Biasing Basics...

[jordan86]

Sort of a "this is a football" moment....I'm entirely self taught on amp building. Well not self taught, but self motivated. I've learned a ton from online sources, amp books, and you fine people here. No EE training. Just a guitar loving tinkerer. I've built and modded enough to understand the main difference between cathode biased and fixed bias, and some of the general tonal differences associated with those. I understand the concept of a cathode resistor vs grounded cathode and where the "negative" voltage is applied in each scenario.

I know that it's often said that cathode biased amps "don't need rebiasing" when swapping tubes.  What I do not understand is WHY.  I know from experience that in fixed biased setups that different tubes can pull different amounts of current, thus requiring the rebias. Why do they NOT do that in a cathode biased scenario?  Or is that just a wives tale and they actually do?  I assume the concept is true, but I just don't understand why the positive voltage on the cathode resistor is so effective at self biasing. I'm more so asking in the concept of power tubes. I understand the majority of preamp tubes are self biased.

Not looking for a lengthy explanation if this has already been covered somewhere else. Happy to read for myself if someone just knows where to point me. I did try the search function FWIW, although maybe I could have been more diligent. Thanks!

[jordan86]

Just found this....under preamp section. Page 13

It is self adjusting. If the average current through the valve increases, so does the bias voltage, which will oppose the increase in anode current, and vice versa. This means that the bias will adjust itself naturally to changes in HT, valve aging, and variations in manufacturing tolerance. There is also less chance of the valve going into runaway and overdissipating in the event of some failure, which can be useful when dealing with power valves.
http://www.valvewizard.co.uk/Common_Gain_Stage.pdf

So in a cathode bias scenario as the current goes up the cathode voltage does as well, which would essentially "spontaneously" bias the tube colder, thus limiting total current? I guess self regulating would be the more appropriate term.

And it is assume that some reason this does not happen in a fixed bias scenario where the cathode is grounded, because it is...well...grounded.

[sluckey]

Different tubes of the same type do draw different amounts of current regardless of whether fixed bias or cathode biased. The reason that cathode biased tubes don't usually need rebiasing is because they are self biasing (another name for cathode biasing). The whole key is the voltage dropped across the cathode resistor. Well really, the key is the voltage difference between the grid and cathode, but since the grid will usually be at zero voltage, it' may be simpler to just think of the key as the cathode voltage. If a tube tries to draw more current (maybe because something in the circuit changed or could be something inside the tube changed, doesn't matter) the voltage drop across the cathode resistor will increase. This increase in cathode voltage will cause the tube to decrease conduction, lowering the current. The decreased current causes the cathode voltage to once again decrease, which tries to bias the tube hotter. This cycle continues until some happy median operation is obtained. This is the self biasing operation greatly simplified. Is this good enough?   

[sluckey]

Ah, see. You figured it out yourself. 

And it is assume that some reason this does not happen in a fixed bias scenario where the cathode is grounded, because it is...well...grounded.

Not only is the cathode held at a constant zero volts, the grid is held at some fixed negative voltage. There is nothing the tube can do to change that. So, fixed biasing is not self adjusting.

[shooter]

my2c;
always check/verify, easy to do, things drift, bypass caps get wonky

[PRR]

If the tube tries to suck double cathode current, the cathode resistor applies double grid voltage. That's powerful feedback. Typically a 2X tube will run near 1.4X current. Since same-type tubes are usually controlled to 20% at normal operating areas, variation is probably only 10% from one to the next.

Not only would fixed-bias allow that 20%: we normally do fix-bias to run cool at idle. But a small change from "cool" may result in un-cool operation. And no cathode resistor to discourage it.

[pdf64]

And if the valve gets too hot then its grids will tend to emit their own electrons - the cathode doesn’t have an exclusive deal on the thermionic emission action 
The positive grid current develops a positive voltage across the grid leak resistor, sends the bias positive, and in fixed bias, the valve can quickly spiral into self destruction.

Hence the grid leak resistance limits tend to be lower in fixed bias than cathode bias.

[mresistor]

my2c;
always check/verify, easy to do, things drift, bypass caps get wonky

You can say that again , found a shorted cathode bypass cap in a Peavey Classic twenty yesterday.