Shunting signal to B+ instead of ground

[anson]

In the past, I've usually shunted tone shaping circuits to ground, such as tonestacks, snd have only used B+ for shunting signal in the form of using a small "smoothing capacitor" across a plate resistor. I think I've read somewhere that a change in output impedance is associated with shunting to B+ as opposed to ground, and someone on this forum (I forget who) praised using B+ instead of ground, and so I thought I might as well try it out and see how it sounds.

I don't have a typical fender style tone stack, and instead use several tonestacks throughout the preamp, but I've sent them all to B+ instead of ground (using caps to block DC where necessary), and have also conected cathode bypass caps to B+ instead of ground (using high voltage voltage caps).

My first impression with all this is that is sounds great! I did it all one cap or tone circuit at a time, to see what role each cap has, and the sound to my ears just kept getting better. Clearer, more beautiful. Even overdriven tones. I connected the tone circuits back to ground, and there was a nticeable increase in "deadness" to the tone, so I connected it all back to B+ again.

Any body else try these kind of things, or would like to shed some light of the principles at work here?

One thing I should add is that the power supply for my preamp is a kindof star-power supply, so each pair of 12ax7 stages has its own power supply tap off of the B+ ( where the EL34 screens get their voltage), and I shunt the tone circuits to the high voltage side of the plate resistor from which the signal came.

[wizard333]

I'm a big fan of taking my high end bleeder caps across the PR to B+ instead of ground (which technically should be the same since the AC sees ground through the big-ass filter caps, but sonically and feel wise, not the same at all, for technical reasons I don't understand but are all too obvious to my ears and fingers) for all the reasons you mentioned.........

But I've never tried taking EVERYTHING, including CKs, to ground through the power rail. Very interesting..........

[Steve Conner]

In theory, B+ and ground are the same thing from a signal perspective, because they differ only in their DC potential: neither has any signal on it.

In practice, the B+ rail may have crud, hash and low-frequency wiggles riding on it that could easily be mistaken for an increase in clarity and liveliness. If you have an oscilloscope, set it to AC coupling, clip the probe tip to the relevant B+ rail (x10 probe for safety's sake please!) and the ground clip to the preamp ground bus, and turn the sensitivity right up. You'll see just what you're shoving into your cathodes and so on.

If you take this practice of returning things to B+ to extremes, you could convince yourself that B+ is really ground and the amp runs off -250V or whatever. :-) Look at what Keeley did to make the Java Boost, with its PNP germanium transistor, run off +9V: it's the same principle at work.

[Enzo]

There are some old Seeburg jukebox tube circuits that do exactly that - the plates are grounded and the cathodes sit at B-.


Ity would seem that if your cathode bypass cap went to B+, then instead of a 25v cap, you'd need a 500v cap. Takes up more room, costs more.

[Steve Conner]

There are some old Seeburg jukebox tube circuits that do exactly that - the plates are grounded and the cathodes sit at B-.

That's interesting Enzo, can you give us any idea WHY they did that?

I'm still waiting for P-channel vacuum tubes, but I suspect they'd need to be made of antimatter :-)

[wizard333]

What Steve said is something that I know, as I said above: the AC instrument signal should see B+ essentially as ground because its seeing ground through a 22uf cap in most cases, plenty large enough to pass 100% of that frequency content right to the chassis.

I've never taken anything but a bleeder cap to B+, but what I still can't figure out is: Why the sonic and tactile results of doing that are so incredibly different from taking the same cap right to ground. In practice, its not even close. Bleeder caps to ground are hateful touch robbing monsters; caps across the plate resistor are much more finger and ear friendly.

The low end ripple hanging out on the pre-amp B+ is at most a few millivolts in any properly filtered amp, and the caps in question are so small that they wouldn't pass anything in that frequency but maybe microvolts. I also can't see how a few microvolts of low end ripple poop could make anything sound good.

And yeah Enzo I had the same thought; his CKs must be HUGE to tolerate that.

[Merlinb]

The low end ripple hanging out on the pre-amp B+ is at most a few millivolts in any properly filtered amp, and the caps in question are so small that they wouldn't pass anything in that frequency but maybe microvolts.

If the caps are large enough to shunt the audio bass signals away then they are also large enough to let PSU ripple in!

[wizard333]

But they arent.

The caps I use for bleeders are at most .0022, usually .oo15 or .001.

Thats going to roll off that milivolts of PSU ripple hard. Its going to end up less than a millivolt, probably a lot less.

And I still dont get how that would make anything sound better?

[Merlinb]

But they arent.
And I still dont get how that would make anything sound better?

I meant the tone stack caps. However, there can still be high frequency noise on the PSU too.

Another possibility is that for 2 triodes running from the same PSU cap, if the PSU caps has some ESR or some questionable high-frequency impedance, then by adding anode caps you are effectively reducing the impedance between the anodes of each stage- effectively creating HF negative feedback between the anodes.

[wizard333]

Hmmmmmm........

[Enzo]

I'm still waiting for P-channel vacuum tubes, but I suspect they'd need to be made of antimatter

P-channel as in positive pressure. or positive vacuum if you will. These tubes have 30PSI inside the tube. the electroncs have to fight their way to the plates, and in so doing they achieve a sort of pride that shows in their sound.

Seeburg negative rail circuits were in the control section, not the audio. Not sure just why. The tubes were used as pulse detector/amplifiers in a core memory unit (Seeburg "Tormat.") but I fail to see how it matters there either. And the system also controlled either a 2050 or a 2D21 thyratron tube - a tube version of a triac. But again I don;t know why polarity mattered. I enjoyed the core memory stuff.

[RudeBoy]

That's interesting Enzo, can you give us any idea WHY they did that?

I'm still waiting for P-channel vacuum tubes, but I suspect they'd need to be made of antimatter :-)

The amp was designed by Lucas?

[anson]

In practice, the B+ rail may have crud, hash and low-frequency wiggles riding on it that could easily be mistaken for an increase in clarity and liveliness.

...I was suspicius of that.


There are three preamp stages in my amp that have bypass caps: two early ones with 0.47 and 0.22u 630v poly caps, and a later stage with a 22u non polarized electrolytic that isn't actually bypassed to B+ right now for the reason of space and $$.

The caps that connect my tone stacks to B+ are 0.022 and 0.047u.

One capacitor that I put to B+ that did NOT have a positive affect on sound (in my opinion) was the 0.1u grid cap for the second stage of an LTP PI. I put it to B+ and the amp still worked, but seemed to have less sustain, tone, balls, etc. It seems to make more theoretical sense anyway to have that cap go to where it normally does, to the junction of the tial resistor and NFB shunt resistor, because that way current can return back to the output tranny (as there doesn't seem to be any reason why that current should play with B+ circuits), and also doesn't have to go through any elecrolytic caps on its return home.

[Merlinb]

It seems to make more theoretical sense anyway to have that cap go to where it normally does, to the junction of the tial resistor and NFB shunt resistor,

Don't forget that by moving that cap to the B+ you basically removed your NFB, and left the PI with a kind of mild NFB and PFB combination.

[Steve Conner]

The amp was designed by Lucas?

Well, does it have trouble starting in damp conditions?