How do supply voltages affect power tube performance?

[bepone]

generally speaking , changing the supply voltage is changing the amp character.. preamp is working differently, in different working points, output tubes have more gain and also new working points, where load /output transformer/ speaker is passing through different set of curves.

never heard about Van Halen "brown sound"? with the amp working at lower voltages?

it is NOT only volume change.. power reducing can be better if preamp voltage remain untouched, and varying only output tubes voltage.. still output tubes will work differently but is doable..

but 6V6's working at 320V have good sound, no worries.

[roberto]

6V6's working at 320V have good sound, no worries.

yep, I agree.

Another point that has not been touched until now is that, keeping the loadline stable (10k or 8k is standard for 6V6GT in PP) and the bias stable at the same percentage of max plate dissipation (80 to 90% is common with those loadlines), a lower voltage means higher currents, so the amp will stay in class A for a wider part of its swing (that could be only class A in some cases, even if it's very uncommon in guitar amps).
The power supply will work with less variations of current, so less variation of voltages even when pushed.

[martin manning]

...power reducing can be better if preamp voltage remain untouched, and varying only output tubes voltage..

Yes I agree. The question posed here is about output tubes, though, where the reduced screen voltage more-or-less preserves load-line location wrt the Vg1=0 knee and headroom. Triodes in the preamp will have reduced headroom as voltage is reduced, which definitely comes into play when they are over-driven. When running lower output tube voltage there is the option of adjusting the preamp voltages with the dropping resistors, within limits of course.

OK, If you have to use 125P23B I would consider using the 540V taps and replacing the Princeton's 5U4 with a 5AR4/GZ34. That should increase the B+ voltage 30V+.

Thanks Martin! I'll try this idea. Fun to play and learn. Thanks for the advice.

You'll need to see where you land with the PT you are using. I see Hammond's PR replacement PT (290AX) has 340 and 275 VAC (680 and 550), which would provide options using different rectifiers, even going solid-state using the low voltage taps. BTW, Fender's '65 Princeton Reverb (the reissue) is running 440VDC on its 6V6's.

[Magnatron]

You'll need to see where you land with the PT you are using. I see Hammond's PR replacement PT (290AX) has 340 and 275 VAC (680 and 550), which would provide options using different rectifiers, even going solid-state using the low voltage taps. BTW, Fender's '65 Princeton Reverb (the reissue) is running 440VDC on its 6V6's.

This is all really fascinating to me and it's obvious I have so much more to learn. I dug out my "Designing Tube Preamps for Guitar and Bass" book and started another attempt to plow through it.

So this discussion has me wondering as well: I assume the higher voltage to the OT with a 680 volt winding would also mean higher voltages to the anode resistors of the preamps tubes as well? For this model, they are the common 100k resistors. Would any adjustments to those be necessary? Would the higher voltages to the preamp anodes change the character of the amp in any noticeable way? Also, I have installed a three-way cathode bias mod to the first tube, going between straight 1.5k resistance to adding a .25uf or a .47uf cap. Would these selections need to be re-examined with the higher voltages as well?

This has all been highly educational! Thanks for your insight thus far!

[martin manning]

No those are the plate loads, and they should stay the same. There are two 18k resistors (one on the board, and one on the can capacitor). Those would be a good places to trim the preamp voltages to get the original 160V on the V1 plates.

[Roe]

lower voltage gives slower response, earlier breakup and more compression with more mids.

... and an after taste reminiscent of cloves...

Got numbers?

?

[martin manning]

Also, I have installed a three-way cathode bias mod to the first tube, going between straight 1.5k resistance to adding a .25uf or a .47uf cap. Would these selections need to be re-examined with the higher voltages as well?

If you trim the dropping resistors to restore the V1 plate voltages the preamp will operate as normal. The values you quote for cathode resistor bypass caps are small, and will cut the bass well into the mid range. The original used 25u, which pretty much fully bypasses the cathode resistor in the audible range. Or, maybe that's a typo? If you meant 25u and 47u, you will probably not hear much difference.

[Magnatron]

If you trim the dropping resistors to restore the V1 plate voltages the preamp will operate as normal. The values you quote for cathode resistor bypass caps are small, and will cut the bass well into the mid range. The original used 25u, which pretty much fully bypasses the cathode resistor in the audible range. Or, maybe that's a typo? If you meant 25u and 47u, you will probably not hear much difference.

The mod circuit I’m using comes from a Two Stroke amp I built a couple of years ago. I liked the sounds so much I thought I would try it on this amp. You’re right, the .25 was a typo. Here is the portion of that circuit:

[WiderGates]

Hello,
your Input switch off R3

[roberto]

The shelving with that configuration has 5dB extension (between bypassed and not), and the -3dB point is around 700 Hz for the 250 nF and 400 Hz for the 470 nF.
It is uncommon but not the first time I see these values (I've used them too in higher gain amps), specially with low wattage amps, to avoid wasting power in a range of the bandwidth that is not of primary importance.

[sluckey]

better

[martin manning]

better

I'd say "typical." The 1M can be disconnected or not, the difference being the input impedance seen by the guitar pickup, which will make a small difference in its resonant peak. The tube's grid is going to be looking at ~10k either way.

[bepone]

for normal / specially low noise application, grid leak resistor must be close to the dedicated cathode resistor, not on the 2m cable away from the first tube (pickup).

[roberto]

Not really. You are live, you want to change the guitar so you unplug your guitar but not the amp, then the grid is no more referenced to ground.

[martin manning]

Not really. You are live, you want to change the guitar so you unplug your guitar but not the amp, then the grid is no more referenced to ground.

Yes, good point.