Time for a progress report on the OTL as there has been a lot. I got to thinking about what early OTL amps lacked that would be an important advancement.
First a few thing that I bet Julius would have done if it was possible then.
1. Eliminate the output capacitor and add a servo to keep the output at 0 volts DC. The early amps he built for himself had no coupling cap or way to zero the output DC. He had a pot to trim it and its hard to know how much it drifted over short and long time. He did design it for 16 ohm speakers which makes it less of a problem but I doubt he could keep it to 100 mV or less. Fifty mV is today's standard and I have achieved about half that and no output capacitor.
It is interesting to note that the circuits Julius published in AES journal were both direct output but none of the amps he, Harvard electronics and New York Audio Labs sold were direct coupled.
2. His early amps had no line isolation. There was a potential hot chassis. In production amps he added one.
3. Employ a PI network (cap, choke, cap) filter for the b+. He used a lot of feedback to reduce output hum to reasonable levels but the amp had to continuously cance hum via feedback. The PI network reduces the B+ hum and noise by 1000X which is 60 dB. He achieved hum reduction by that large amount of feedback.
4. Have matched plus and minus supplies which he had in his units but had a single supply in the production models.
Further improvements are:
Simplify the circuit and have the gain tubes in the driver be a triode rather than his very high gain (x1000) low current pentode. The triode gain stage I use is just 50X. This results in much less feedback.
