Tube Tech MP1A weirdness

On my bench I have a Tube tech MP1A for service and it's acting weird. When I flip the 48V switch on channel 1 a few times it starts humming like hell. Flipping the pad-switch doesn't reduce the noise. The noise does go away if I plug something into the DI input. The humming stops if I flip the switch a few times again but this is pretty random and sometimes takes a while.
I also checked the PSU voltages and they seem (too) low. For 240V I measure 220V and 270V -> 227V. 35V is 30V but phantom 47,6V.
Channel 2 is still working ok.

Any ideas where to look? I've already changed the tubes. PSU recap maybe? Input transformer?

cheers,
Corneel

The DI input separates the input transformer from the rest of the circuit. So if this eliminates the hum, it must be coming from the circuit around the input transformer.
Strange, because even if there would be a ripple on the phantom power voltage, this should be eliminated by balanced input.

By the way: if you are busy with the MP1A, then I would increase the value of the 15 uF capacitor connected to SW3A.
This will improve the bottom end, especially at higher gain settings. (I would use something like 47 uF or so.)

RuudNL said:

The DI input separates the input transformer from the rest of the circuit. So if this eliminates the hum, it must be coming from the circuit around the input transformer.

Click to expand...

Yes, it is puzzling. I almost get the impression that the flipping of the switch is randomly causing a ground-loop, but it doesn't make sense...

are there any  dc blocking caps?

RuudNL said:

By the way: if you are busy with the MP1A, then I would increase the value of the 15 uF capacitor connected to SW3A.

Click to expand...

The largest filmcap I can find is a 33uF WIMA. Can I also use a electrolytic cap for this? My 'standard mod' as per Jakob is replacing the 4,7 feedback elco with a film-capacitor.

pucho812 said:

are there any  dc blocking caps?

Click to expand...

yeah, but they are after the input transformer. check schematic here: mp1a-schematic

small update:
I tried to do some measurements, but as soon as I start probing with my multimeter the buzzing stops. Particularly after I touch the + side of the 4,7uF cap. I also measure 44+ V over there..

ye olde plot thickens.... the pad switch and volume switch also cause the hum to appear..  ???

Can I also use a electrolytic cap for this?

Click to expand...

I could not think of an argument why you could not use an electrolytic here!
The side connected to SW3A is always positive referred to ground anyway.
If you belong to the 'phreakz' (they are the worst!  ;D ), then you could use a film capacitor in parallel with the electrolytic.

have you checked for cold  solder joints? have you put it on a scope?

Nele said:

small update:
I tried to do some measurements, but as soon as I start probing with my multimeter the buzzing stops. Particularly after I touch the + side of the 4,7uF cap. I also measure 44+ V over there..

ye olde plot thickens.... the pad switch and volume switch also cause the hum to appear..  ???

Click to expand...

Looks very much like an oscillation. You must check with a scope the output of the second triode. Possibly changing the tube may cure the problem.

pucho812 said:

have you checked for cold  solder joints? have you put it on a scope?

Click to expand...

not yet.. I'm gonna hook it up to the scope tomorrow or friday.
the noise seems a bit low-freq for oscillation. it sounds more like a ground-loop..

FWIW I would just reheat all the joint making them shiny and new. It's not going to hurt anything and it might actually fix your problem. Secondly if it sounds like a ground loop then check for things not going to ground properly.  you mentioned the pad switch and the rotary switch are also causing this hum. IS that with or without phantom?

pucho812 said:

you mentioned the pad switch and the rotary switch are also causing this hum. IS that with or without phantom?

Click to expand...

Either way, phantom on/off doesn't matter. What makes it very annoying is that it's not easily re-produceable. I first thought it was only the phantom-power switch but I accidentally found that flipping the pad-switch or volumeswitch can cause it as well. It must be something between the input and the first tube-stage.
There is a flat cable between the frontpanel controls and the preamp pcb so maybe there is a bad connection there...

Nele said:

pucho812 said:

have you checked for cold  solder joints? have you put it on a scope?

Click to expand...

not yet.. I'm gonna hook it up to the scope tomorrow or friday.
the noise seems a bit low-freq for oscillation. it sounds more like a ground-loop..

Click to expand...

When a circuit oscillates, although it produces a generally inaudible frequency, it makes it very sensitive to any stray field, thus amplifying any latent hum.

abbey road d enfer said:

When a circuit oscillates, although it produces a generally inaudible frequency, it makes it very sensitive to any stray field, thus amplifying any latent hum.

Click to expand...

aha! thanks for the info. will check tomorrow!

cheers everyone for the help!

small update:
- if I remove the first ECC83 on channel 1, psu voltages are ok, but if I measure them with the tube inserted they are way too low: 270V measures 227V, 240V measures 220V..
- I hooked it up to the scope and there is ripple on the phantompower and heater-supply, but only with the first tube inserted...

will check solderpoints tomorrow

If there is ripple then there  could be a bad electrolytic cap or several bad electrolytic caps in the PSU

one thing I noticed in the psu-schematic is that in the schematic there is a 2200uF/63V cap in the 33/48V section and in my unit there is only a 1000uF cap.

Nele said:

- I hooked it up to the scope and there is ripple on the phantompower and heater-supply, but only with the first tube inserted...

Click to expand...

IIRC, the tubes heaters (connected in series) are run off the 48V - meaning phantom and heaters on same circuit...

Jakob E.