an SET circuit, aren't you very worried about all the DC current you will have to run through the voice coil to achieve the correct bias for the valve? Seems to me that half the problem with SE amps in the first place is gapped transformers that are compromised to allow for DC current?

Interesting ideas.

In the intervening time since my first post on this, my thinking has changed somewhat. Though my motor design has the potential for excellent heat sinking, I agree that it's probably asking too much of a voice coil to carry much idle current.

One exception to this could be the inside-out concept Mohan floated where the amp modulates a field coil which moves a permanent magnet or DC-carrying voice coil. The field coil could be engineered as bulky as necessary to handle idle current since it is not called upon to move. It could also be wound for any impedence your heart desires. To avoid forward displacment of the transducer because of idle current, you could wind dual field coils fed directly by a pair of output tubes in balanced single-ended mode so their idle currents create canceling fields.

One thing I haven't straightened out in my mind is whether or not this system would create "leakage inductance" and therefore limit bandwidth. Any thoughts?


Great news on your progress. Let me know when you have a full range driver. I would like to try.

I am working on a single ended amplifier for electrostatic speakers like Quad ESL57. So I am studying the behaviour of these speakers. Perhaps some new ideas could come up that could be applied to cones as well.


Sounds like a variation of a previous theme.

Back in the 70's, Accoustat made electrostatic speakers that took a line level input and used tubes (HV Television deflection types as I recall) as the "amplifier" No signal transformers needed.

Made plenty of sense, especially if you were using tube amplifier.

A Tube amp had a high impedance to low impedance transformer. Could be thought of as a high voltage/low current to low voltage/high current, step-down.

Go through some speaker cable.

Then at the electrostatic speaker, go to a low voltage/high current to high voltage/low current transformer, step-up transformer for the high voltage electrostatic.

An awful lot of iron to turn a down, then trun it back up again.

Maybe some historical research could give you some insite. Some schematics of the Accoustat designs might be of help. White papers published at the time could also be useful. I do not know if any patents were issued.

Jim Strickland, the brains behind Accoustat, is still (as far as I know) still working at Rockford Fosgate/Hafler.



I have already designed this amp and now waiting for power and output transformers. Because this amp will be offered as a kit, it will not be a direct drive. A separate transformer to replace the existing unit inside the speaker will be provided. This is the plan at present.

Yes, the logic of step-down and then step-up does not make sense from a minimalist point of view. I have raised the issue with many knowledgeable people in the electrostatic scene. So far my understanding is that ESL manufacturers prefer to have a more versatile product that can be connected to a variety of amplifiers.

I am sure that the idea is not new. All electrostatic manufacturers have known this for years. There are also a handful of direct drive amplifiers for ESLs. Accoustat made one in the nineties, using 6HB5 tube. Because of lethal voltages involved, any direct drive amplifier needs to be packaged with the ESL. I liked the old Beveridges a lot and also had telephone discussions with Harold.


I was just trying to site an example that answered the very first question. My answer was one case of where an ideal (or close to ideal) load was interfaced with a tube. I assume it was an optimised tube/driver engine.

I did not see the type of engine I brought up in among the threads. I trying to point a different direction.

Of course, the are many type of tubes, and many types of loads. Some combinations will match up better than others, for a variety of complex reasons.

I am a big fan of trying to learn from the past and take advantage of other peoples research. Or would that be mistakes?



Yes, I am also learning as I go along. Our forefathers have already thought out most of these things and we are in the process of uncovering their findings. Please see my posts in the Speakers forum “Tube amp for ESL”. There are others in this forum thinking on similar lines.

Are you an electrostatic fan? I am. I like Lowthers and Electrostatics.
I have STAX-F83 at home and Lowthers at my factory.


Sorry about the 4+ month to reply. Somehow I over looked a your question to me. Normaly I would reply to any one who took the time to address me directly. The recent post prompted me to look at the thread again.

To answer your question, Yes I do like elctrostatics. (or their close cousins, the ribbons) My reasons is because they push my good buttons: Detail, accurate bass and imaging. Their minuses (low volume levels and "low" bass output) are not high on my list. Most speakers exagerate both bass tonal quality and levels.

Stax are wonderfull. Most of my headphone listening is in a work type enviroment, so Grado does the job for me. I do not know of the Lowether. I know they enjoy a good reputation.

Regards (and sorry for the delayed reply)


Ps Been to Australia, great country. Any one reading this would do well to visit.
Any further progress on your speaker design?

Hello Mohan,

Well, the answer is not as much as I'd like, on this design anyway. This driver topology requires its own novel means of suspension, and designing that has been the bottleneck. Still working on it.

In the meantime, I've also been working on a new compression driver that should offer a very low Fs and good treble performance without the internal phase cancellation of other drivers.

Actually, I've been expending most of my effort on a new bass driver with extremely long stroke made possible by a suspension design I cooked up while working on my other full-range driver. This suspension is very stiff across the axis of travel, but highly compliant along the axis of travel, exhibiting almost no inherent return force until near its limit. This makes possible an extremely linear bass driver with a very high Qms that will offer something like 60mm of one-way Xmag, 70mm of one-way Xsus, and an Fs in the single digits to low teens! It's pretty much a complete rethink of the bass driver. I placed special emphasis on a strong motor--the magnetic return circuit boasts a cross-sectional area about 3 times that of most high-excursion drivers, so it will be able to throw lots of flux across the gap. I hope to have a R&D prototype fabbed in the next few months. Needless to say, I'm pretty excited about it.

I am not laying claims to be an expert. My company Australian Electric Valve Importers Pty Ltd., specialises in high power (Broadcast) and high frequency applications using electron tubes. Government projects occupy a greater proportion of my time.

I enjoy listening to music and so I design Tube Amplifiers as well.

In 1992, I purchased a small (3,000 sq.ft) factory/warehouse in Bayswater.
Address is 22 Michellan Court, Bayswater, VIC 3153 (Melway Map64A4). I am there most days between 1100hrs until dark. You are most welcome to visit me

You may contact me (03)9720 8720, 0500 566 616 or on mobile 0409 415 315.

Visit for sure.

You bet I'll come over sometime or other, Mohan. I am working in Henderson Rd Rowville, off Ferntree Gully Rd about 1 mile east of Stud Rd so it's only ~10 minutes away. I flap my gums on an international forum and get to meet people in my own back yard. Fancy that!
Last edited by a moderator:
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.