JTM45 Build

[JazzGuitarGimp]

The thick trace segments on the tops and bottoms of the waveform is usually indicitive of high frequency oscillation.

[Littlewyan]

The one Martin is talking about is the one that had the 'accident' but I'm sure I already tried replacing this looking back at my notes.

What is interesting is how the presence doesn't make the issue worse. Added to which if I turn up the middle or bass the bias voltage issue goes away and the power valves act normal even though the sound is still bad. I haven't measured the PI plates when doing this though. But as soon as I leave just the treble up the bias voltage on V4 starts to go more positive.

We've ruled out the power section and the NFB and I'd say we've ruled out V1 as the issue starts to occur when I inject the signal into the mixer resistor. Least the PI 82K Anode voltage still swings positive by quite a bit as I turn up the treble. So it must be between the mixer resistors and the PI.

[Littlewyan]

Ok here is my plan. I've just tried to scope the amp and I can't see any oscillation on the input grid on the PI but I can see it on the other grid. Not sure if this is just the NFB from the power section so I will remove the power valves again and perhaps stick a load on the amp as the voltages were a tad high and I was worried about the caps. Then I should be able to see if the oscillation is on the grids or not.

[JazzGuitarGimp]

Many times, just the little bit of added capacitance by connecting your scooe probe to a grid is enough to cause parasitic oscillation to stop. It's better to scope the plate of a suspected tube as this will not interfere with the oscillation. If the plate has an unwelcome waveform, the grid does too.

[Littlewyan]

Ok I've got something! I put the volume on 4 (kept it lowish as the input signal is 2Vpp) and had all tone controls on 0. As soon as I turn the treble control even just a tiny bit the waveform shifts to the right a bit and then as I turn the treble up more half of the waveform shrinks. Whereas turning up the middle or bass don't do this, nor does turning up the middle and THEN treble. I filmed it. Will try and upload it now.

[Littlewyan]

Ok I can't upload it as the iPhone quality is too good so the video is 54MB even though its only 10 seconds long -_-.

Just found the anode of V2a does the same thing as the volume is increased.

Actually I think its just the scope.

Just tried swapping out the 0.1uF you mentioned Martin but its made no difference .

What if one tiny little strand of wire was touching another turret. This probably wouldn't show up on a meter but could cause an issue yes?

I think I'm going to have to pull the board out, redo all of the wires underneath it, then put it all back in. I bent the wires round underneath the board so they rested against the chassis but some got twisted. I'm wondering if a few strands are touching somewhere.

[martin manning]

Depends on which wire this tiny strand belongs to and which turret. It would probably show up on a meter if there is electrical contact. Where is this suspected short?

Another idea is since you can hear this problem you could make up a listening probe, have someone play through it or whatever while you follow the signal.

[Littlewyan]

How do you mean a listening probe?

I don't know where the suspected short is, I just can't see what is causing the issue.

Is it normal for the PI to swing really positive on the 82K side and negative on the 100K side? Or am I wasting my time trying to fix that. As when I'm trying things to fix the problem I'm not playing the guitar I'm just monitoring that.

[martin manning]

Listening probe: http://billmaudio.com/wp/?page_id=1254

[sluckey]

Is it normal for the PI to swing really positive on the 82K side and negative on the 100K side?

No it's not. Apply a 0.2v peak to peak sine wave to the input jack. Set your volume to deliver a clean sinewave to the input of the PI. The signal on each PI plate should be a clean sinewave that is bigger than the signal at the input. They will be 180° out of phase with each other but should be nearly the same amplitude.

With power off, measure resistance from pins 2, 3, 7, and 8 of the PI to ground. One lead on chassis, the other directly on the tube socket. What are the numbers?

[Littlewyan]

The signal generator I have at hand doesnt have any adjustment on it.

Pin 2 1.003M
Pin 3 14K
Pin 7 1.011M
Pin 8 14K

[sluckey]

The signal generator I have at hand doesnt have any adjustment on it.

Put a pot on the output of the sig gen. Wire it like a typical voltage divider volume pot. Need to get the signal down to a level that does not overdrive the PI so you can see if it is behaving normally.

Resistances look OK.

Measure the voltages on pins 1, 3, 6, and 8. Also measure voltage at the node where two 1MΩ, 470Ω, and 10KΩ come together. This is the real grid voltage.

[Littlewyan]

Pin 1 - 202V
Pin 3 - 30V
Pin 6 - 192V
Pin 8 - 30V

The node where the 1M, 470R and 10K come together measures 30V.

This is all with a clean output on the PI.

Edit: One thing that bothers me is when I soldered the under board wires onto the turrets something bubbled around the turrets on the board in certain areas. Also on one turret the solder kept making itself a bubble then bursting! I have never had that before. You dont think this board has become conductive do you? Its phenolic material.

Added to which I can measure a very small voltage on the treble cap, pi input cap and the treble pot. Its tiny though, about 20mV.

[martin manning]

That doesn't sound good. What do you get if you push your meter probe into the board around the turrets carrying high voltage?

[guitarmike2107]

I have seen phenolic boards bubble, but that was after to much heat and solder, I don't think it was actually the board bubbling but the resin.

As far as I know there is nothing in a paper based phenolic that would cause conduction, the last time there were conducting boards was due to them using graphite or similar to make the FR4 boards black. i.e. boards don't become conductive, it would be a manufacturing fault and would likely affect the whole board I would think?

Your top connections look fairly clean/no bubbling

Did you solder the underside wires after soldering the top connections?

Since I supplied the board it better bloody not be conductive!