Troubleshooting IMD and hum in a 1959SLP

[Gtr0]

Well, in the meantime, I turned on to take some voltages as is....

There is DEF a problem here. On one side of the output tranny, the amp hums when checking the B+ on pin 3 and does not give a very stable VDC reading. The other pair seem fine.

Powered down, cleaned and rehit a bunch of joints, even though they looked fine. Traced (again) through the power section all the way to the plates, all seems fine.

Changed tubes from another working amp - same thing... hum on Pin 3 of one side of the output tranny... it ends up reading 485 VDC here while the other dside reads 515VDC. Pin 5 (bias) all read -49VDC.

CT to pin 3 on all sockets read good... 16ohm on one pair and 18ohm on the other.

My next thoughts are to just replace everything on the side that is acting wierd... sockets, any resistors etc, even if all is tracing out okay. What else could I do??

[eschertron]

Well, in the meantime, I turned on to take some voltages as is....

There is DEF a problem here. On one side of the output tranny, the amp hums when checking the B+ on pin 3 and does not give a very stable VDC reading. The other pair seem fine.

Powered down, cleaned and rehit a bunch of joints, even though they looked fine. Traced (again) through the power section all the way to the plates, all seems fine.

Changed tubes from another working amp - same thing... hum on Pin 3 of one side of the output tranny... it ends up reading 485 VDC here while the other dside reads 515VDC. Pin 5 (bias) all read -49VDC.

CT to pin 3 on all sockets read good... 16ohm on one pair and 18ohm on the other.

My next thoughts are to just replace everything on the side that is acting wierd... sockets, any resistors etc, even if all is tracing out okay. What else could I do??

What's B+? I'd suspect trouble with the socket where pin 3 reads high. 515vdc suggests the tube is not conducting? So either the plate pin or the cathode pin not making good contact with the tube.

edit: of course, an open screen connection would also prevent the tube from conducting. Maybe a 1k screen stopper is actually 100k in circuit? Checking voltages with tubes out would not reveal this. What are all the voltages on the power tube sockets with tubes in? Just for grins, swap tubes and repeat the measurements.

[Gtr0]

B+ at the filters is 515VDC

The 485 side is the side humming while checking plate voltage.

[Helmholtz]

I think you are triggering a power amp HF oscillation when touching the plates with your meter probe. Symptoms are typical: lowered DCV reading and humming.
The meter lead causes a positive feedback by parasitic capacitance. It only happens on one side of the OT primary as the other side is 180° out of phase.

You should get symmetrical plate voltage readings if you disconnect the NFB loop. Sometimes it also helps to turn presence fully up.

[Chuck H]

I think you are triggering a power amp HF oscillation when touching the plates with your meter probe. Symptoms are typical: lowered DCV reading and humming.
The meter lead causes a positive feedback by parasitic capacitance. It only happens on one side of the OT primary as the other side is 180° out of phase.

You should get symmetrical plate voltage readings if you disconnect the NFB loop. Sometimes it also helps to turn presence fully up.

I think you meant "touching the grids".?. This sounds like a probability, but I've never seen it. I'm interested to hear if this is what's happening because it seems like a rare problem and an intuitive observation.

[Helmholtz]

I think you meant "touching the grids".?.

No I meant plates. Power tube plates are the highest gain and amplitude "outputs" of a tube amp. Connecting a long lead to them produces a transmitting antenna.
The oscillation eats power and thus loads the power supply, increasing ripple. Ripple at the speaker output is further increased by OT primary impedance asymmetry at HF. As now there is HF signal at the plates, average plate voltages must drop.

With a scope having sufficient bandwidth you could see the HF.

I am familiar with this artefact especially from servicing Marshalls. It's much more likely in amps using a lot of global NFB.

Anyway it's just a measuring artefact and no real amp problem. But the oscillation may be very high amplitude and is likely to stress the OT.

