Ok if I recap a 8uf bias cap with 47uf ? (Traynor YGM-3)

[Iplayloud]

I'm recapping this old Traynor YGM-3 that was motorboating terribly and I'm all done except for the first bias cap C27 right after the diode. In this lockdown situation my local electronic shop is closed. I could wait for a mail order but.. On standby the voltage peaks at 59v, all I have anywhere close is a 47uf 100v cap. Thinking of the Deluxe Reverb, whose bias cap went from 25 to 80 over over the years I'm thinking it should be fine. My only concern is maybe it interacts with the tremolo circuit? Honestly I never really understood those tremolo circuits... the bias seems to come from two paths at different voltages. It's a bias tremolo so maybe it swings between both paths and C27 vslue might affect it? Any input appreciated. Thx

Schematic:
https://music-electronics-forum.com/attachment.php?attachmentid=51588&d=1545229903

[nickb]

47uF 100V will work just fine.

[Jon Snell]

You may have to adjust the value of R2 as 47u is a larger 'tank' value than the original 8u and may store a different voltage.
Check it when fitted for the correct bias voltage. It may be OK.

[nickb]

You may have to adjust the value of R2 as 47u is a larger 'tank' value than the original 8u and may store a different voltage.
Check it when fitted for the correct bias voltage. It may be OK.

I think you mean R44 (150K). The difference in voltage is rather small, around 350mV. That's about 2% so that means you'd have to use a 1% resistor to affect a change. It really isn't that critical. OTOH the ripple is much reduced, a good thing.

[Jon Snell]

I think you mean R44 (150K). The difference in voltage is rather small, around 350mV. That's about 2% so that means you'd have to use a 1% resistor to affect a change. It really isn't that critical. OTOH the ripple is much reduced, a good thing.

[ATTACH=CONFIG]57991[/ATTACH]

Note to self; Must engage brain before assuming the wrong rersistor.
Of course, that is what I meant, although, I have no idea where R2 came from. Senior moment🤣

[Iplayloud]

Ok yes there's a mistake in the redrawn schematic. R2 is there twice, back to original schematic. Thanks for the quick replies!

[Jon Snell]

Issue solved. It wasn't me at all!

[galaxiex]

Ok yes there's a mistake in the redrawn schematic. R2 is there twice, back to original schematic. Thanks for the quick replies!

Yes, mistakes in that schem.... R26 is also there twice.
R26 in the bias circuit and R26 grid of V4

I think the extra R2 is supposed to be R twenty something...

I found those mistakes when working on my YGM III.

[Helmholtz]

The 150k series resistor together with a 47µ cap make for a very long ramp-up time of bias voltage (maybe 15s to 20s).
This may cause the power tubes to redflash after turn-on.

I think it makes sense to go with lower bias cap values in circuits where the bias supply is derived from a HV winding and a large series resistor is used. The large series resistance takes care of sufficient filtering with lower capacitance.

OTOH, higher capacitance is desirable in amps with a dedicated bias winding and lower series resistance.

[nickb]

The 150k series resistor together with a 47µ cap mean a very long ramp-up time of bias voltage (maybe 15s to 20s).
This may cause the power tubes to redflash after turn-on.

I think it makes sense to go with lower bias cap values in circuits where the bias supply is derived from a HV winding and a large series resistor is used. The large series resistance takes care of sufficient filtering with lower capacitance.

OTOH, higher capacitance is desirable in amps with a dedicated bias winding and lower series resistance.

The ramp up time is only ten seconds. It's fine.

[Helmholtz]

The ramp up time is only ten seconds. It's fine.

Did you simulate or measure ramp up time?

According to theory the 90% voltage value is reached after t = 2.2*R*C, which woud be 15.5s. The second filter stage and the shunt resistor tend to further slow down bias ramp up.

If the delayed bias ramp up is critical will depend on HV ramp up time.

[nickb]

Did you simulate or measure ramp up time?

According to theory the 90% voltage value is reached after t = 2.2*R*C, which woud be 15.5s. The second filter stage and the shunt resistor will further slow down ramp up.

If the delayed bias ramp up is critical will depend on HV ramp up time.

I simulated. Both 8u and 47u have pretty much reached the final value after ten seconds.

The circuit is too complicated for such a simple 2.2RC formula. If it were DC on one side of the diode the R would be the Thevinin equivalent i.e around 28k so that (2.2RC) would be ~3 secs ( ignoring the second RC network). It's not DC (current) but chopped half sine waves so takes quite a bit longer.

Here is the plot showing 8uf and 47uf:

[Helmholtz]

I simulated. Both 8u and 47u have pretty much reached the final value after ten seconds.

Thanks!

It's not DC (current) but chopped half sine

You're right, that certainly makes a difference.

If it were DC on one side of the diode the R would be the Thevinin equivalent i.e around 28k so that (2.2RC) would be ~3 secs

I don't think the Thevenin source equivalent applies here. Any (additional) load current must delay and not speed up the initial charging of the 47µ cap.


Anyway, I certainly don't doubt your simulation results.

[nickb]

Thanks!



You're right, that certainly makes a difference.



I don't think the Thevenin source equivalent applies here. Any (additional) load current must delay and not speed up the initial charging of the 47µ cap.


Anyway, I certainly don't doubt your simulation results.

Here is a Thevenin equivalent compared. Looks the same to me ...




At this point the OP is probably banging his head on the workbench wondering what he did to cause all this. I liked my original answer. "It'll be fine"

[Helmholtz]

Thanks, seems you were right and I was wrong. (I definitely have to improve my simulation skills.) After all I'm here to learn something and I think I understand my fundamental error now.

To the OP the difference is an estimated 15s or a simulated 10s bias ramp up.