One thing I have noticed is most of the SE and SET schematics I see online use cathode or "auto bias" for the power tubes. All of the SE amps I have built use this type and I now wonder if there is a reason, and what are the advantages/disadvantages of going with fixed negative grid bias on an SE amp.
Does it require changes to the driver circuit? Are the output tubes harder/easier to drive? Would you bias the output tube at the same mA as it was with cathode bias and would the negative voltage be approximately the same?
I would suspect you would get more/cleaner power with fixed bias, not have the heat/wasted power of the cathode resistor and would eliminate a capacitor in the signal path/cathode of the output tube.
Still learning and experimenting, I am thinking about trying this on one of my SE amps that does have a tap on the power supply for building this type of biasing circuitry.
Does it require changes to the driver circuit? Are the output tubes harder/easier to drive? Would you bias the output tube at the same mA as it was with cathode bias and would the negative voltage be approximately the same?
I would suspect you would get more/cleaner power with fixed bias, not have the heat/wasted power of the cathode resistor and would eliminate a capacitor in the signal path/cathode of the output tube.
Still learning and experimenting, I am thinking about trying this on one of my SE amps that does have a tap on the power supply for building this type of biasing circuitry.
With fixed bias maximum value for g1 resistor for power tubes is much smaller giving lower input resistance. So it might be necessary to redesign the driver stage accordingly.
1.) The first advantage of fixed bias is eliminating cathode R//C complex. Without this, the tone colourisation would be decreasing.
2.) IMHO the "classical" fixed bias (capacitor coupled driver,negative bias via large resistor) is not optimal, the blocking distortion may occur.
3.) The interstage driving (negative bias on the secondary "lower" point) is better, but in general the previous stage output impedance aren't enough low.
4.) The satisfactory solution is the properly designed cathode or source follower (direct coupled to the power tube grid). It has enough low impedance and enough driving capacity (current) to drive the power tube even to A2 mode.
The disadvantage of these is negative bias termination, which would damage the power tube.
BTW I use negative biased amplifiers for decades, and this occurred only in one case: the bias supply PSU diode damaged, caused HT fuse blowing.
2.) IMHO the "classical" fixed bias (capacitor coupled driver,negative bias via large resistor) is not optimal, the blocking distortion may occur.
3.) The interstage driving (negative bias on the secondary "lower" point) is better, but in general the previous stage output impedance aren't enough low.
4.) The satisfactory solution is the properly designed cathode or source follower (direct coupled to the power tube grid). It has enough low impedance and enough driving capacity (current) to drive the power tube even to A2 mode.
The disadvantage of these is negative bias termination, which would damage the power tube.
BTW I use negative biased amplifiers for decades, and this occurred only in one case: the bias supply PSU diode damaged, caused HT fuse blowing.
Typically 5% (maybe 10%) more heat. In return you get easier drive and VERY much safer operation: if the tube sucks too hard it turns itself off.
You do not get "more power" unless you drive the tube into serious distortion (as a few datasheets do).
Some tubes are recommended for fixed bias. IIRC the 6as7 often recommends not to use fixed bias but servo/cathode instead.
Folks have already mentioned the issue of grid circuit resistance and potential idling current runaway. Not to discount this, because it's a real thing, but proper design, including appropriate resistance values, anticipation of DC bias potentiometer contact failure, and even fusing if felt necessary, have been done for about a century. Not impossible.
Assuming that you've adjusted B+ voltage down by the same amount as the previous cathode voltage, power output doesn't change. Distortion when playing real music, where cathode current averages ("DC") very slightly differently at different signal levels, is slightly lower, but slightly enough that it's hard to measure. (The reason is that even-order non-linearities in triodes add to cathode current - the more linear, the less change. In cathode resistor biased stages this change envelope modulates bias voltage.)
The lower "safe" value of grid circuit resistance is a parasitic to be dealt with. In your "Schade" feedback designs, the issue of driving power is already becoming significant. Maybe some change in this parasitic load should be considered as just another factor at the summing junction, a constant to be included in your plans.
All good fortune,
Chris
Assuming that you've adjusted B+ voltage down by the same amount as the previous cathode voltage, power output doesn't change. Distortion when playing real music, where cathode current averages ("DC") very slightly differently at different signal levels, is slightly lower, but slightly enough that it's hard to measure. (The reason is that even-order non-linearities in triodes add to cathode current - the more linear, the less change. In cathode resistor biased stages this change envelope modulates bias voltage.)
The lower "safe" value of grid circuit resistance is a parasitic to be dealt with. In your "Schade" feedback designs, the issue of driving power is already becoming significant. Maybe some change in this parasitic load should be considered as just another factor at the summing junction, a constant to be included in your plans.
All good fortune,
Chris
Thanks Chris and to everyone who responded. It's given me a lot to think about and consider before I try this, if it even makes sense to. It sounds like very some are saying minor if any gains, but then others say it improved the sound of the amp. Clearly the amp going nuclear isn't a pleasant thought either!