Help troubleshooting - I'm out of ideas

[Dr Tony Balls]

Hi all. Working on a new amp and trying to implement a "one-tube reverb" a la the Hoffman thread provided below and i'm having problems. Hoping someone can sanity-check me and figure out what i'm missing.

https://el34world.com/Forum/index.php?topic=7957.0

The amp itself works and sounds great with the reverb tube (V2) removed, however with it inserted i'm getting significant 120hZ hum. Here's my key observations so far along with a schematic:

- The hum goes away by turning the reverb control
- The hum is present even with the reverb tank removed




This would limit the source of the hum to something happening between the reverb tank and the reverb control, i.e. the area around V2B. Here's what i've tried and noted so far:

- Grounding the grid of V2B does not eliminate the hum
- Replacing the cathode cap and/or resistor do not eliminate the hum (C15 & R31)
- Replacing the plate resistor and or coupling cap do not eliminate the hum (R32 & C16)
- Chopsticking wires does not eliminate or reduce the hum
- Replacing V2 does not eliminate or reduce the hum
- Disconnecting the grid connection to V2B (at the tube socket) DOES eliminated the hum


What am I missing here? Plate/Cathode voltages of V2B are 144 and 1.03 VDC, respectively. Below is a shot of the area, though its a tight build so maybe hard to discern whats going on:

[Stevem]

Your schematic is not showing the needed ground to the reverb pan, is this how you built it?

With these Fender type reverb set ups the input jack on the pan is grounded to the pan.
The output jack of the pan is grounded to the pan and the cable going from the amp to the pans output jack is grounded where the jack on the amp is bolted down at on the chassis.

The cable going from the amp, to the input jack on the pan had its shield lifted so that no ground loop is formed.

I also do not understand why that schematic shows, has a C12, as it’s not needed since C11 is already performing the needed DC plate voltage blocking function while coupling the audio to that tubes grid!

Also just in general you should not daisy chain any of the power supply filter grounds.
They should each have a separate run and go to ground near where the V+ From recto bridge gets grounded near the PT.

[Dr Tony Balls]

Your schematic is not showing the needed ground to the reverb pan, is this how you built it?

With these Fender type reverb set ups the input jack on the pan is grounded to the pan.
The output jack of the pan is grounded to the pan and the cable going from the amp to the pans output jack is grounded where the jack on the amp is bolted down at on the chassis.

The cable ( the send ) going from the amp, to the input jack on the pan had its shield lifted so that no ground loop is formed.

Sorry, no the connections to the pan are shielded and grounded. However as noted in the original post i'm getting hum even with the reverb pan removed and grounding the grid of V2B doesnt stop it. So its unrelated to the connections to the pan.

[Dr Tony Balls]

PROBLEM SOLVED: turns out the V2B cathode ground is too sensitive to be grounded on a bus-style ground scheme with the power supply. separating those cleaned it right up.

[Stevem]

Ok good, but C6 is not needed, so take it out!
The more caps you pump the audio thru the more phase shifting takes place which can lead to oscillating conditions!

[Bergheim]

Ok good, but C6 is not needed, so take it out!
The more caps you pump the audio thru the more phase shifting takes place which can lead to oscillating conditions!

Doesn't C6 keep the PI input from being referenced to ground via the 100k mixing resistor and 2M reverb pot? If so, removing that cap will mess up the PI bias.

[sluckey]

C6 is absolutely needed for the reason you state.

[Dr Tony Balls]

I believe that C11 is technically not needed (it wasnt initially there) though that does put voltage across the reverb pot.

[sluckey]

I believe that C11 is technically not needed (it wasnt initially there) though that does put voltage across the reverb pot.

C11 is needed to block dc plate voltage from getting to the reverb pot.

[Stevem]

If C6 is needed then why is such a cap not needed off of the input jack?

[sluckey]

If C6 is needed then why is such a cap not needed off of the input jack?

The grid of that preamp tube operates at zero volts so there's no need for a cap there. The grid of the PI tube operates at some considerable positive voltage. A typical Fender AB763 PI may have more than 70V present! The cap prevents any previous circuits from upsetting that bias voltage. For example, look at this PI...

http://sluckeyamps.com/phoenix/phoenix.pdf

Aiken has a good write up on the LTP PI. Might explain it.

[pompeiisneaks]

If C6 is needed then why is such a cap not needed off of the input jack?

Because there is no direct path for DC to get to the input grid of the first gain stage, as there's literally nothing but an instrument jack there. But since there are multiple anode connections linked to that point (through the previous gain stage and through the reverb return) and the PI cathodes, it's possible that DC can get to places it shouldn't.

As explained the dc on the PI cathodes can leak backwards to the other stages

Edit: Sluckey explained it better as I was typing but I figured mine would still be 'similar' in concept...

[Bergheim]

The grid of the PI input is referenced to the junction between the 470R bias resistor and the 22k tail resistor via the 1M grid leak. In a regular gain stage (common cathode gain stage), both the cathode and grid are referenced to ground via the grid leak and bias resistor, so one can look at ground in a regular gain stage as the bias/tail junction in the PI. Thus, no matter the tail voltage (in relationship to ground), the relationship between the cathode and grid voltage stays the same. That's the way I visualize it, anyway. In the PI, the tail voltage normally is 30-100 volts above ground. When the cathode of the PI is, say, 100V and the grid gets referenced to ground (0V) via resistors, odd things happen. I once wired a pre PI master volume after the PI input cap, and if memory serves me right it distorted/farted out badly even at barely audible volumes.

[Stevem]

Thanks for the correction info!