To bypass cap or not to bypass cap

[torquil]

Hi! I have a question for the ones learned in tube amplification:

What is the difference (apart from the gain factor) between the following two options for a triode standard configuration (everything else being equal):

1) No cathode bypass cap
2) With a cathode bypass cap that is large enough not to influence the frequency response at hearable/relevant frequencies. E.g. if |Z| of cap is equal to Rk at f = 10Hz.

I'm mostly interested in the frequency response, but maybe there are other differences as well?

Best regards
Torquil Sørensen

 

[donnyjaguar]

There's no point using too-large a value of bypass capacitor. All it will do is let through sub-audible signals that could cause instability (cycling/motorboating) with the power amplifier section.

 

[rmconner80]

Ouput impedance for the stage. This and a wealth of other great information is explored in the following:

http://www.aikenamps.com/CommonCathode.htm

 

[DT7]

What the Aiken article touches but doesn't go into great detail on is the effect of the negative feedback with an unbypassed cathode resistor. You get a cleaner (more accurate) signal with an unbypassed cathode resistor, due to the feedback...as long as the plate load remains light. If you load down any gain stage too much you'll begin introducing distortion into the signal. Unbypassed cathodes have a higher plate resistance and, therefor, require a lighter plate load to keep the signal as clean as possible.

 

[donnyjaguar]

DT7, by negative feedback are you referring to local cathode degeneration?

 

[DT7]

DT7, by negative feedback are you referring to local cathode degeneration?

Sorry, I'm not familiar with that term.

What I'm referring to is the change in cathode bias voltage that occurs when the tube begins conducting A/C. This change opposes the incoming signal at the control grid through the grid resistor that references the grid to ground. The effect acts as negative feedback. When the cathode resistor is bypassed, the A/C signal passes to ground without creating additional opposing voltage at the cathode...no feedback occurs.

 

[torquil]

Thanks guys! I have a Ceriatone Hiwatt clone, and was considering some modifications. One was to increase gain in the first triode by including a bypass cap on the Normal channel, to increase its gain.

However, I will instead try something else: include the unused half of V2, but only on the Bright channel. Also possibly remove/change the cathode bypass cap on the first triode in the Bright channel as well.

That way I keep the fine tone on the Normal channel, and will have an option for more preamp distortion when turning the gain knob on the Bright channel.

Hmmm, this could mean I cannot mix the channels successfully, since one has an even number of gain stages and the other has an odd number, so the signals will approximately be mirror-images of one another, at least when playing clean(ish).

The Normal channel is actually modded to suit bass guitar. But its also glorious for clean guitar. So for me, it is a one channel bass amp, or a two channel guitar amp! The Normal channel is great for bass when treble/mid/presences are turned a bit down, and also great for clean guitar on many EQ settings.

 

[donnyjaguar]

DT7, you just described cathode degeneration. I just wanted to make sure we were talking about bypassing a single voltage amplifier as opposed to a pair of tubes in an output section with shared cathode resistor and the typical associated NFB loop.

 

[torquil]

For future reference, I found a page containing some measurements of THD with and without cathode bypass cap. As mentioned above, it seems that THD is lower without the bypass cap, even for the same output signal strength (at least under those conditions). However, it doesn't seem to mention the size of the capacitor in question.

Ref: http://www.angelfire.com/electronic/funwithtubes/Amp-RC-Dist.html

Torquil

 

[JJman]

LED bias?