Bugera 333XL doides control grid to ground?

[Alex R]

Have a look at this - power stage on a Bugera 333XL. What are D1 and D2 for? Control grid to ground? I know they'd only remove half the wave each side but... well I've never seen them before and don't get the theory. I guess I've always thought of the PI as not splitting the wave in two quite as firmly as these would.

[pdf64]

I think that the diodes are probably just ensuring that the control grids can't become positively biased under any drive conditions from the LTP.
Thereby preventing excessive cathode current / plate dissiption.

[Alex R]

*diodes lol

[Alex R]

Thanks, yes fair enough - I must be being dim - but what about the signal?

[Enzo]

The grids are sitting there biased to -45v or whatever. So those diodes only conduct if the signal peaks are larger than 45v on the positive side. Most of the time that won't happen. But if you did manage to shove an over 90v p-p signal into those grids, the diodes would turn on for the positive peaks over 45v. SO your signal would have positive peaks clipped off. The purpose of the diodes is to prevent the grids going positive and drawing current.

Keep your signal below 90v p-p and the diodes will never do anything. They ONLY conduct for the portion of the signal OVER 45v.


And in every place I wrote 45v, you can change to whatever the actual bias voltage is in your amp.

[Malcolm Irving]

What does the 'BIAS SELECT' switch do? I can't see clearly from the schematic, but if it switches between 'fixed bias' and 'cathode bias' the diodes could be needed to prevent the bias getting out-of-control (positive) during the switch over. (Just speculating here.)

[Dave H]

Thanks, yes fair enough - I must be being dim - but what about the signal?

At first sight it looks like the positive going half of the signal will be clipped by the diode (as you said in the OP) but the signal is AC coupled from the PI to the negative bias voltage so the signal swings symmetrically above and below the bias voltage not symmetrically above and below zero. When the signal peak exceeds the bias voltage it is clipped by the diode but it would still be clipped by the tube grid without the diode.

The diodes don't "remove half the wave each side" from the PI outputs. Both PI outputs are the full wave with one output inverted (180deg out of phase)

[Malcolm Irving]

What does the 'BIAS SELECT' switch do? I can't see clearly from the schematic, but if it switches between 'fixed bias' and 'cathode bias' the diodes could be needed to prevent the bias getting out-of-control (positive) during the switch over. (Just speculating here.)

OK. It switches between two fixed bias levels (for EL34s and 6L6s) but maybe the bias could get a bit undefined when the switch is half-way between the two?

[Enzo]

The diodes are not related to the bias switch.

[Malcolm Irving]

The diodes are not related to the bias switch.

Yes, I agree. I think the bias switch is safe in the half-way position.

[Enzo]

Look at the schematic. The bias switch is a slide switch, not that that matters, but it shorts across a resistor. So there is no half way contact condition. It moves until the resistor shorts. So either resistor is shorted or not. There is no open condition in the middle for example.

SO setting the switch in the middle will result in one bias or the other. It may wind up the wrong one for the tubes that are installed, but the tubes will still have bias.

[Gingertube]

The diodes will do a couple of things.
They limit the +ve grid swing on the output tubes.
By doing that they:
1) limit grid current (grid rectification current) in the output tubes (prevent or at least limit drive into AB2 operation)
2) limit peak anode current swing (reduction of maybe as much as 25% - inferred from the datasheet graphs for an EL34).
So a "light" output tube protection thing and maybe some "light" output tranny protection too.

They will NOT help blocking distortion as the coupling caps charge up from grid recification current will happen from diode forward current instead.

Cheers,
Ian


P.S.
Looks like it was "stolen" from the Peavey JSX 120 (what a surprise - NOT)
http://www.thetubestore.com/lib/thet...-Schematic.pdf
SCroll down to the power board schematic - it is the same circuit.

[teemuk]

It's a Peavey circuit, naturally.

As said, those diodes do not distinctly affect "grid clipping" (and associated effects) because they basically just replace the grid conduction (and associated clipping) with clipping of the diode, which has about similar characteristics.

Peavey designs generally have aimed into featuring characteristic effects associated with grid conduction (bias shift and dynamically increasing crossover distortion) and they even implement them to several of their solid-state amplifiers that "model" tube amplifiers.

However, and this has been mentioned earlier as well, eliminating actual grid conduction has distinct impact to screen characteristics: During grid conduction the screen begins to draw substantial amounts more of current, and in interaction with other circuit behaviour (such as gain compression) this does not always produce very "well-behaved" overdrive characteristics and may even lead to instability. And it's "harder" for the tubes too. So this was a characteristic Peavey designers obviously wanted to avoid.

[Gingertube]

That screen current limiting thing makes sense. I'm a hobbiest (EE in the day job) but over 20 years I've repaired/restored maybe 15-20 amps with output stage problems. Melted screen grid resistors have been apparent in at least half of those.
On old Fenders swapping out the 470 Ohm 2W Screen grid restors is usually the first thing I do (for 1K 5W Wire Wound) - seen the old 470 Ohm resistors measure up to 1K (ignoring the ones which were NOT open circuit that is).
Also seen lots of 6L6GC tubes with noisy sagging screens.

Probably more important for amps with EL34, the true pentode EL34 has higher screen currents typically than the beam power tetrode 6L6 style tubes. That has given rise to the EL34 having a reputation for fragile screens. Their screens are actually no more fragile they are just subject to higher stress.

Cheers,
Ian

If it is good enough for the master copier, then I'll try copying it too for my next build.

[Mike Sulzer]

What is that extra tube plate labeled F to the right of the lower left EL34?

Another question is what is the sonic difference between a.) limiting screen grid dissipation with grid-to-ground resistors, and b.) raising the screen resistors?