The dreaded hum

[soundguruman]

I recently ran into this where my g20 hummed with a buzzy sort of hum even with new filter caps etc. by buzzy hum I mean it had both 60hz and quite a bit of higher harmonics and noisy hash. Tube selection didn't yield anything.
Turning the volumes to zero or pulling the preamp tubes (one per channel) would make it quiet so I ruled out the PI and power tubes and focused on the heaters. I agree with soundguruman that the heater traces are too close to important stuff.


I tried elevating the heaters with a 9v battery and it went dead quiet. I made a quick divider off of the lowest dc rail that gave me about 100v of elevation and it did NOT help the hum or the hash at all that I could tell. I didn't really get this.
Empirically I found that a .001uf cap from each heater line to ground made it quiet again. I suspect the value of the cap isn't that important. However, the cap had to be installed down at the far end of the heater traces and not on the balance pot end.

I'm still pondering why just straight elevation had zero effect when the battery made it perfect. The dc rail is very clean on the scope. I also haven't got my head around why the caps solved it as the impedance is megs at 60hz. I would have expected the hash to go away but not the hum. Also, it's still not quite as clean as the battery but it's more than acceptable as the volume has to be cranked to hear anything at all. I will continue to ponder this.

When you elevate the heaters you still have (small amount of ) ripple /AC in the heater supply.
That ripple is entering the audio path, and is amplified.
You hear Buzz.

When you run the heater off pure DC, the ripple is absent. No Noise.

It's not "that much" of a mystery. The amplifier - amplifies!
If there is noise in, there is noise out.

Especially in a high gain amp, the layout is very important.
Many times you can create a layout which avoids AC ingress into the audio path.

But in several situations, the layout or design is not ideal.

More and more builders are realizing this.
More and more builders are attempting quieter amplifiers - with higher gain.
You are entering that realm of performance.
It's a good way to go. Good for you!

Now you start to understand why...
Many manufacturers of high gain amps are turning to DC heaters.
AND, that DC has to be "battery like." NO ripple at all.
Requiring large filter capacitors, or voltage regulation, to reject even the smallest amount of ripple.

It WAS the HI Fi Guys, like Jim McIntosh, who started us down that road, to DC heaters.
And they were right.
You CAN have high gain, with very low noise, TOO.

You don't have to put up with the noise.

[R.G.]

I recently ran into this where my g20 hummed with a buzzy sort of hum even with new filter caps etc. by buzzy hum I mean it had both 60hz and quite a bit of higher harmonics and noisy hash.

I don't know the answer to your exact question. I believe it may lie in the low internal impedance of the battery as opposed to the DC bias point. This is somewhat corroborated by the caps' low impedance at RF helping cure it. The notes about the caps curing it one place but not another along the heater supply line also indicates RF issues, as the inductance of the heater wiring would add/subtract from the caps' influence. It may be that putting the caps right on the heaters on the first preamp tube socket is the best place. However...

Your description is suspiciously like a problem with rectifier buzz and/or fluorescent light buzz. You can get otherwise-incurable rectifier buzz by simply not having the power transformer return wire from the ground side of the first filter cap in the wrong place. The PT wires to the first filter cap simply must be on a separate wire, not on the chassis, and not on any part of the power/signal grounding.

Another way you can get rectifier buzz is RF emissions from the rectifiers slamming off. I don't know the G20, but if it uses solid state rectifiers, they can turn off abruptly enough to make the wires connected to them radiate pulses of RF into other wires. If you're seening real 60Hz, not 120, this would indicate that one rectifier, not both, might be doing this. The caps cure also indicates you may be getting RF pickukp of similar pulses.

And then there's fluorescent buzz. Does it do this in other rooms/buildings that don't have fluorescent lights?