50Hz Hum in Plexi Build

[shocki]

When I first tried my plexi with 50+50 Filter caps I had almost no hum. Now I switched to 32+32 and like the feeling with those caps a lot better. But I do get a loud 50Hz hum (at least it sounds like 50Hz)
This hum is there with no guitar plugged in, no PI tube and does not get louder when I turn up the volume. So in a band situation you wouldn't really notice. But still it is not very nice.

I tried to move wires around and tried different grounding points. But nothing helps. If I don't find a solution I have to go back to 50+50 filtering I think. Maybe somebody has an idea how to solve this. I know the earlier Marshalls all used 32u+32u caps or even lower. So this should actually work. Also could be the caps are bad (they are brand new). Is there a way to measure this?

I used this layout. http://www.ceriatone.com/images/layo...exi50-Lead.jpg

[jpfamps]

I assume you are getting 100 Hz hum, not 50 Hz hum.

32uF/32uF should be quite enough at idle.

It's not unknown for new caps to fail. I usually find looking at the power supply on a 'scope the best way of identifying problems in circuit.

Duff caps can be identified by a capacitance meter, however I've often found caps that test fine on the meter, but clearly aren't OK at working voltages.

[shocki]

Don't have a scope ;( Just measured the caps with my meter. They seem to be OK. BTW I also hear this same hum frequency when turning on the amp. But very very soft. I hear the sound the transformer actually makes. I agree it sounds more of a 100Hz than 50Hz hum.

[shocki]

OK I found it. Somehow the bias was totally of spec about 120ma. No idea how this happend. No it is back to 35ma and the hum is almost gone. I think it could even be better with the 50+50 filtering. There is only a little bit of hum left. But I think without a scope it is really hard to trace.

[Gingertube]

You are probably now at the best you can do without further mods.

The hum could probably be reduced by using a better matched pair of output tubes.
It can definitely be reduced by adding a bias balance control or individual bias adjustments for each output tube.

If the idle currents through the 2 halves of the primary are identical then the 100Hz (120Hz in the US) residual ripple from each side will cancel (since they are out of phase). AS the idle currents become less balanced the hum will get louder.
The Fender style bias balance control is quite good as you can set it by ear. Turn the master volume to 0 and then set balance control for minimum hum. This actually accommodates minor magnetic balance imperfections in the OT primary halves too.

I will not build an amp without either an bias balance control or individual bias adjustments for each output tube, and when I see a design that does not do either I conclude that it is designed down to a price and not up to a standard.

Cheers,
Ian

[shocki]

You are probably now at the best you can do without further mods.

The hum could probably be reduced by using a better matched pair of output tubes.
It can definitely be reduced by adding a bias balance control or individual bias adjustments for each output tube.

The current for both tubes is like 5mA appart. So one tube is biased at around 32mA the other one around 38mA. Maybe matched tubes would solve the issue you are describing.

[soundguruman]

When I first tried my plexi with 50+50 Filter caps I had almost no hum. Now I switched to 32+32 and like the feeling with those caps a lot better. But I do get a loud 50Hz hum (at least it sounds like 50Hz)
This hum is there with no guitar plugged in, no PI tube and does not get louder when I turn up the volume. So in a band situation you wouldn't really notice. But still it is not very nice.

I tried to move wires around and tried different grounding points. But nothing helps. If I don't find a solution I have to go back to 50+50 filtering I think. Maybe somebody has an idea how to solve this. I know the earlier Marshalls all used 32u+32u caps or even lower. So this should actually work. Also could be the caps are bad (they are brand new). Is there a way to measure this?

I used this layout. http://www.ceriatone.com/images/layo...exi50-Lead.jpg

[ATTACH=CONFIG]22312[/ATTACH]

A. Move the bias supply filter ground over to the PT side of the chassis.
B. Move the preamp ground over to the input jack side of the chassis.
C. If there is less hum with 50 / 50 then USE 50 /50! ( I would for sure)
D. Oscillation in the preamp will cause hum in the output.
E. Imbalanced output tubes will cause HUM.
F. Spiral filament preamp tube in cathode follower will cause hum.
G. NOW LOOK! The design of the OLD Plexi has NO screen grid resistors. Check your schematic. The screen grid resistors MUST be added.

[shocki]

A. Move the bias supply filter ground over to the PT side of the chassis.
B. Move the preamp ground over to the input jack side of the chassis.
C. If there is less hum with 50 / 50 then USE 50 /50! ( I would for sure)
D. Oscillation in the preamp will cause hum in the output.
E. Imbalanced output tubes will cause HUM.
F. Spiral filament preamp tube in cathode follower will cause hum.
G. NOW LOOK! The design of the OLD Plexi has NO screen grid resistors. Check your schematic. The screen grid resistors MUST be added.

A. This is not really easy to do now ;(
B. Do you mean I should remove the ground bus across the pots?
C. I think you might be right
D. I have the hum also with no preamp tubes and no PI
E. You are right. I will try to get matched ones or change bias to dual bias.
F. This is also concerning the preamp section I guess. I am not 100% shure what you mean.
G. I used 1K screen grid resistors. A little bit hard to see.

[Steve Conner]

I don't see any point in doing A or B. Use some logic: The amp didn't hum with 50uF, but now it hums with 32? Then the 32uF capacitors are the cause of the hum. It's the price you need to pay for the dynamics you wanted. The test with the PI tube removed ruled out just about all grounding and oscillation issues.

Better matched tubes, a bias balance control, or colder bias will all help.

Re point F: Some Sovtek tubes had a low heater-to-cathode voltage rating and would break down when used in a cathode follower. I think this is what SGM refers to. I was under the impression that they didn't just hum, they broke down and shorted out.

[shocki]

I don't see any point in doing A or B. Use some logic: The amp didn't hum with 50uF, but now it hums with 32? Then the 32uF capacitors are the cause of the hum. It's the price you need to pay for the dynamics you wanted. The test with the PI tube removed ruled out just about all grounding and oscillation issues.

Better matched tubes, a bias balance control, or colder bias will all help.

Agree with this. I think changing the the caps at the preamp to 32+32, Powertubes and PI to 32+32 and leave the first cap to 50+50 would be good compromise.

[shocki]

BTW. The amp has absolutly no hum now. I had the amp set at triode setting. This way I had this slight hum left. In Pentode mode there is absolutly no hum at all.

[Gingertube]

Yes,
Matched tubes would help. The Mtched pairs you buy these days are often not as matched as we would like but are probably better than 5mA difference.

Cheers,
Ian