[Gtr0]

I think you are triggering a power amp HF oscillation when touching the plates with your meter probe. Symptoms are typical: lowered DCV reading and humming.
The meter lead causes a positive feedback by parasitic capacitance. It only happens on one side of the OT primary as the other side is 180° out of phase.

You should get symmetrical plate voltage readings if you disconnect the NFB loop. Sometimes it also helps to turn presence fully up.

Thank you for the info!

I do have an oscope that recently stopped giving me a trace, and have a new one on the way, but that wont be for at least a week.

At any rate, I replaced the 4 power tube sockets again. I originally replaced them with what I thought were known working ones... so this time I used brand new... and well if nothing else I spent more time doing a solder job on these points at least.

And if what your saying is true, I should expect the same behavior?

the presence circuit is, as always ;-), on full.. 100kΩ r to the 8Ω tap from a quick look...

If I get the same behavior, I will unsolder it nfb from the 8Ω tap and recheck.

Thank you!!

[Helmholtz]

Look at it this way:

If the voltage drop of 30V across a primary half was caused by high idle (DC) plate current, this would mean a total current of 1.77A and a dissipation of 866W for the 2 tubes (assuming a typical primary half DCR of 17 Ohm for a 100W Marshall OT). Not very likely

There is simply no component (tubes, resistors, sockets, transformers) in an amp that could take that much power without melting, burning, exploding - and of course fuses would blow.

[Gtr0]

So after replacing tube sockets, rewiring everything on the power amp basically, I got the same result... that is until I turned down the presence pot from 10 to 1... then I got the same reading on pin 3, 514 VDC.

So it made me look at the presence wiring... and while it actually is 47k to 8Ω tap (earlier I quoted 100k), I can find no error in the wiring.... does this sound correct?

Helmholtz, I have not seen this before, is this typical that the presence circuit will do this to voltages?

Thanks all for your wisdom!!

[Helmholtz]

Helmholtz, I have not seen this before, is this typical that the presence circuit will do this to voltages?

Assuming the presence works as intended, the position of the presence pot does not actually change plate voltages, it just changes the amount and phase of HF NFB, thus influencing stabilty conditions (phase margin).
What actually changes plate voltages is your positive meter lead by causing oscillation.

Did you try without NFB?

[Gtr0]

Assuming the presence works as intended, the position of the presence pot does not actually change plate voltages, it just changes the amount and phase of HF NFB, thus influencing stabilty conditions (phase margin).
What actually changes plate voltages is your positive meter lead by causing oscillation.

Did you try without NFB?

No I did not, only because turning the pres knob down seemed to rectify the situation... should I? I placed the positive lead to the plate and turned the pres pot and the voltages were going back to normal (rising toward 515 VDC) as I tuned the pot toward zero.

An another thing I have noticed since rewiring is that the hum does not seem to be a factor now, although the ghost notes still remain though only noticeable on high solo notes.

[Helmholtz]

... should I?

Not really necessary, at least it won't give me additional information.

Ghost notes (= intermodulation distortion) mean that some residual hum or ripple mixes up with your guitar signal. Without a scope it will be hard to find the culprit.

Do you have a dummy load? If yes, you should insert a 100mV/400Hz sine signal at the input, adjust controls to get 35VRMS at the power tube grids and measure output power.
It is also a good idea to wire 1 Ohm resistors between each EL34 cathode and ground. This allows to individually check the idle currents of the tubes. Won't have any effect on sound.

[Gtr0]

It is also a good idea to wire 1 Ohm resistors between each EL34 cathode and ground. This allows to individually check the idle currents of the tubes. Won't have any effect on sound.

ha. *just* removed them yesterday because I take bias a different way... but I am intrigued... it'll take me a few days, but I will do this!

Thanks for the info!!!!!

[Chuck H]

... the voltages were going back to normal (rising toward 515 VDC) as I tuned the (presence) pot toward zero.

Do you mean to say that adjusting the presence control is changing the plate voltage?