Humming tubes

[Lynxtrap]

I was recently trying to source a strong 60Hz hum in a Fender Bassman. I replaced the output tubes (one year old Sylvania 6L6GC) and put in some old Sovteks, and the hum was gone. What might be the reason? Had the tubes suddenly gone bad?

[Cliff Schecht]

In a push-pull stage (any balanced push-pull stage) there is a natural tendency for the stage to cancel common-mode noise (i.e. signals that are common to both the push and pull devices). 60/120 Hz ripple in the supply line gets cancelled by push-pull action, as does other noise to a lesser extent. We luck out with 60/120 Hz because at low frequencies common-mode rejection is high but this term degrades with frequency (meaning at higher frequencies the stage is not able to cancel the noise as well).

What you were dealing with is that the matching between the two tubes wasn't good enough. In a 6L6 amp I don't like to go much past 5mA of mismatch because going much past this will start allowing hum to creep into the output stage. This is one reason to match output tubes, otherwise you will never reap the full benefits of noise rejection in the output stage.

With that said, some mismatch in the output stage can be a good thing. Another thing push pull stages do well is kill second-order harmonics when things are well balanced. Adding in some intentional mismatch can actually tune this second harmonic cancellation to a certain extent. I personally do not like matched tubes in my amps that rely on the power stage for distortion (ala Trainwreck's, Marshall's and some Fender's) but this is a personal choice. It's probably good to just get a matched set and be done with it unless you have a lot of tubes sitting around that you can play with.

[Lynxtrap]

I don't think that was the case here, Cliff. This is a Silverface with bias balance, and that didn't help. And the hum was way louder than that of unmatched tubes, it was loud as a short circuit.
It also came suddenly: the amp was quiet, then suddenly it started humming. It went away when I pulled the PI tube, so at first I thought it was a problem with the PI. I went over the filters, resoldered almost the whole PI, checked everything.
Then I put in the old Sovteks - dead silent.

[martin manning]

If the hum was 60Hz, then it is likely that some internal failure allowed the AC heater supply to produce a large signal in the plate-cathode circuit on one side. Sudden tube failures (internal shorts, e.g.) are not uncommon, and I'd say you were lucky that there was no colateral damage.

[Lynxtrap]

If the hum was 60Hz, then it is likely that some internal failure allowed the AC heater supply to produce a large signal in the plate-cathode circuit on one side. Sudden tube failures (internal shorts, e.g.) are not uncommon, and I'd say you were lucky that there was no colateral damage.

Well, actually it was 50Hz since I live in Europe, when I think about it
Thanks, Martin. Heater supply seems logical if it was a failing tube. I was puzzled that it obviously was AC-hum, but never came to think of the heaters in the output tubes. Not even when I re-soldered the heater wires at the PI...
But how come there was no hum when only the output tubes were in (preamp and PI tubes pulled)?

[martin manning]

But how come there was no hum when only the output tubes were in (preamp and PI tubes pulled)?

That doesn't make sense, does it. I can't think of any power tube failure that would do that, unless you are dealing with an intermittent failure where disturbing the amp to pull the PI tube temporarily corrected it.