JTM45 Build

[guitarmike2107]

What a bummer.

Measuring the pots may not show a fault by the way. But worth mentioning that you had bypassed the treble pot previously in a conversation you had with me with no effect

Did you unbundle all those wires as suggested with the rebuild?

Does it show the same symptoms without the attenuator. What attenuator is it?

Did you find the larry grounding scheme. I would defo do that or follow merlins/rubys/many others adviser.

Right now you are stabbing in the dark. Work methodically, Sort out the possible issues you think you may have and move onto the next on. I.e.
Grid stops, grounding (needs lots of work), lead dress (those bundles) etc

Time for an updated picture please

[Littlewyan]

The Primaries and Secondaries are unbundled.

Tried without the attenuator, same issue (that hurt).

I have modified the grounding slightly. I cut the buss bar in half to separate the pre amp from the PI and I connected the speaker ground to the PI ground.

I went through the amp with a scope last night and my meter taking note of every little change as the issue occurred. I may have possibly found some oscillation, I saw a little blip on the waveform that moved up and down the waveform as I turned the controls. Not sure if it was oscillation or not.

However what I did find was when I turn the treble up by itself the 82K Anode on the PI pretty much stops drawing current and the 100K side draws a LOT more current which is why the voltage drops. Thats why the bias voltage drops on V4 as V5 goes up, because V4 isn't really being driven much at all at this point and is probably the cause of the bad sound.

Now whilst watching the waveform on the PI anodes with all tone controls at 0 and volume on 10 its a square wave with one corner missing. When I bring the treble up the side thats missing rises up and then levels out with the top. However when I bring the treble up over 4/5 the other side of the squarewave starts going down so I end up with the same wave with the treble on 0, just inverted.

Nothing changes before the PI and the issue still occurs with the power valves out so Colossal narrowed it down to the PI and Bias Circuit. I've checked all of the resistors and each one measures within tolerance, which is why we suspect there is a leaky cap somewhere.

[joeboo88]

Ryan the amp still looks beautiful, i figured that with all the changes your amp would end up looking like mine .....keep plugging away

[sluckey]

This idea is kinda out there but so is your problem. Should only take a couple minutes to test it.

The way you have the threaded end of your bias diode connected makes me nervous (or is that a factory lead attached to the threaded stud). If you replace it with a 1N4007 does it help the problem?

[Littlewyan]

The diode came with the leads already attached if thats what you mean? I tested the diode with a meter and it tests fine on the diode setting. I can measure resistance across it which reads just over 1Meg and I imagine its because its in the circuit. I haven't tried replacing it but can do as a test. When I use the scope on the bias circuit do I need to set the scope to DC or AC to detect ripple? As when it was set to AC I didn't pick anything up?

Didn't have time to do any tests tonight, will hopefully have time to look at it again tomorrow night. Least we've narrowed it down now to the PI/Bias Circuit.

Thanks for the compliment Joe.

[sluckey]

When I use the scope on the bias circuit do I need to set the scope to DC or AC to detect ripple? As when it was set to AC I didn't pick anything up?

Needs to be on AC and set to a sensitive input level to measure ripple. Set trigger source to line.

I've never seen a factory fresh stud mounted diode with a lead soldered to the threaded end. Doesn't mean they are non existent. Usually stud mounted diodes are meant for high current applications and the stud provides a means for mounting to a heatsink. . A bias supply in a tube amp is a very low current application.

[Littlewyan]

Tested the diode and its fine. Tested the bias supply caps, PI Filter Cap and the 47pF PI Cap and all test ok. Also tested the 100K Anode Coupling Cap and its fine.

[Littlewyan]

Just disconnected the PI from the tonestack and hooked my signal generator up to it and couldn't get it to swing at all. I sent a 4Vpp signal into it, used sinewave, squarewave, trianglewave and frequencies from 100Hz up to 50Khz. Nothing. So perhaps its not the PI.

[guitarmike2107]

Remove the power valves, Disconnect the bias circuit from the grid leaks and ground the grid leaks as if it was cathode biased power section, now if you still have the same issue then it is not the bias circuit causing it

[Littlewyan]

It still occurs with the bias circuit disconnected and grid leaks grounded.

Do you think we can rule out the PI? I even set my oscillator to sweep when I tested it and the PI voltage stayed stable throughout the test on both sides.

Just tried a 100K grid stopper on V2A and that didn't make a difference so I think we can rule out that stage. That just leaves the cathode follower (V2B) and the tonestack.

Reeltarded mentioned how he once had a valve socket that gave him problems by making noise. Maybe my V2 socket is bad? Can't be V1 as I put my signal generator onto the mixer resistors bypassing the V1 circuit and had the same issue.

[Littlewyan]

Ok lets think this through logically here. The PI on it's own is fine, however when hooked up the tone stack something shifts the bias so much that one side is basically cut off and the other runs harder. If we look at the bias for a 12AX7 as the grid becomes more positive it draws more current, and as it becomes negative it starts to get starved of current. So either something is dragging the 82K side negative or pushing the 100K side positive. Its only got to happen to one side as obviously the signal will be inverted onto the other side and cause the opposite affect.

It must be oscillation as the PI by itself is fine and can only be a leaky cap if there is a cap in the Tonestack that leaks and the PI Input Cap leaks. Here is what I saw on the scope when I created the issue and then stopped it by turning the middle up. There is something on the top left part of the waveform that disappears when I turn the middle up.

Also I should mention that I monitored the PI output and found that when I turned the middle up to stop the issue the following frequencies got squashed. 7Khz, 16Khz, 23Khz and 30Khz. They also varied so sometimes I saw 32Khz instead of 30Khz. Anyway I traced these back to V2A's anode/V2B's grid and found that they also went down there when I turned the middle up. This must point to coupling between stages right?

[Littlewyan]

My response appear before your's for some reason Miles even though I posted it hours after :S

[Reeltarded]

I tested it 100 times too. Pulled tube slightly and checked positively with a signal every time on every pin to every termination and across components. Nutty.

You start redefining insanity with these kinds of issues.

You are doing a great job of getting it apart and together again.. if nothing else you will be an assembling ninja for the suffering.

Hope.

[guitarmike2107]

You can bypass v2. Put a wire from the mixer resistors over to the pi cap. Will need to increase the input signal but I think you can do that?

You still have a jj in v2

[Littlewyan]

I could hit the front end with a 4Vpp signal which should give me more than enough signal to slam the PI. I'll try that tomorrow.

I did try lifting the tonestack earlier and found that as I turned the volume up the PI voltage dropped at first as it would because the HT is dropping, but then at 2 on the volume the PI 82K voltage started to creep up and then after about 3 it just went back down. So the issue is obviously still there.

Also I think even with the middle control up the high frequencies are still not getting through as turning the treble up or down doesn't make a huge difference and the PI voltage still moves as I turn up the treble. So the middle is obviously just masking the issue enough for the PI to function and whatever high end the treble pot is supposed to allow through is probably getting cancelled out.

Also yes I still have a JJ in V2 which should be fine. My pre amp voltages are lower now don't forget as I have the additional HT dropper resistor.