Bypassing Plate resistors

[PlinytheWelder]

I see that on a lot of high gain amps that the plate resistors in a stage are sometimes bypassed with several hundred to several thousand pF's.
Is there a rule of thumb to play this game?
I guess that it's a low pass filter of some sort, but I wonder how it sounds?

[dave g]

You're just shunting high frequencies to AC ground - simple as that. A plate bypass cap is effectively the same as a cap to ground.

[Bob-I]

I see that on a lot of high gain amps that the plate resistors in a stage are sometimes bypassed with several hundred to several thousand pF's.
Is there a rule of thumb to play this game?
I guess that it's a low pass filter of some sort, but I wonder how it sounds?

It's done for 2 purposes, 1 to prevent HF oscillation and 2 to kill the bacon, kill the sizzling, piercing highs.

Some amps also put this cap from plate to cathode, same affect, shunts the HF AC to ground.

[dave g]

Putting a snubber from plate to cathode a la Dumble is only effectively the same as a shunt to AC ground when the cathode resistor is bypassed. Otherwise, it's acting as an additional source of negative feedback and will sound a little different.

[j-po]

Bypassing with an electrolytic is not perfect though. ESR will allow some nfb via the ceramic cap.

[dave g]

A teeeeeeeeeeeeeeny tiny amount, sure - but the effect is negligible.

[Andy Le Blanc]

It does work, start small and sub other values etc..... it will attenuate the
top end, and work as a way to "un-brite" the tone. If your ears are shot in
the high freq you'll need another pair to help judge, or play it out once or
twice to get the values right.

[PlinytheWelder]

Yeah, I just picked up a Laney GH-100L for really cheap and after I do a cap job on it and possibly redo the grounding scheme, I'm going to mod it a bit just to see where I can take it...

[tubeswell]

... plate resistors in a stage are sometimes bypassed with several hundred to several thousand pF's.
Is there a rule of thumb to play this game?

Takes out HF oscillation in hi-gain amps. You get the same freq response as having a cap between plate and cathode, or between plate and ground, (because HT and ground are one and the same as far as AC is concerned). However if you have an anode bypass cap, the cap sees a smaller voltage differential than the other two methods, but it can also let HT noise through so you need a stable B+ (i.e. plenty of filtering/decoupling)

As long as the input impedance of the following stage is large compared to the output impedance of the stage you're talking about (i.e. as long as Zin is at least 5 x larger than Zout - Rule of Thumb), for a -3dB rolloff (which is about the minimum amount of vol reduction the human ear can detect/distinguish) the formula is:

pF = 1/(2.Pi x f x Zout), where: f = roll-off freq in Hz; and Zout = ohms[/img]

[benoit]

a -3dB rolloff (which is about the minimum amount of vol reduction the human ear can detect/distinguish)

Respectfully, -3dB is not the minimum audible difference detectable. It's certainly not a large audible difference but most people can easily hear 1 or 2 db changes unless they're in a fairly noisy environment. The -3db point is just the cutoff frequency of the filter circuit which is where the power is halved.

[tubeswell]

Well I find it hard to detect a 3dB change. Other people probably have better hearing than me.

[benoit]

Well I find it hard to detect a 3dB change. Other people probably have better hearing than me.

Could well be. The sensitivity depends on where in the audible band the dip is. We hear much better in the midrange. It's also further affected by how loud the sound source it is as our ears respond differently to higher SPLs, so maybe it was just circumstances. Anyhow the important part is the the filter info/math was solid and I was, as I said, splitting hairs.