Who fights the power that fights the power?

[TransLucid]

I would like to suggest a load-line analysis approach, because I think the load line will show fairly clearly what will happen in overdrive. If you change the B+, redraw the load line and see what happens at each end...

On another point, preamp stages are class A and therefore draw constant average current when operating linearly. So, as mentioned previously, the only sag you'll get will be due to the output stage dragging down the whole supply. However, I could imagine an overdriven class A stage drawing more current during the positive half of the input waveform due to grid conduction, causing it's own sag.

Bob

[DIY_Guy]

You are running a lot off that one power supply node. I usually give each dual triode it's own node in the power supply. It's just what I'm used to seeing in most amp builds, and most of the books I've read seem to recommend that number.

[Chuck H]

Two things that will really wake that amp up (and possibly correct some tonal problems). You are running a whole bunch (technical term) of NFB. Waaay more than normal. That 1500R shunt at the bottom of the PI should be changed to a MUCH lower value. If your OT is 8 ohms try a 47R resistor there. You should also change the dual ganged master to a 250k value. That will put you in the right range for the vintage amps your emulating. Your voltages look fine, but I would agree that another decoupling stage for the "voxish" channel is a good idea. Then maybe a twisted pair filament wiring elevated on the 6V6 cathode. Looks like it'll be a super fun amp when your done tweaking it.

[jfutral]

Yes on the 8 ohm OT. Would changing the 1500 to 47 be the only change suggested? Or also reduce the 47k? I can't for the life of me remember where that came into play. Maybe when I was thinking plexi at one time.

Funny thing, the only 47R I have is a 5 watt. I went with two 1/2w 100Rs in parallel. I figure 50R is pretty darn close, especially at +/-10%.

For the Vox decoupling, should that be a branch from the same point as the Fendish? Or series? Keep with the 10k dropping resistor?

Still want to learn more about the mechanics behind the power supply.

I had considered the larger value for the MV at one point, but since I am clueless, I never moved on it. Kind of the story of my life. Wait. Wow, that was sad even for me. I never considered myself much of a nihilist before.

Thanks,
Joe

[d95err]

Why decouple the stages? Only to adjust voltages along the chain?

Each decoupled stage adds better filtering of the DC, resulting in less noise. Also, if a lot of gain stages run from the same power supply node, you can get feedback through the power supply which can cause oscillations.

[Chuck H]

Yes on the 8 ohm OT. Would changing the 1500 to 47 be the only change suggested? Or also reduce the 47k? I can't for the life of me remember where that came into play. Maybe when I was thinking plexi at one time.

You could try a 680R/22k tail. Probably closer to what you need than 470R/47k. Just ball parking though.

Funny thing, the only 47R I have is a 5 watt. I went with two 1/2w 100Rs in parallel. I figure 50R is pretty darn close, especially at +/-10%.

Amp got louder didn't it? Probably also sounds smoother and less buzzy too.

For the Vox decoupling, should that be a branch from the same point as the Fendish? Or series? Keep with the 10k dropping resistor?

You can just Y two 10k R's at that last node and give each branch it's own decoupling cap. 22uf for the Fenderish side and 8uf for the Voxish side.

Still want to learn more about the mechanics behind the power supply.

You can download Duncan PSU at their website. It's a pre programmed PSU spice program. Also, look up "tube amp FAQ' on GEO, Aiken's technical info and the Valve Wizard web sites. Enough to keep you busy for a looong time.

I had considered the larger value for the MV at one point, but since I am clueless, I never moved on it. Kind of the story of my life. Wait. Wow, that was sad even for me. I never considered myself much of a nihilist before.

The standard value used here for Vox and Fender amps is 220k for the grid load/bias feed resistors that your MV is replacing. The amp will have more dynamics with the higher value.

[jfutral]

well, my schematic was wrong. I have a 47k resistor on the NFB, not 820. Not sure where that came from. Added the second branch on the B+ and was surprised at the outcome. The voltage on the Fender side is now around 188-190v. Tightened up the sound somewhat, particularly when I push into overdrive. But by far (surprisingly to me) the biggest help in the over all tone was moving to 250k dual pot for the MV.

Also after reworking the wiring for the heater leads didn't seem to help my (albeit very faint) hum I decided to add 100R resistors. I wasn't sure if I should leave the center tap connected or not, but I did anyway. The hum is all but gone. I'm very happy with that. I'll post an updated schem on another thread I have been discussing the noise.

Anyway, my volume of TUT came and now I am wondering about adding a reverb (because the cabinet has a tank in the bottom) and an FX loop. I thought about maybe a foldback reverb, but I think I may just invest the time into a standard reverb because I want the FX loop after the reverb. And if I don't try now, when will I?

Joe