Help with AC noise issue on 70s SF Twin Reverb

[glebert]

I'm working on resurrecting a 100 watt Master Volume SF Twin from the dead. Master volume control and boost removed, putting it as close to AB763 as I can. I have the Vibrato channel mostly working (reverb working but no tremolo yet) but it has a pretty bad hum (pretty sure 120Hz but haven't scoped). No hum with the PI pulled, so don't think it is tube mismatch (only running two 6L6s currently). If I roll off the mid control and put the volume to a particular spot the hum is gone so presumably it is hitting all the stages with varying amplitude and I can get them to cancel out.

Caps are fairly new, of course could have a bad one. One other weird thing (to me, at least) is that when I measure the AC ripple before the standby switch the AC gets significantly worse when the switch is ON. It is less than 100mVAC when in standby and 3.4VAC when taken out of standby. In that condition the voltage at the B node is maybe 80mVAC, the C node is about 50mVAC, haven't measured beyond that. My question is if it is normal for the AC ripple to get considerably worse when there is a higher current load on the PS? If not, what could cause it? For some reason I am wondering about the six diodes in the full wave rectifier, wondering if a problem in there could cause more ripple when the PS needs to deliver more current.

TIA,
Greg

Basically this schematic except the main caps are 100uF instead of 70uF.
https://ampwares.com/schematics/twin_reverb_ab763.pdf

[catalin gramada]

The ripple get significantly big in respect with current draw. Two 1000v series diodes per side should be pretty safe for modern ones. Max reverse voltage should be minimum 2x so two series is plenty.Check the spec of diodes you have. I.ll check if voltage is equal over stacked reservoir caps.

[glebert]

Lee's ripple with the load ? I.ll check if the voltage is equal over the reservoir stacked caps.

More ripple under load (to me switch is ON=out of standby, but that is confusing too).

[catalin gramada]

More ripple under load (to me switch is ON=out of standby, but that is confusing too).

I get 9.5VAC ripple in mine with a 50uF cap (two 100uF stacked).and 100mVAC when off. and get no substantial hum as time the power stage is decent balanced.
More ac ripple current meant shorten the capacitors life so there is a reason for big sized cans. As general rule the cap current rated ripple should not be lower than current the circuit draw (twice is better).

[pdf64]

Those amps rely on the chassis to conduct 0V common current, via dozens of connection points. Build up of surface oxide corrosion etc over time degrades those connections.
It might be beneficial to remove every relevant pot switch, socket and the brass plate, and clean up the chassis and plate around the holes, so as to reestablish a good electrical connection.

[Helmholtz]

One other weird thing (to me, at least) is that when I measure the AC ripple before the standby switch the AC gets significantly worse when the switch is ON. It is less than 100mVAC when in standby and 3.4VAC when taken out of standby. In that condition the voltage at the B node is maybe 80mVAC, the C node is about 50mVAC, haven't measured beyond that. My question is if it is normal for the AC ripple to get considerably worse when there is a higher current load on the PS? If not, what could cause it?

Absolutely normal.
Theory says without any load current ripple must be zero because nothing is discharging the caps between the voltage peaks.

Ripple voltages look ok except node C, which I would expect to show lower ripple.
Maybe tack another cap here for a test.

[glebert]

Absolutely normal.
Theory says without any load current ripple must be zero because nothing is discharging the caps between the voltage peaks.

Ripple voltages look ok except node C, which I would expect to show lower ripple.
Maybe tack another cap here for a test.

Makes sense, for whatever reason I tend to discount that current has to be coming from the capacitors between peaks.

Thanks to all for tips to try.

[glebert]

OK, think I might have this figured out. Turns out there is a difference between the AB763 and the 100W Master Volume amps that I hadn't noticed.

Mast Vol


AB763


My amp did not have the 100 ohm resistors to ground (Did I mention I'm the 4th tech to work on this since the last time it worked?). As a test I alligator clipped a 100 ohm from one side of the 6.3V to ground and the hum goes almost completely away. Go figure!

After I put these resistors in, time to move on to the tremolo. Wish me luck!

[Tom Phillips]

...My amp did not have the 100 ohm resistors to ground (Did I mention I'm the 4th tech to work on this since the last time it worked?). As a test I alligator clipped a 100 ohm from one side of the 6.3V to ground and the hum goes almost completely away. Go figure! ...After I put these resistors in, time to move on to the tremolo...

Nice that you discovered the missing heater hum balance resistors. I'm sure we could generate a very long thread listing faults missed by (or caused by) past techs (or wanna be amp techs). Sometimes it turns out to be an original factory mistake. The really sad cases result in expensive repairs and sometimes permanent damage to the amp.