I have wanted to build a 13EM7 amp ever since exploring the late Fred Nachbaur's circuit here:
Fred went the bare bones, lo-buck but elegant route (the man was a genius - look at his other projects), and came out with a very good sounding amp.
I wanted to expand on his idea, keeping the essence of the circuit the same, but moving out of the realm of a simple starter circuit to something much more capable and polished.
I reworked the loadlines on both sides of the circuit. Starting on the output side, I amped the B+ up somewhere north of 420 volts; when yoked to a 5K ohm output transformer, the resulting loadline is very linear. This would also be a good place to point out that the plate curves on the 13EM7 look very similar to the 2A3 - more on that later.
Driving the output triode of the tube from 'roof to floor ' on this load line, will give you right at 2.5 watts of output.
For the full output, a voltage swing on the output triode grid of ~100 volts is required. The preamp triode (very similar characteristics of a 12AX7) easily accomplishes this with light loading. On the preamp stage I also went with simple cathode bias, bypassed with a capacitor to block any local feedback or degeneration at that stage.
The output triodes are run fixed bias, just like in the original. I did include a trimmer to allow experimentation with different bias points. In the end, after hours and hours of listening, the amp sounds best where Fred had it to begin with - right around -35 volts.
I do have to say that this amp certainly has the sound of a 2A3 - I think if you were to do a double blind listening test, most listeners (even picky audiophiles) would be hard pressed to distinguish between the two at listening levels below 2 watts output. It has all the warmth and soundstage that you would expect from a single ended triode, but with a crisp clarity and none of the 'mush' that you sometimes get with these type circuits. I'm becoming more and more of a believer that running triodes at higher plate voltages and with larger load impedances than generally specified can really make for some really spectacular sound.
No negative feedback in any form is used on this amplifier. I good friend and wise man once told me: ‘Global feedback is a cheap Band-Aid used to cover up any multitude of circuit atrocities’. I couldn’t agree more.
The power supply is tube rectified (6CA4) and since I had a headphone jack installed, it had to be quiet!
CLC filtering with the massive 16H Thordarsen choke and light capacitive filtering (50uF +50uF) gives a very stiff and quiet high voltage supply. All filaments are also DC fed. The chassis is segregated into two sections with all AC confined to one side, the amplification components to the other for minimum noise. The internally mounted bias transformer is also shielded and grounded.
Specifications and performance are listed below and also graphed on my website. All specs. are the real deal, measured on calibrated equipment, and not a product of my imagination or ego.
Input: 2X RCA stereo. 2V PP signal for full rated output.
1) Stereo Binding posts for 6-16 ohm speakers. Rated output of 2.2 watts/channel @ < 0.5% THD into an 8 ohm resistive load.