If the solder joint on the part was noisy and it DOES NOT make a LOUD noise at the speakers, the part IS NOT in the main signal path.

If the solder joint on the part was noisy and it DOES make a LOUD noise at the speakers, the part IS in the main signal path.

If you double the value of the part and the volume or frequency response does not change, the part IS NOT in the main signal path.

If you double the value of the part and the volume or frequency response does change, the part IS in the main signal path.

1. The inductors and resistors between filter sections in a good B+ design are not in the main signal path. The diodes, HV power transformer and all but the last B+ capacitor are also not in the main  signal path.They still can impact the sound, but not directly.

The last B+ capacitor usually is in the signal path. The last B+ capacitor is the "single cap" or "bank of caps" that makes up this last capacitor value.

A good B+ design has multiple filter stages. The first stage of the B+ is the one that attaches to the transformer. The last stage of the B+ is the one that attaches to the tube.

Minor amounts of distortion and noise in the first stages of the B+ (or B-) don't have a big effect on the signal. The job of these parts is to provide power to the circuit and to keep current ripple and transformer noise off the last B+ capacitor.

Remember, you buy B+ parts for safety and ruggedness as much as circuit performance.

* Use flameproof metal oxide or flame proof wire wound resistors in the B+.
* Use capacitors with margin on their voltage and current ratings. The cap's rated voltage should be at least what the power supply will rise to if the power supply does not have a load on it with the 115V  at 130 VAC.

If you use a 1n4007 in the power supply instead of an ultrafast like the UF1007 (or better), the first part of the B+ chain is NOT doing its job right because it is making noise that gets into the signal path. Or it could mean that the ringing on the B+ is getting into the filaments and from the filaments into the cathodes of the tubes.

If you use an audiograde film resistor between sections of a CRC B+ supply you are wasting your money. Film resistors do not handle surges well and will fail if used in the B+ of a supply.

I hate simplifying things this far but: If you can short the output of the B+ to ground for 3 seconds (don't try this, just work the math out) with a film resistor in the B+ and nothing makes smoke or fuses, the film resistor is properly rated for power supply use. 99% of even oversized film resistors made will fuse if this happens. Most reasonably sized metal oxides and wire wounds will just get real hot.

1. All connections and wires (signal and return) to the volume control (attenuator) are critical in the signal path. If you have a filter cap attached to the attenuator ground, all of the filter cap's current noise will be injected into the attenuator through the resistance and inductance of the attenuator ground wire!

2. The bias set resistors in a tube stage are critical. If they get hot, their value will change. The the resistor's value changes, it makes distortion.

2. Any feedback / gain set resistors in a tube stage are critical. If they get hot, their value will change and thus the "volume" (amplitude) of the signal will change. Meaning  if the resistor's value changes, it makes distortion. These parts should be heavily derated for power and possibly picked for <= 25ppm/C temperature coefficients.