Sluckey’s diode across the intensity pot mod on a 6G9-B Tremolux

[Mark]

Put the tremolo tube on a separate power supply node. Connect a 1K to the screen node and then a 40µF cap from the 1K to ground. Use this new node to supply the tremolo circuit. Does that help the thumping?

Thanks for your suggestion, I did try it and it didn’t get rid of the thumping. The output was reduced to 3.57Vpp from 5.61Vac. It’s a most peculiar issue, which seems quite unusual.

IMG_5252.jpeg

EDIT: One of the things I learnt or remembered was that the CRO on AC mode attenuates the low frequency waveforms. I mistakenly used AC in the picture above, it should be DC and the voltage would more than likely be no better than not having the extra power supply node.

[Helmholtz]

I find it so strange that the thump is quieter with the P.I. removed.

That's a direct consequence of that strange PI input signal, which should not be there.
As this signal gets amplified by the PI, producing opposite phase PI plate signals, it gets processed like a normal (guitar) signal without cancellation by balanced output tubes.

I would also consider a ground contact problem.

[martin manning]

I would say the first thing to do is make sure both the B+ and bias supply ripple voltages are low (all-new filter caps), and that the output tubes are well matched. You can adjust the bias level to keep the output tubes safe, and when this bias modulating system is working it sounds very good. IIRC the reason Fender abandoned it was because of the frequent adjustment that it required. Beyond that the issues are likely to be things that are not shown on the schematic, like undesired coupling between leads, components, or via the circuit board material.

[Mark]

I find it so strange that the thump is quieter with the P.I. removed.

That's a direct consequence of that strange PI input signal, which should not be there.
As this signal gets amplified by the PI, producing opposite phase PI plate signals, it gets processed like a normal (guitar) signal without cancellation by balanced output tubes.

I would also consider a ground contact problem.

I did lift the earth for the oscillator and move it to the chassis earth. I didn’t move the screen power supply node earth though. It made no difference to the thumping.

[Mark]

I would say the first thing to do is make sure both the B+ and bias supply ripple voltages are low (all-new filter caps), and that the output tubes are well matched. You can adjust the bias level to keep the output tubes safe, and when this bias modulating system is working it sounds very good. IIRC the reason Fender abandoned it was because of the frequent adjustment that it required. Beyond that the issues are likely to be things that are not shown on the schematic, like undesired coupling between leads, components, or via the circuit board material.

Hi Martin, sorry for the slow reply, my son is to be sitting for his HSC exams and I have dropped everything to help him.

I didn’t replace the filter caps as they had been previously replaced with Sprague Atom 16uF caps. I did measure all of them with my Peak ESR70, the values of the caps were good as was the ESR. I think the value of the caps would have to be increased in order to remove the waveform from the power supply or a more sophisticated design would be needed.

I agreed the better the output valves are matched the better the common mode cancelation. However I have found the louder the amplifier is played the quicker the valve matching diminishes, so I don’t put a lot of faith in the output valves staying matched for long.

Why Fender stopped using bias modulating tremolo?

I decided to ask Grok why Fender stopped using bias modulating tremolo before I add my observations.

GROK: “Bias voltages are higher on 6L6 circuits than on 6V6 ones, so the LFO signal often cannot swing the bias enough for deep, consistent modulation unless the tubes are biased relatively cold. Cold bias can introduce crossover distortion, reduce headroom, thin out the tone, or make the amp sound less lively when the tremolo is off. At hotter, more typical bias settings preferred for tone and tube longevity, the tremolo depth becomes weak or disappears—especially when the amp is turned up. This trade-off is repeatedly noted by amp techs and builders.”

My thoughts are the tremolo doesn’t only decrease the bias it also increases it. I found if I biased the amp for 60% dissipation (40mA when the rail is 420VDC), waveform on the grid of the 6L6 was going above 0V. My TAD Bias Master was showing around 55mA as an average measurement. I think extensive use of the tremolo may cause additional wear upon the output valves.

I thought the board being conductive might have been a factor too. I have encountered it before in other Fender amps. The cathode and grid components are located next to the oscillator and driver components so I mounted all of these components off the valve pins to eliminate this as a possibility. The reading were the same in this case.

I end up putting a 47nF cap across the plate resistor of the oscillator and across the intensity control. I also used John G’s suggestion of using a 330K resistor in series with the diode. It was very quiet after that and the current only dropped about 5mA’s as opposed to dropping to 5mA’s.

IMG_5323.jpeg