hum problem in new homebrew build

[AcePepper]

Working on a new homebrew amp. 30 watts, Class A-style, 4xEL84, cathode biased, GZ34 tube rectifier. Circuit based on Marshall 18 watt-style PI w/doubled power tubes. Using an AC30 tranny set. Preamp is modded 800-style.

I have power and preamp grounds seperated. "Power" grounds (B+ CT, plate and screen filters, power tube cathode resistors and filament resistors) grounded on power tranny bolt. "Preamp" grounds (PI, preamp cathodes, filters and pots) on a ground buss screwed down near input.

Got a 60hz hum. Low-level, does not go up/down/change with volume.

No center tap on filaments. 2x100 ohm resistors-to-ground instead. Have changed those. No effect.

Tried pulling tubes, one at a time. V1 pulled, no change. V2 pulled, no change.

V3 (PI/driver) pulled, HUM STOPS. This makes me think hum is centered around the PI tube/circuit. Yes?

Tried swapping tubes, no change, still hums.

Tried replacing all caps and resistors in PI/driver circuit, no change.

Tried swapping power tubes, no change.

Tried cleaning up lead dress in several places, no change.

Tried grounding filter cap for driver in a different place, no change.

Tried swapping OT primarys, no change.

Could this be filament hum?? If it's in the audio, where could it come from? Why wouldn't it stop when I pull V1 or V2 (preamp tubes)?

Running out of ideas to try. Please help!! Suggestions??

[AcePepper]

I should probably mention, all voltages look normal and the amp is passing signal and functional, aside from the aforementioned hum.

[km6xz]

Are you sure it is 60hz and not 120hz? If it is 60 it would be induced from portions of the circuit that are using AC, such as heaters or magnetic coupling. Are the power supplies for preamp and power amp separate or are you creating two different return paths for the same common ground. Differences in ground potential between the two ground paths will cause hum, but it ought to be 120hz if power supply ground currents and be controlled by the gain settings. Other options are transformer generated ground currents if there is no air gap.
It really is not class A, but a hot AB so there would be cancellation of common mode hum induced into both sides of a push pull topography, unless there is an imbalance in the transformer or tubes. or their drive from an unbalanced PI.
Does the hum lower when increasing the bias(decrease cathode current)?

[Enzo]

The PI is a small tube, and many people call all the small tubes "preamp" tubes. But the PI is part of the power amp, try moving the PI ground over to the power amp ground point.

Stan's point is VERY important. 60Hz and 120Hz are the same note, just an octave apart, and since they won;t be sine waves in the system, the overtones can make it hard to determine which. But the two come from totally different sources, and each source of hum has its own cure.

[AcePepper]

I know it's not really "Class A". I just put that for reference, as in "like an AC30 or Marshall 18 watt". I should have just said "cathode biased." My bad.

Pretty sure it's 60hz hum, not 120hz. Hum is constant and not controlled by gain or master vol. Hums when all volumes are down.

Each pair of power tubes (inside pair and outside pair) has it's own cathode resistor. Outside pair is wired to a "1/2 power" switch that disconnects the cathodes for that pair. Each pair of tubes has a 150 ohm cathode resistor w/50uf bypass cap. Had 2 x 120 ohm resistors, changed to 2 x 150 ohm to decrease cathode current/be easier on the tubes. No change in hum.

Hum does go away pretty instantly when B+ is turned off by standby switch.

As stated, I do have two ground points (plus an A/C ground near the power cord socket). Power section grounds are on power tranny bolt. Preamp grounds are on a bus on the turretboard and grounded on a screw near the input. Have tried moving preamp ground to same place as power ground (a la KOC's "galactic grounding"). Makes no difference. No change in hum.

As far as I can tell, PI is extremely well balanced. Plate voltage on PI plates are within 1 volt of each other.

My PI is is like that from a Marshall 18 watt. Same part values. No negative feedback.

Hope this helps. Any more ideas??

[AcePepper]

Enzo,

Thanks for the tip! I've built lots of other amps where the PI is grouped/grounded with the preamp and never had this problem before. I'll try seperating the PI grounds and moving them to the power ground point. I have tried moving the ground for the PI's filter cap to the power ground point. Made no difference to the hum.

The sound is more low-frequency "hum" than "buzz". That makes me think it's 60hz hum. However I'll try any ideas for 60hz or 120hz hum.

[Enzo]

120Hz is also a low note. I always check with my scope, regardless of what my ears think. If I touch my probe tip, that puts 60Hz on the screen - my body picks up the 60Hz fields in the area. I then move the scope probe to view the hum signal. If it has the same number of humps on the scope, it is 60Hz, but if it has twice as many, then it is 120Hz.

Hum can come from radiated fields picked up by the circuits, it can come from poor grounds, and it can come from power supply ripple. If your guuitar cord was crappy or the input jack messed up and causing hum, when you switch off the power, the sound fades out over a second or two, and the hum would as well. What does stop instantly when power is removed is the AC power to the system. The ripple curents instantly disappear. So in the couple seconds as the amp fades, you can hear the hum disappear, while your guitar sound might linger and fade.

