Pro Jr Point-to-point hum issue

[klooon]

First, never, ever use an ungrounded socket. If it's just a slight sensation likely it's just stray capacitance between the windings. Using an ohm meter to check that the earth from the plug has a low resistance to the chassis. Also check that live goes to the power switch and fuse. Don't just look at wires. Assume that the wire colours can't be trusted so measure from the plug. Check the neutral does not connect to the chassis

I swapped them around, but no change. Also in the Netherlands, you can plug it either way (180deg), so swapping L and N. I swapped it with the plug and also in the amp, but no difference...

You hum issue is elusive. Disconnect the grid from the screened wire and 100 ohm grid stopper that goes to the volume control wiper and short the grid to the cathode using as direct a connection as you can manage. What happens to the hum?

I just tried this. No hum at all....
When I did not connect something to the grid, (so not jumpering grid and cathode, thus leaving the grid unsoldered), there was a very big oscillation... beeping noise that increased so much that I needed to power off the amp...

[nickb]

I just tried this. No hum at all....
When I did not connect something to the grid, (so not jumpering grid and cathode, thus leaving the grid unsoldered), there was a very big oscillation... beeping noise that increased so much that I needed to power off the amp...

Good. That narrows it to the grid circuit.

Reconnect the screened wire to the grid pin 7 and connect the other end ( used to go to the wiper) with the the screen to the ground end of the 1.5k cathode resistor. How's the hum? Now try positioning that wire so that it's parallel to the 1.5k resistor and the wire to the cathode. What we are trying to do it minimize the area of the loop formed by grid->ground->cathode resistor->cathode. Better? If better then you are getting magnetic fields coupling into the grid circuit, a bit like the secondary of a transformer. You can reduce the coupling by keeping the loop area to a minimum, you can position it are right angles to the intrusive magnetic field which practically speaking may be very hard to to, or try to figure out what is creating the field. The usual culprit is heater wiring, so twist and reposition. The HT wiring can do it as can the power transformer itself.

[klooon]

Good. That narrows it to the grid circuit.

Reconnect the screened wire to the grid pin 7 and connect the other end ( used to go to the wiper) with the the screen to the ground end of the 1.5k cathode resistor. How's the hum? Now try positioning that wire so that it's parallel to the 1.5k resistor and the wire to the cathode. What we are trying to do it minimize the area of the loop formed by grid->ground->cathode resistor->cathode. Better? If better then you are getting magnetic fields coupling into the grid circuit, a bit like the secondary of a transformer. You can reduce the coupling by keeping the loop area to a minimum, you can position it are right angles to the intrusive magnetic field which practically speaking may be very hard to to, or try to figure out what is creating the field. The usual culprit is heater wiring, so twist and reposition. The HT wiring can do it as can the power transformer itself.

Thanks! First time I tried to connect the screened side to the ground end of the 1k5 and it was dead quiet. Then I started moving around the screened wire (including the stiff end with the resistor attached) and then it got worse as if the amp was going to explode. A big increasing noise/hum. But I can't get it back to the situation where it is quiet. Even worse: I removed the screened wire and tried jumpering the grid and cathode and it makes it noise.. Maybe I destroyed the tube feet? Maybe install a new one and resolder V1 all over?! Maybe lower the heater wires for V1? Any advice/good practice picture?

[nickb]

Oh dear! Well there are a number of possibilities so you will have to troubleshoot logically. But a quick visual and push and shove check that no solder joints have come loose, especially on grounds and filter caps. If that doesn't reveal anything pull all the tubes and try it out. Start replacing one EL84, then the other then the phase inverter V2 and finally V1 and see when the noises start.

[klooon]

Oh dear! Well there are a number of possibilities so you will have to troubleshoot logically. But a quick visual and push and shove check that no solder joints have come loose, especially on grounds and filter caps. If that doesn't reveal anything pull all the tubes and try it out. Start replacing one EL84, then the other then the phase inverter V2 and finally V1 and see when the noises start.

I did the check and it's only noisy when V1 comes in again. When leaving grid bare (no connection with any wire), there is a huge hum... I then I put a small unshielded wire from the grid to the hot end of the cathode resistor 1k5. And it is silent. Then putting a shielded wire with a 100R at the grid lug and only connect the shield to the ground side of the 1k5 cathode resistor and it is very noisy; moving around with the wire and it increases in hum and noise very badly.

[Helmholtz]

When leaving grid bare (no connection with any wire), there is a huge hum... I then I put a small unshielded wire from the grid to the hot end of the cathode resistor 1k5.

Both these conditions, open grid and grid shorted to cathode leave the tube without proper bias, a very undesirable and useless state. Open grid makes the tube highly sensitive to EMI/noise and grid to cathode short prevents any amplification. So your hum results are obvious.

The grid needs to have a DC path to ground, typically accomplished by a 1M grid leak resistor or other resistive path. It makes sure that the grid potential is at 0 Vdc.

[klooon]

Both these conditions, open grid and grid shorted to cathode leave the tube without proper bias, a very undesirable and useless state. Open grid makes the tube highly sensitive to EMI/noise and grid to cathode short prevents any amplification. So your hum results are obvious.

The grid needs to have a DC path to ground, typically accomplished by a 1M grid leak resistor. It makes sure that the grid potential is at 0 Vdc.

Thanks for the explanation. Good to know that "this" is normal.
Your second comment. It seems that there is no path with a 1M or so to ground, maybe the 250k from the volume to ground when volume is zero? Is something missing in this design?

