Which high power SET

If it's topology, it means harbeths with its low sensitivity are generally incompatible with any SET amp, irregardless of it's power output. Given a hypothetical world where both push pull and SET generate the same power output, does push pull drive harbeths better simply because of topology?

Yes!! What is not getting the air time it should in this discussion is the fact that to take advantage of an SET, **you must not push it past about 20% of full power**!

You can go past that amount with a push pull amplifier and not have ill effects.

This is all about your amplifier investment dollar- with a speaker of efficiency that low, your investment dollar in an SET will not be served due to the fact above.

The impedance curve is another issue! It is pretty evident that the Harbeth is expecting that the amplifier used with has some negative feedback to linearize the output of the loudspeaker. Your EL34 amplifier has that feedback. If you put a zero feedback amp on that speaker its likely that the range of that high impedance in the midrange will take on some prominence.

To understand more about this phenomena, see this link:

http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php

I guess Atma's point also has relevance to my situation, as 20% of 200W is still 40W, which should be able to push my speakers to a little more than 100dB, which is pretty loud.

The bonus is that at 40W, the amps are still in Class A1. Above that they begin to move into Class A2, which would mostly be for big dynamic peaks.

In any case, they work superbly with the RSIIb speakers. However, if you were to buy them commercially (if they were available) you'd be paying a LOT of money, as the parts cost alone was over $11K, with about a year of my time spent in planning and building. The 2.3kV voltage on the output tubes might also cause a little consternation in the family unit...

So I guess the moral of the story is...you can power a low efficiency speaker with an SET very effectively, but it's better if it's a monster-powered SET so you don't operate anywhere near full power for most music, and you'd better be prepared to spend a lot of money doing it.

Would you agree with that, Atmasphere?

Ait, congratulations on what is certainly one of the most spectacular DIY amplifiers ever!

As a point of information that may be relevant to the thread, do you know what its output impedance is, or alternatively, its damping factor?

Also, what load impedance is its output tap optimized for?

Regards,
-- Al

Allow me to back up Ralph's point on impedance from personal experience.

For years I owned a pair of highly regarded speakers which the manufacturer rated at 90 dB efficient. I auditioned many amps with those, both tube and SS. One day a friend brought over a tube amp he scratch built which he estimated would be rated about 30-40 wpc. Since he'd had good results on a few other speakers we were both surprised when it simply fell on it's face attempting to drive mine.

After some discussion he suggested two reasons for the failed performance. First off, my speakers were designed for time/phase integrity so had first-order crossovers. The crossover board could be viewed by removing a cover on the back of the speakers and that revealed a very complex (number of components) board. My friend joked that all that was needed was the addition of a transformer and they could power themselves.

The second reason was the impedance plot. It dipped slightly below 3 ohms at two frequencies. I don't remember exactly now but they were around 50 Hz and at 10K Hz.

His amp worked with other 90 dB speakers but failed with mine since they could not adequately drive that load.

In fact I owned a pair of Ralph's MA-1s for nearly a year. With non-complex music they sounded wonderful. But with complex music (full orchestra symphonies, big band jazz, etc.) at moderately loud levels, even their 100 watts was not adequate. That was the fault of the mis-match, not Ralph's amps.

Al, they have an output impedance of about 1.2R, which is in large part due to their 13K : 4 output transformers (Monolith Magnetics S-833); they spec them at 12k:4 but my measurements say 13k. The OPTs weigh 62lbs each and have an amorphous double-C core with varilay winding and teflon interwinding insulation. I'm running the output tubes at 2.3kV and 160mA, which is well within the specs of the 833C with forced air cooling (400W max plate dissipation). I had plans to add in some local NFB around the 833 to drop the output impedance further, but listening tests told me it wasn't needed - bass is detailed and well controlled as is.

The Infinity speakers are actually a relatively benign load below 100Hz, ranging between 6R down to 14R at 40Hz and about 10R at 20Hz. They dip to 1.6R at 9kHz, but there's not much energy up there and the top end sounds just fine even with the low damping factor in that region.

The power supply on the amps is massive (about 3/4 of the amp chassis is PS) with 400 joules of energy storage just for the output stage! That helps a lot with transient power demands, as does the additional 35 joules of energy storage for the driver stage; very helpful when grid current is demanded in A2, at full power it consumes 200mA of grid current!

So, the robust power supply combined with a relatively benign low frequency region in the speakers, allows me to get away with a zero NFB SET amp where you might not think it possible. Again, it's all about the match among system components, a speaker that dipped to 1R in the LF region would most likely be a boomy disaster.

PS: do you live anywhere near central NJ, Al? If so, you could stop by for a listen any time. Just let me know...