Steve,
Nice topic. Here's the story on how these amps came to be.
It may be interesting to note what speakers were in my system at the time the amps were being developed. I restored a pair of ESL-57's and the RM-10 was born. I ran it on a breadboard for 5 years on those speakers while I was proving out the 35 watt EL-84 application. RM-10s have been putting out 35 watts per channel for 15 years now with good tube life, regardless of the nay-sayers who believe only 17.5 watts is possible. An RM-10 owner told me today his technician friend said the RM-10 couldn't be 35 watts per channel. So he measured it on his bench and got the 35 watts. I asked what comment followed. The tech said he still didn't like it because it wasn't
normal to have 35 watts, 17 watts is
normal. I hope I am never so lazy as to design a
normal amplifier. You can design a 17 watt EL-84 amp in about 1 hour. Just lift a common driver stage and the application in the RCA manual tells you the rest. It states the required B+, bias, transformer impedance. You call up the tranformer guy and tell him you want a power transformer with the RCA specified voltage and current and an output transformer with 8,000 ohms impedance, get as much bandwidth as you are willing to pay for and you are done. Now you can focus on what is
really important, how it looks and what popular brands will be on the menue. aa Though I say this in jest, it is the method by which most amplifier are
designed. Sarcasm aside, this process holds no interest for me. The 17 watt design is first seen in RCA Tube Manual RC-19 (1959) and has been is unchanged, since the tube came out. There are just 3 applications: Class A SE 5.7 watt, Class AB 11 watts and Class AB 17 watts. Fisher, Scott and Dyanco made enough of these respectable little guys and most people think that's all that can be done.
On the bench the RM-10 measures a bit over 40 watts one channel driven and 35 both driven to clipping but the 40 watts is more accurate for music. I chose 35 because it is in the model number of the Dyanco ST-35 and SCA-35, though 35 is the power for two channels together. When the RM-10 came out many assumed that the 35 watts quoted was the total power, not per channel. Anyone who has a RCA tube manual post 1959 can look up the 17 watt application that runs at 300 Volts. Mine runs at about 670 V yet the tubes run cooler and longer than the 17 watt amps.
Time passed and I wanted to try the ESL 63s. Mono RM-10s ran them just fine but I wanted a stereo amp and the RM-200 appeared. The RM-200 was created to solve two problems.
1. How to make an amplifier with fewer tubes that drove more difficult loads than the RM-9.
2. How to make that amplifier simple, elegant, economical in energy use, transportation cost and cost of upkeep. Part of that backfired on me when my Asian distributor said "My customers don't understand why you charge more for an amplifier with half as many tubes as the RM-9 it is replacing"? That's like buying amplifiers by the pound. I also informed him that the RM-9 would have to go up to $5,000 due to several shocking material costs. I wanted to keep a 200 watt amplifier in the line. For 6 years now it has been class A $$$ rated in Stereophile rec. components issue. At least they aren't tube counters. The review is in the archive at:
http://stereophile.com/tubepoweramps/560/ For some reason the link on my website takes you to the main page so use the link above.
Electrostatic speakers do present challenging loads for amplifiers as do some dynamic speakers with impedance dips or reactive loads. I do not agree that OTL amps and ESLs mate well. In the bottom end when the impedance is high and the current is almost nothing they will bump up the bass level and play louder than most transformer coupled amps, but in the mid and treble range they often cannot supply the current demands and distort in some abrasive ways. While I had the QUAD 63's a no feedback Circlotron (not Atma Sphere) amp was auditioned on the 63's. The bass was overbearing. I measured a rise of +8 db at 50 Hz. This is no small thing. At 8 KHz the treble was down about 6 dB so we were listening to a 14 dB frequency response error from low to high. Of course it sounded different.... and a different I could not live with though I was told that one happy customer had just that combination. Well, he sure had boomy bass to spare and not much else.
An amplifier has to have good damping to handle loads of wide impedance range. Feedback (or a lot of tubes, or both) is essential to do that. The Music Reference/Counterpoint SA-4 could because it had both of the above. I have been using an OTL (also called Direct Drive) amplifier on my ESL for 5 years now. It operates at the full ESL voltage of 2000 V and is very reliable with long tube life (first set still going strong with no signs of wear). There are only 2 output tubes per channel and the tube plates go right to the ESL electrodes. It is neither the Acoustat circuit or the Beveridge circuit and runs at less than half the voltage. Sometimes I actually do designs at lower voltages. This may be shocking
but I use a single ESL as a patio speaker. The amp sits in my lab and I have 3 silicon rubber insulated wires about 20 feet long that run the 2000 volts to the speaker. The speaker is thus light and easy to move around and the tubes are safely inside the lab on the amplifier breadboard. Makes a great patio speaker because the sound doesn't go all over the place which makes the neighbors happy. Because it does not diverge as much as a small cone speaker the listening level is more uniform for those sitting near or far. It's in a wood frame with a one foot square plinth for the bottom. I have no fear of moving it around barefoot on the concrete while it's playing. Could this be the first Electrostatic Patio speaker?
Hope this history of the amps kicks off a discussion of user experience be it in agreement or not.
