DCAudioDIY.com

_________________
Walt

I am using a 1K adjustable cathode resistor so I have more control over the GM70 operating point. Currently running around 100mA. Educated guess at this point is around 850 ohms. Cathode bypass capacitor is a composite of a 47uF Obbligato Gold, 2uF/200V AudioCap PCU, and a 0.22 uF Russian Teflon. I would probably go with a larger value for more extended LF, but it is what I had on hand.

David

Well -- you are already down below 4-Hz. so I would not sweat any bullets about the bass.

From what I saw in your pictures -- you did your homework in terms of staying away from electrolytics and using a nice high-wattage resistor to stave off hysteresis issues.

But -- I expect that the self-inductance of the cathode resistor (assuming that it's not a non-inductive resistor) may be an issue together with the cathode bypass caps. Parallel capacitive/inductance reactances causing an unacceptacle resonance that may at frequency that is causing you some pain.

You might try a nice big non-inductive resistor or an array of non-inductive resistors at the value of choice.

Changing to fixed bias may not be a good idea because you then lose the dynamics that the cathode bias helps with.

One trick that has history is what G.E. did in the grand old days of tubes. They used both. Cathode bias together with fixed bias that provides a starting point.

Jadis uses that scheme -- at least in the Defy-7.

This was 10-cents and I expect change.

_________________
Walt

I have some 880 ohm 50W non-inductive aluminum cased resistors that I can use. It is very close to the value I would need at my current operating point. I will need to put them on an heatsink to help dissipate the heat. I do like the idea of a mixed bias scheme, maybe having a 0-30V of adjustable fixed bias with around 75 volts across the cathode resistor at my operating point. Hmm.

David

_________________
Walt

I am curious. How does fixed bias reduce the dynamics?

ray

Looking carefully at the plate curves of the GM70, it really does not need much bias voltage adjustment to significant effect the operating point given a mixed bias scheme. Around -20 to 0 volts is all that is needed to provide adjustment from approximately 140mA to 70mA, when accounting for the drop across a fixed value resistor impacting the effective plate voltage across the tube. This means I can probably select a cathode resistor only slightly smaller that what I am using now (say 750 ohms) which would effectively limit the default current at 0V fixed to a slightly lower high value than above (maybe 130mA), and applying negative bias to adjust down to the desired operating point. I like this approach, since failure of the bias supply would still protect the tube (limited to rated dissipation) and most importantly, the output transformer. Failure of the bias supply for a 100% fixed bias amp is not good for both the tube and output transformer.

Good point about the failure of a fixed bias supply. I have repaired many an amp with old tube sockets and new tubes with slightly thinner pins that fit loosely. Bias supply failure can occur right at the G1 pin. Worse even than transformer failure was one of the ensuing fires. It was contained, but had the potential for a real catastrophe.

_________________
Walt

or



Now I am confused. I suspect I am misunderstanding something.

I am aware of the dangers of fixed bias, but I had not heard of the idea that cathode bias with a resistor and capacitor produced better dynamics than no resistor and cap. My understanding is that the main advantage of fixed bias with no cathode resistor or capacitor is not having the cathode capacitor to deal with.

My understanding is also that the power supply is the main contributor to whether an amp gives the impression of 'dynamic'. I am curious whether there are other factors.

ray