If caps are expensive, signals are small, and low B+ buzz is desired, you pick the dropping resistors to drop 20%-30% of raw B+.
The old Heath mono phono preamp started from 440V and dropped to 130V at the first stage. This allowed very low hum and a small total size (small 4-section cap).
You can figure "20%-30%" by assuming the tube-stage is a "resistor". We know the usual 12AX7 stage with 1.5K and 100K sets its plate at 70% of the stage B+. The 100K plus tube acts very much like a 333K resistor.
For 20%-30% drop into a 333K load, simple math will tell you to use a 83K to 142K dropper.
You can put two sequential stages on one dropper/filter. Then the resistor must be half value: 42K to 70K.
We have 4-stage amps. The later stages don't need full filtering. And three sequential stages on one filter may motorboat. So we have two filters. The one which passes current for four stages must be 1/4th value: 22K-33K.
When signals are high and you want large B+ to cover the signal, you drop 20%-10%, even 5% per section, and up-size the caps to kill the hum/buzz.
This leads to values near a dozen K. And because drop is low, the exact value is not critical. Obviously Fender used 10K and 4.7K kinda interchangeably. While 130V is different from 330V, 330V and 360V are practically the same.
For 12AX7-size tubes at resonable currents, several 20uFd-40uFd caps will clean-up even with "small" droppers (a few K). So we really don't worry about it.
It is possible to run more than 2 stages on one filter without motorboating. Especially with the short bass response favored for guitar. And many-stage gitar amps have loss networks between stages so gain does not build up too fast. If you are Peavey building a million amps, you do the hard math to feed a dozen sequential crunch-stages off one giant cap. In one-off DIY, that's too much brain-pain, you buy a pack of caps and be done with it.
Topic: Power supply resistors (Read 5762 times)