Tremolo Changes after EL84 to 6V6 Swap

[tictac]

Hello guys,

I have a Princeton Reverb type amp that originally had EL84 power tubes. It has a bias modulated tremolo circuit which works fine with the EL84's.

Later, I changed the power tubes to 6V6's. Amp sounds great except now the tremolo is very weak. With the intensity cranked to 10 the tremolo can only faintly be heard.

Negative bias had to be increased from -16VDC to -33VDC to accommodate the 6v6 tubes

I'm trying to figure out how to restore the tremolo intensity. Would this be a matter of increasing gain of the tremolo output?

I was thinking about adding a bypass cap to R62 but a pair of fresh eyes would be helpful.

Thanks for your feedback!

TT

[cdemike]

I think you're right that the increase in bias voltage likely is the culprit. Your oscillator's amplitude didn't change, so with the larger overall bias voltage, the proportion of swing for trem is lesser. How did you settle on a 100k drain resistor? I've typically seen larger for trem oscillators, and even though yours is set up as an amplifier for an oscillating signal, getting more output there would get you some of the way. I did some loadline drawings for a project where I considered using an LND150 input stage, and I've attached the load line document I worked from. You have a pretty hearty B+ supply, so increasing the drain resistors to 220k, 300k, or even 470k would be my first plan, but that may require adjustment to the source resistor's value as well. I don't know if it'll make as much difference for a non-audio signal, but you could also try bypassing the source resistor with a capacitor.

If that doesn't get you there, you could also try using a source follower after the amplifying stage to lower output impedance.

[sluckey]

R63 and R64 voltage divider only allows 60% of the trem signal to reach Q6. Making R63 larger will give more trem signal.

[tictac]

Thanks for those ideas sluckey,

I was thinking about making R63 larger but I had previously changed R72 to 100k and was getting nasty distortion from Q6 output as a result. I'm thinking upping R63 may cause similar issues. I could raise it a small amount and see how it goes. I really like your idea of raising R65 to 220k. Maybe 500R for R62?

TT

[Stevem]

Well -33 is well in the ball park for what a stock PR wants!

[FUCHSAUDIO]

You forget the EL84 and 6V6 have a different bias point, so the modulation will be weaker with a tube needing more negative voltage (and more tremolo swing too), to get the same effect.

You could raised R-72 to get more output from the chip (clever circuit btw), or toy with the divider into the LND-150 (R-63 and 64) to get more swing to the grids as well. Maybe take R50 to 220-K to load the tremolo signal a little less.

Have fun.

[cdemike]

(clever circuit btw)

I forgot to comment on this initially, but I'm also really enjoying looking over this circuit. Thanks for posting, tictac! I'm looking over the ICL8038 datasheet now. I'd love to see how you configured your power supply and regulated the oscillation frequency if you're willing to post it.

[martin manning]

Not sure what the amplitude coming out of the 8038 will be, but guessing 10V p-p I would try something like this. It should get the positive swing up to ~0V at max intensity. The right side is 6G9-B Tremolux.

[tictac]

I'm back,

Finally got into the amp...

Adding a bypass cap to R62 did the trick. The trem doesn't become audible until the Intensity pot is around 11 o'clock but it has plenty of depth as you turn it up and all the useful speed is there.

I think I'm going to leave as is. Making further changes requires removal of the entire pcb which is a time-consuming chore

TT

[tictac]

I can't take credit for this design but here's more of the trem circuit...

TT

[lonote]

Is the trem circuit on it's own board & just added to the PR or did you mod the existing PR layout to fit it?

[tictac]

The amp is not a PR but is similar. The pcb contains all circuitry for the entire amp

[lonote]

Got it, I thought it might have been an add-on.

Thx