Quote:
Originally Posted by G8HQP Dave
CCS as a cathode bias for output valves is theoretically a bad idea, although undeservedly popular. For PP an unbypassed CCS enforces current balance, but it also enforces fixed average current when what you actually want for low distortion is fixed quiescent current. For single-ended, a bypassed CCS can create LF phase problems in addition to fixing the wrong current. The LF phase problems can then encourage motorboating.
|
As I discovered.
The other phenomenon I discovered was on odd occasions the LM317 wired as Current source wouldn't start to conduct but went into a sort of sulk, usually sorted by switching off and back on again.
It seemed like a great idea to keep the EL84 cathode current steady but it also stops the valve finding its own bias point if the HT isnt quite right.
Anyway, the main thing is its rock solid now, no instability issues and still hum free which is a definite result on a SE amp with AC heaters. I did take great care on twisting and routing the heater wires and referenced to earth using a pseudo centre tap with a couple of resistors to earth.
I will have to get the sig gen out and get the amp on the scope to see what power its putting out at clip, My suspicions are its only giving a couple of watts but for most of the time that's plenty, even with insensitive 86 dB/Watt smallish standmounters. And I think it's going to be well worth another version using some decent size output transformers and a bit extra HT around the 300V mark.
I have attached files of the circuit I used, including the one with CCs and Zener diode feed to screens. Built both but in my opinion it works just as well without the semiconductors.
Oh and the PSU is a simple full wave job using UF4007 into 100uF-10H-100uF and I only used a 4.7uF to decouple the HT supply to the ECC81 but it works fine and dandy.
Its a lot simpler than the Mullard 3-3, you can argue that the Mullard probably squeezes more out but for a simple power amp it does work rather nicely.
A.