That leads me to think maybe the ground path for the PI is shared by some ripple currents.

[km6xz]

If the hum did not change in amplitude when changing the cathode current, that eliminates about 3/4 of the circuit. If it indeed 60hz, look for heater or magnetic induced fields from the transformer. Does the transformer buzz or hum audibly, and how far apart are the OP PT, and what orientation are the laminations in relation to each other.
One thing to consider is the transformer and whether it is an actual AC style OP transformer. It is not the same as a transformer that is used on conventional AB amps. An air gap is larger in the laminations of a Class A OP(so the higher idle DC current flowing in a quasi-Class A OP transformer primary would otherwise increase the risks of saturation) which causes the inductive reactance to lower, requiring higher turn counts, for a design goal inductance. That can make the transformer more susceptible to induced magnetic fields from the PT unless care is taken to orient the laminations for minimum coupling between transformers. If the core is saturating due to not having the gap, like a conventional 4 EL84 AB amp transformer used in a high idle current application, the inductance drops a great deal resulting in very high current being pulled from the supply when the core is saturated. That would generate 120hz hum from the reservoir caps just not having enough reserve storage. But if there was a magnetic induced hum, due to coupling between transformers, it would be 60hz.
Try unbolting the power transformer and lifting it clear of chassis contact. Does that change the hum level? If so, try rotating it to see if there is a null point where the hum drops a lot.

Do you have a sensitive voltmeter that you could measure drop between the grounds. Any voltage difference between the input ground and other parts of the chassis would be seen as a signal in a single ended signal path just as surely as if it was coming from the guitar cable center conductor. You mention the PI is balanced DC but that does not tell us anything about the balance of signal that is capacitively coupled or when signal levels increase over idle. It is possible to have hum signal enter that way at the PI grids, on one side so the normal cancellation of a push-pull circuit of common mode noise would not be in effect.

What test instruments do you have? A good dual trace scope with channel summing and one channel phase invert? If it does have that, you can use it as a differential balanced input not referenced to ground so you can track down ground current noise and imbalances.

[Bruce / Mission Amps]

Here's a quick way to test for 60Hz or 120Hz hum...

Unsolder one of the 6.3v leads from the tranny to the first place it goes.... That cuts ALL the filament supply to the tubes.
Now use a jumper clip lead to reconnect it.
Start up the amp and let it get hot.
Turn the amp up so you can hear the hum.... reach over and disconnect your clip lead.
The tubes are still hot and will continue to conduct for a few moments...
but if the hum stops instantly... you have a 60Hz hum.
If the hum does not stop... yup, 120Hz hum from the B+ rail or a little ground loop somewhere, such as the negative end of the filter cap of the phase inverter section with respect to the audio ground point of the same area.

[eschertron]

I'm pretty it's covered elsewhere in the forum; I learned to test for ripple-based hum by clipping another cap in parallel with the first cap in the power supply. The bigger the cap, the more change you'll have in the ripple.

[AcePepper]

Back at it now. I tried Bruce's idea about the filaments. Hums stays after filaments are disconnected. So it seems to be a 120hz hum.

Last night I tried replacing all filter caps except the main 30+30uf cap on the plates and screens. No effect, still hums.

Gonna try seperating the PI grounds and move those to power ground.

As for trannys, power and output trannys are several inches apart, mounted 90 degrees to each other. Choke is in between PT and OT, facing same as PT.

Will try to post a gut shot pic later.

Any other ideas?

[loudthud]

Not sure exactly what PI you are using, (and I'm not going to go hunting around for it on some website I'm not familiar with) but if it's anything like the AC30, there are two inputs. Since you have no feedback, there must be one input that is more or less grounded. That ground needs to come from the quiet ground in the preamp, not from the noisey power amp ground.

[AcePepper]

PI is a long-tail type. 2nd input is grounded. All PI grounds are grouped on a ground bus with the preamp grounds and grounded near the input. Currently using Marshall 18 watt PI values. May try the AC30 PI, values are a bit different.

[AcePepper]

OK,

Tried moving PI ground to power ground. No effect, still hums.

Tried AC30 PI values. No effect.

Tried replacing PI tube socket. No effect.

This thing is kicking my ass. Doesn't seem to matter where what's grounded, hum is all the same.

Will try to post pics tomorrow. Any more suggestions??

[acorkos]

I built an ac30-style amp a few years ago. I wanted it in a compact chassis. Ended up with a hum problem like yours...and I tried about everything you did. The only way I got rid of the hum was to remount everything on a much-longer high-power tweed twin chassis. I think it was the physical distance between power and output transformers that was the issue. During troubleshooting, I disconnected the OT, pulled it off the compact chassis, reconnected w/alligator-clip leads and started rotating it...but only after moving it away from the power transformer did the hum decrease noticeably. So I bought the larger chassis.