Shared cathode bypass

[Randall]

What's your take on shared cathode bypass cap and resistor, like how Fender ties the second triode of each channel in the blackface stuff? I just read on an "amp guru" blog that it is better to separate them?

[bob p]

from 17 years ago:

cathode caps and resistors - AMPAGE Archive

Looking at that thread, I guess we need to add Dutch to the MIA list.

[Justin Thomas]

That's a "Like" for missing Dutch.
As for the topic at hand, I don't know of any technical reasons, but I tend to separate them for tone options... more of a problem to me are the dual caps with a common ground that Fender occasionally used...

Justin

[Randall]

I read thru that very long and sometimes confusing thread. I get that the cathode resistor sets bias and gain, and that 820 ohm stands in for two 1500 ohms in parallel. I also get that the bypass cap is a high pass filter, almost get it. Where I have a hole in my understanding is how this works in a cathode.

The cathode is pulling electrons from ground thru the resistor, and are modulated by the grid as they are being attracted by the anode and sent on their way as an amplified representation of the grid signal. Correct? I have never thought of the electrons coming from. Ground as anything other than a raw supply source, with no frequency. Like water into a hose before the sprinkler or shower head makes it into a spray or a stream or a pulsating pattern.

help me out here.

[Gingertube]

Cathode Bypass caps.
At high frequencies the bias resistor is fully bypassed (low capacitor impedance) and you get maximum gain from the stage.
At low frequencies (depending on cap size) the bias resistor is not bypassed (high cap impedance) and so you get a different lower gain (because of the signal feedback developed across that resistor when not bypassed) .
They call this a "shelving" response because it has fixed max gain amd fixed mionimum gain with a frequency dependent slope between them and we usually talk about the 1/2 (way between the two gains) response point.

A single shared bias resistor on 2 triodes has a potential problem. If one triode conducts more idle current the other tends to give up idle current by approx. that same ammount.
Separate resistors allow you to make sure each triode runs at the design idle current.
Separate resistors also has the benefit that you can tailor low frequency response in each by choosing the bypass cap for each.

Pretty picture showing response for a 12AX7 stage with 1K5 bias resistor and various bypass cap values.
The Valve Wizard -Cathode Follower

In general when you split a single R and C into separate ones:
1) You a halving the idle current in each of the new resistors so for same bias violtage double the resistor.
2) With double the resistor then you might think that you just need to 1/2 the capacitor for the same 1/2 response frequency point. This is a good start point but it is not the entire story. This would be the entire story if it were just the R and c in parallel you are interested in but it is not.
3) It is actually the R the C AND the impedance to signal ground looking into the cathode of the tube which are all in parallel. There is a bit of algebra you can do to work out that impedance to 0V looking into the cathode but most don't do it. WE just say it is roughly equal to 1/gm of the triode, for a 12AX7 triode with a gm of 1.6mA/V that gives 625 Ohms.
4) so it is 1K5 parallel 625 Ohms = approx 440 Ohms which is in parallel with the cap.
5) with the common 820 R you have 2 lots of the impedance looking into the cathodes so the effective impedance is 820 parallel 625 Ohms parallel 625 Ohms = 226 Ohms.
6) that shows that it is the impedance looking into the cathode which dominates (rather than the bias resistor value) and just halveing the cap will most always be good enough (440 Ohms effective rather than 226 Ohms effective across the cap).

One of those things which we "just do" knowing it works but having forgotten why. Just halve the cap value with confidence.

Hope this is of help.

Cheers,
Ian

[joe_bfstplk]

Zombie thread alert - heck, zombie USER alert

from 17 years ago:

cathode caps and resistors - AMPAGE Archive

Looking at that thread, I guess we need to add Dutch to the MIA list.

Dutch here, have been elsewhere for a bit....

[Justin Thomas]

I didn't read the thread, but good to see some old familiars... Back when I'd spend entire nights in the UVA computer labs learning what meager bit I could...

Justin

[SoulFetish]

Now we need to add Bob P to the MIA list. Hope he's okay

[the fatch]

I read thru that very long and sometimes confusing thread. I get that the cathode resistor sets bias and gain, and that 820 ohm stands in for two 1500 ohms in parallel. I also get that the bypass cap is a high pass filter, almost get it. Where I have a hole in my understanding is how this works in a cathode.

The cathode is pulling electrons from ground thru the resistor, and are modulated by the grid as they are being attracted by the anode and sent on their way as an amplified representation of the grid signal. Correct? I have never thought of the electrons coming from. Ground as anything other than a raw supply source, with no frequency. Like water into a hose before the sprinkler or shower head makes it into a spray or a stream or a pulsating pattern.

help me out here.

Electrons are flowing from the cathode to the plate and onward through the plate resistor to the B+ line. The electrons leaving the cathode are being replaced by electrons flowing from ground through the cathode resistor. This stream of electrons is a current, and since the tube and the cathode resistor are in series, the same current must be flowing in each of them. According to Ohm's Law, if a current flows in a resistor, a voltage must be developed across it. If one end of the resistor is connected to ground, the cathode must be at a voltage above ground equal to the voltage developed across the resistor. An AC signal at the grid causes the current in the tube to vary, which means that the current in the cathode resistor must also vary, which in turn means the voltage across the resistor, i.e. the voltage at the cathode, must vary. This varying voltage at the cathode is a 'negative feedback' signal which acts in opposition to the signal at the grid, reducing the gain of the stage. A bypass capacitor removes the feedback signal by conducting (shunting) it to ground allowing the stage to attain maximum gain.