Quicksilver Mono Amp Model 1300 - creating Auto Bias wanted

[tiefbassuebertr]

Tube amplifier without resistors in series to the cathode pins (i. e. without Auto-Bias version) needs still a negative voltage for idle current adjust. In case of faulty variable resistor the tube runs on the maximal possible quiescent current because the negative voltage isn't longer present.
This issue is present by two Quicksilver monos, which I have on the table.
In the attachement you will see the schematic in original (genuine) condition (first and second page of pdf attachement) so as in the modified condition (third page) with removed bias voltage unit and added resistors/Caps in the cathode lines (not yet tested).
Are there things that must be considered additional?

[costis_n]

Looks good, but keep in mind that the auto-bias will "steal" voltage from the plate voltage. If you bias for example at -30V, you will have PlateVoltage - 30Volts. So you will have a little less output power. It may or may not be a problem (most possibly not)

[kevinkr]

Connect a large value resistor (relative to the value of the pot, say 100 - 220K ) between the negative bias supply and the wiper of the pot - this will prevent pot failure from resulting in ruined output tubes.

Autobias probably is not a good idea with these amplifiers as it will introduce yet another LF pole in the closed loop response and if these are like the other Quicksilver amps I have worked on they are Williamson topology and small changes can make them unstable. I have yet to find a good sounding large bypass capacitor for cathode bias, although perhaps one of the Janzen film caps might work well. This will also substantially reduce output capability and shift the output stage operating point significantly. (Depends on what output tube is used, the hit with KT88/6550 is significant, 8417 less so.)