Univox U45b Tremolo

[hylaphone]

Hi folks, I have a U45b on the bench, sounding pretty good after replacing a handful of drifted resistors and leaky caps. Ditched the grid leak bias on V1 for a more familiar cathode bypass, and I'm happy with the results.
I also slowed down the tremolo, it's nice and deep but imparts some very harsh distortion. Voltages look ok, I've tried swapping tubes, adding a grid stopper to the PI, etc. Can anyone suggest a remedy?
I don't have a scope to play with, unfortunately.
Had good luck with the Paul C Mod on another bias-vary-trem amp with a similar problem, but I am unsure how the 6BM8's would react to the higher cathode voltage?

[Jazz P Bass]

Is the distortion apparent at lower volumes?
The reason I ask is the fact that the Tremelo, at higher volumes, can drive the output into distortion.
If you are playing clean at a loud volume and then engage the Trem, the Trem can put the signal 'over the edge'.

[hylaphone]

Yeah it's only at higher volumes. Is there a way to tame the distortion while retaining the trem depth?

[Jazz P Bass]

The Trem is driving the amp into clipping.
That is a consequence of the design.
You are most probably playing right on the edge of distortion & the Trem is taking it over the top.
If only the bass notes are driving the amp into clipping when the Trem is engaged, you could try a lower value cap in C15.

[Enzo]

Or maybe reduce R21?

[hylaphone]

Good suggestions. After swapping a few values, I'm realizing it's a fine balance between taming the clipping, and losing the lower speed trem entirely.
Too bad, the amp is really impressive otherwise!

[Chuck H]

Is this a garbled, phasey sort of distortion or just clipping? I ask because the trem takes the bias in both directions. Not only can it cause the power tubes to clip, but it can also cause them to cut off. In fact it SHOULD do both when a high intensity is used. Check the actual bias of the power tubes. Experiment within a safe bias range and the trem intensity and hopefully find a sweet spot.