Non top boost AC-30 in for repair.

[Mark]

Yes the hum then stops. If I short the 1M grid leak resistors in the PI out, this stops the hum too.

[rooster]

This is a great read, guys. Mark, going back to your volume pot values, I would imagine the 500K is a much better idea than what you have there. My build has the TB, the Normal, and EF86 channel going on. As to the volumes of the three channels, the TB channel will knock you over at 12 noon, the Normal quite livable, the EF86 is very much on a parr with the Normal. However, (all my volume pots read 500K), when you move these two volume controls of the Normal and EF86 channels into the 2 o'clock range and higher, then the output increases considerably.

Here's my thinking, all the pots in your amp started out at 500K (or should have), and so the carbon traces have been thinned out over time with cleaning and use. Do you think that Dick Denny arbitrarily used 500K pots instead of 1M? And there is also the strong possibility that the traces did not wear out evenly. Your 12 o'clock on one pot could be 400K on one and 100K on the next.

[Mark]

Rooster thanks for your reply. I have since changed the volume pots. I think the amp sounded a little better for changing the volume pots of the brilliant and normal channels. I changed the ceramic bright cap to a silver mica and I found the brilliant channel sounded harsh so I used the original ceramic cap.

[pdf64]

It’s pretty common for the tracks of original pots in 60s British amps to measure far above their nominal, eg up to twice nominal.
I think it being due to wear is unlikely, as such large changes would greatly affect the taper, which doesn’t appear to be the case. eg log pots are still around 10% of the total track resistance at 50% rotation, linear pots measure around 50%, regardless of the track resistance being so much higher than nominal.

[Roe]

It’s pretty common for the tracks of original pots in 60s British amps to measure far above their nominal, eg up to twice nominal.
I think it being due to wear is unlikely, as such large changes would greatly affect the taper, which doesn’t appear to be the case. eg log pots are still around 10% of the total track resistance at 50% rotation, linear pots measure around 50%, regardless of the track resistance being so much higher than nominal.

Yes around 30%+ is typical on plexis, although it varies greatly

[Mark]

I have come to the conclusion that the earthing scheme on a vintage AC-30 is pretty awful. It’s a pity someone has come up with a mod for that.

[pdf64]

I guess that as designed, with the heater circuit wired single ended, due to the resulting 50Hz buzz, 100Hz loop hum wasn’t apparent

[Mark]

The heaters have two 100 ohm resistors to quieten the heater noise.

The noise emanates from the PI stage.

[Colossal]

The heaters have two 100 ohm resistors to quieten the heater noise.

The noise emanates from the PI stage.

Mark,

I can't recall without re-reading the thread, but did you try adding input grid resistors on both the inverting and non-inverting sides of the PI? I read Pete's reply about hum just being a consequence of the relatively high gain design on the AC30.

On a separate, but sort of related idea, I have always noted that the Marshall 1974/Wattkins/Lite IIb circuits used 470k/470k for the PI input grid leaks, reducing the overall input impedance to 940k with a slightly more centered bias of 820R, where the Vox is 1M/1M with the slightly colder 1k2 bias, which can give some of the shattered glass high end when pushed hard. Now that is not really 50Hz related, but it is interesting to note the differences in an otherwise fairly similar design. I have built Lite IIbs that were just as quiet as a church mouse when idling dimed. Long time ago, I built a Rocket type amp and ran into the same vexing PI hum that was only cured by splitting the ground bus into preamp (grounded at the input socket) and high ripple current power amp grounds.

[Mark]

Thanks for your reply Colossal. The AC-30 doesn’t have grid stoppers on it. I have built an 18 watt amp and I’m
aware of the different P.I. stage.

I did have this sort of issue when building my Rocket amp, but it turned out to be a conductive circuit card. Once the card was replaced with a Garolite card, the amp was very quiet. As you’d imagine, I have tested for a conductive card and found it to be okay.

[Mark]

The latest discovery on the AC-30 with channel bleed issue was the bleed could be eliminated by increasing the 22uF shared common cathode cap to 100uF. Perhaps this is the reason the early Tweed and Marshall amps used the 250uF common cathode cap?

I resoldered and cleaned every joint and terminal in the amp and it made no difference whatsoever.

[Colossal]

The latest discovery on the AC-30 with channel bleed issue was the bleed could be eliminated by increasing the 22uF shared common cathode cap to 100uF. Perhaps this is the reason the early Tweed and Marshall amps used the 250uF common cathode cap?

I resoldered and cleaned every joint and terminal in the amp and it made no difference whatsoever.

Mark,

That's an interesting observation, about the possible use of the big 250uF on the Tweed and Plexi amps. I was going to ask if you'd tried splitting the cathodes...giving each it's own cathode resistor and bypass cap.

[Mark]

I tried that too and it stopped the signal bleeding into the other channel. It was similar to a Treble model except I used 22uF for both channels.

[Reeltarded]

at zero volume there is bleed on the two ac30s i normally play through

the sound in your demonstration is not a lot different than what i expect (i used to think it was pots not fully closing at zero rotation) until you crank the vibrato. eeek

I like the idea of snubbing the PI a little and I did the same when servicing the 64 with added TB and that amp is particularly terrifying with attitude.. face stayed the same but no longer picked up Tokyo Rose if using that stupid Super Hard-On pedal with the amp dimed and the guitar at zero..

That RF is part of a couple recordings we did in the early 90s. It will be missed.

[Mark]

Thanks for your reply, what do you mean by snubbing the P.I. exactly?