[Helmholtz]

It seems that there is no path with a 1M or so to ground, maybe the 250k from the volume to ground when volume is zero?

Yes, R6 and R7 take care of the grid ground reference for the second stage.

[nickb]

I did the check and it's only noisy when V1 comes in again. When leaving grid bare (no connection with any wire), there is a huge hum... I then I put a small unshielded wire from the grid to the hot end of the cathode resistor 1k5. And it is silent. Then putting a shielded wire with a 100R at the grid lug and only connect the shield to the ground side of the 1k5 cathode resistor and it is very noisy; moving around with the wire and it increases in hum and noise very badly.

You need to connect the inner wire AND the screen together for the tests I suggested in post #32. Sorry, I see I didn't make it clear. Having no DC path to ground for the grid is not a valid mode of operation.

Anyway, we have learned that connecting the grid to the 1.5K is silent. Is it still silent if you make a big loop with it (stray magnetic field test)?

We got off track a bit so we might have to repeat some of our steps.

Connect ( inner + screen) to lug three of the volume pot what happens? If hum then check that ground wire from lug 3 to ground and / or try repositioning the screened wire, perhaps run it under the board close to the chassis mirroring the 1.5K resistor path.

[klooon]

You need to connect the inner wire AND the screen together for the tests I suggested in post #32. Sorry, I see I didn't make it clear. Having no DC path to ground for the grid is not a valid mode of operation.

Anyway, we have learned that connecting the grid to the 1.5K is silent. Is it still silent if you make a big loop with it (stray magnetic field test)?

We got off track a bit so we might have to repeat some of our steps.

Connect ( inner + screen) to lug three of the volume pot what happens? If hum then check that ground wire from lug 3 to ground and / or try repositioning the screened wire, perhaps run it under the board close to the chassis mirroring the 1.5K resistor path.

I have replaced the tube socket and resoldered everything. The heater wires are now on the chassis and not up in the air. Everything is connected except the grid wire.
Yes, the grid to 1k5 hot side is silent when using a short wire. And a long wire is not. It is picking up this hum. Exactly this hum.... Moving around the wire is more or less hum...

[nickb]

I have replaced the tube socket and resoldered everything. The heater wires are now on the chassis and not up in the air. Everything is connected except the grid wire.
Yes, the grid to 1k5 hot side is silent when using a short wire. And a long wire is not. It is picking up this hum. Exactly this hum.... Moving around the wire is more or less hum...

You were supposed to connect the screened wire (inner + screen) that originally went the the wiper to the GROUND end of the 1.5K, not the HOT end, assess the hum and report of the effect of changing the loop area/ shape. If silent then move it (screen + inner) to the lug three of the volume pot and repeat the experiment. It's important to do this things exactly else we'll reach the wrong conclusions.

[klooon]

I have replaced the tube socket and resoldered everything. The heater wires are now on the chassis and not up in the air. Everything is connected except the grid wire.
Yes, the grid to 1k5 hot side is silent when using a short wire. And a long wire is not. It is picking up this hum. Exactly this hum.... Moving around the wire is more or less hum...

Then, if I understand you correctly, I put the screened wire (the one with the inner to 100R attached to the Grid) and the other end, both inner and shield attached together, to the lug 3 of the volume pot. Then there is a hug hum and moving it around it picks up more buzz. But the hum is still there.
I have put the inner and shield then to the ground side of the 1k5. Still the same hum (maybe a bit less buzz).
I also ran it under the board, but the hum is the same...

[nickb]

Then, if I understand you correctly, I put the screened wire (the one with the inner to 100R attached to the Grid) and the other end, both inner and shield attached together, to the lug 3 of the volume pot. Then there is a hug hum and moving it around it picks up more buzz. But the hum is still there.
I have put the inner and shield then to the ground side of the 1k5. Still the same hum (maybe a bit less buzz).
I also ran it under the board, but the hum is the same...

Correct - thanks To have such a big difference it's almost as though the ground wire to lug 3 is connected to the wrong place or even not at all. If you connect your test screened wire ( inner+shield) to the point that the ground wire from lug 3 connects to the board, how is the hum? That should be just a short distance from there to the cold end of the 1.5K cathode resistor which is hum free, right?

I'm assuming this is what you have layout & results so far:-

[klooon]

Correct - thanks To have such a big difference it's almost as though the ground wire to lug 3 is connected to the wrong place or even not at all. If you connect your test screened wire ( inner+shield) to the point that the ground wire from lug 3 connects to the board, how is the hum? That should be just a short distance from there to the cold end of the 1.5K cathode resistor which is hum free, right?

I'm assuming this is what you have layout & results so far:-
[ATTACH=CONFIG]58000[/ATTACH]

Unfortunately, the cold side of the 1k5 is not hum free. I thought yesterday, that it was for a moment. But up to now I could not reproduce it. So all it humming. This is 100Hz hum. The 50Hz and its multiples are changing in level when moving the wires around. But the 100Hz hum stays...

[nickb]

Unfortunately, the cold side of the 1k5 is not hum free. I thought yesterday, that it was for a moment. But up to now I could not reproduce it. So all it humming. This is 100Hz hum. The 50Hz and its multiples are changing in level when moving the wires around. But the 100Hz hum stays...

I think that puts us back to square one. Focusing on the 100Hz hum, try this. Hook it all back up the way it should be, screen to lug 3 and inner to lug 2 of the volume control, set vol to '1' and note hum level. Now disconnect the negative end of 22uf that goes to node "1" on the layout diagram and connect it to node "2" instead. Any better?