AC Voltage on the B+ rail

[hasserl]

I've got a BF Dual SHowman on the bench that came to me having not been turned on in years (decades?), with 4 mismatched power tubes and original electrolytics. It does have a 3 prong cord installed. Anyway, without even trying it as it was I went ahead and replaced all the filter, bias and bypass caps, then checked resistors and replaced a few of those that had drifted out of spec, and swapped out the power tubes with a matched quad. Upon firing it up it's got a bit of 120hz hum that is just more than it should have, enough to be bothersome to me. I found it has about 14 - 15 volts AC on the B+ rail, measuring right on the board where the line splits off, one going to the OT center tap and the other going to the choke. There is a good sawtooth pattern on my scope right there too. I suspected maybe one of the two first filter caps were bad, so I temporarily clipped on a couple of other caps with some jumper wires to run in parallel with the existing fresh caps, effectively doubling the capacitance, and the voltage reduced down to about 8 volts, along with a similar reduction in hum. Then I removed the fresh caps from under the dog house completely, well, I didn't actually remove them, I just lifted one leg of each one so they are out of the circuit, and just left the temporary caps in-circuit using jumper wires, and the voltage went back up to about 14.5vac, the sawtooth pattern and the hum returned to the previous level. So I think I eliminated some bad filter caps as the cause of the ripple. What do you guys think? What else is there that could be adding to the AC ripple?

One other thing, the original cap bleed off resistors were only 100k, so I replaced them with 220k as shown on the schematic. I don't think that would be the cause, but thought I would mention it just to include as much as I can here.

[Leo_Gnardo]

What else is there that could be adding to the AC ripple?

Poor connection of filter cap negative side to ground, same for power transformer HT winding center tap.

Open or not-well-soldered hi voltage rectifier? Granted, that would result in 60 Hz hum but worth a look.

[nickb]

The main filter cap is only 35uF in effect ( two 70's in series) and the four 6L6's present quite a load. Therefore, there will be some hum. When you say 14-15V, is that peak-peak or RMS?

[hasserl]

I clipped an extra ground wire onto the ground leg of the caps to one of the transformer screws, no change in anything. Voltage and hum are still present. The PT center tap is soldered to the chassis.

[hasserl]

The main filter cap is only 35uF in effect ( two 70's in series) and the four 6L6's present quite a load. Therefore, there will be some hum. When you say 14-15V, is that peak-peak or RMS?

That's measured with DMM set to the AC volts setting.

[nickb]

That's measured with DMM set to the AC volts setting.

Meters vary quite a bit in how well they measure RMS. Some simply have a DC rectifier to get the average and then scale that to get the RMS. This is fine when you have a sine wave but it fails for a sawtooth. What you measure is about what I'd expect for such a meter.

If the hum is too much for you, you can always put the extra caps back in.

[g1]

I'm wondering how much the mismatched power tubes are a factor, can you try a matched set?

[hasserl]

I'm wondering how much the mismatched power tubes are a factor, can you try a matched set?

It is a matched quad now. I never even fired it up with the other tubes in it.

[J M Fahey]

If old caps produced 14V ripple and new caps the same, then old caps are as good as new, at least as far as capacitance.

Plus 14V ripple sitting on ~450V +B is actually quite good.

Main suspect is mismatched power tubes and no, I don't trust vendor's words, add a 1 ohm resistor to each cathode and wiggle them around until both halves match idle current .

Also 35uF total capacitance is pathetic by today's standards, many use 2 x 220uF caps in series, do the same.

Then ground PI input grid (before the coupling capacitor feeding it audio signal) to the same point where the 100 ohms NFB resistor is grounded, and check how much ripple/hum is left at the speaker output jack, which is what matters.

You might try a couple PI tubes to check whether situation improves if you wish.

[Enzo]

Just pull the PI tube, does it still hum? if so, it is coming from the power tubes. If that kills the hum, it is entering the system at or before the PI.

I'd have to agree that 3% ripple is OK for B+.

Mismatched tubes will work fine, but they will not balance out and cancel hum.

[g1]

Ok, I did not read the first post clearly enough, he has installed matched output tubes and has been using them, so that is a red herring (unless they have become unmatched).
Sorry about the sidetrack.

swapped out the power tubes with a matched quad. Upon firing it up it's got a bit of 120hz hum that is just more than it should have

[Mick Bailey]

Sometimes tubes sold as matched aren't, and some amps do better on mismatched tubes shuffled to get the best symmetry according to the OT balance. Thinking mainly AC30s here, but I've encountered this with other amps. Ever stuck a matched set of tubes in a Fender with an output tube balance control and had to turn it right over away from centre to get the hum down?

[loudthud]

Did you clean and re-tension the power tube sockets? It won't matter how good the match if one socket is bad. Check the drop across the screen resistors as a good indicator of balance. I bought a Traynor amp once that hummed because one side of the PI tube was stone cold dead.