Drive voltage vs. bias voltage

[Chuck H]

I was taking some readings on one of my amps (2xel84 cathode biased) and I noticed that the tubes start clipping with 8.7VAC signal at the grid and the cathode is at 12.7VDC. I thought I read here many times that the bias voltage dictates the grid voltage at clipping and that they are ideally analogous. When I consider what I "think" I know about tubes it made sense. I guess I just never checked the right things in the right order to notice this before and I've never specifically looked for it having taken what I've learned here at it's word. But I'll bet there's a little caveat in the theorized ideal operation and if anyone knows what it is I hope you'll chime in.

I've been basing the fine tuning of my zener bias tweaks for el84 cathode biased amps on erroneous information for almost a decade. And now I feel pretty stupid for never checking the grid drive voltage in relation to the cathode voltage!?!

TIA

[catalin gramada]

it was rms? then have a lot of sense 8.7x1.4142 means 12.3v peak, pretty close what you get

[Leo_Gnardo]

Yup that's how it works. Try to drive the grid positive with respect to cathode, you'll find that most drive schemes can't supply the necessary current and now you're in clip territory. The grid starts to act like a biased-on diode at that point and draws mad current. Some of our hard headed friends in hi fi have devised methods to successfully do this (scan Audio Asylum) but it takes a strong drive circuit. Can't say I've seen such a thing in regular guitar amps. Maybe in bass behemoths like Fender 400PS but they are in a unique class and rare enough I've never seen one in 50 years of amp experience.

[Chuck H]

Ah! Yes. I suppose if I ever did check it I was using my old DMM which wasn't an RMS meter. The meter I'm using now is. That's probably at least part of my confusion (albeit a small part )

Just for kicks I measured the grid to cathode voltage when I was doing the test. I knew there would be at least a small difference and I know, of course that the grid to cathode voltage is the ACTUAL bias voltage. 12.37V

[Leo_Gnardo]

Gotta confess I wasn't hip to this until maybe 6 or 7 years ago I was reading Audio Asylum threads on design & building class B2 amps, just curious how those mad scientists went about it. For the rest of us, no need to go there.

It does shed some light though on how power figures reduce when you have tired old tubes and drop bias voltage well below common values to try & panel-beat them into continuing to work. When I see some old veteran 6L6 amp, bias voltage 39 volts it's a sure bet fresh tubes & proper biasing will perk up the power figure by 20 or 30 percent. OTOH some blues players prefer the sound of their clapped-out tubes.

[Chuck H]

I don't think I've ever had the opportunity to babysit aging tubes. So couldn't say about their tired sound vs. the sound of newer tubes. I've had some aged tubes in several amps I've owned or serviced, but never any that exhibited a "tired" sound. My own experience has been that they fail before that. That could be a byproduct of my time in this arena starting about 22 years ago (instead of a hundred like some of you guys ). Nearly all the tubes I've worked with were Russian or Chinese. There were a few southern Euro tubes around. I remember my old brown base Tesla's lasted A LONG time with steady, but light use and I eventually sold that amp with those tubes in it. Any other Euro or American tubes I've seen were in amps so old that one or more of the set was dead before the amp got to me. But I could believe the older Euro and American tubes were probably longer lived die hards that could be pushed into longer service. Not so with tubes of the last two decades I think.

[Dave H]

I suppose if I ever did check it I was using my old DMM which wasn't an RMS meter. The meter I'm using now is. That's probably at least part of my confusion

Even if it's not true RMS the old meter will be calibrated RMS for a sinewave so as long as you are measuring a sinewave the two meters should read the same.

[Enzo]

yep, even if the RMS reading isn't accurate, it will be close enough, and the instantaneous voltage at peak will be roughly 1.414 times the "AC" reading. As was explained in post #2, that means your signal was indeed clipping at roughly the bias voltage.

[R.G.]

Can't say I've seen such a thing in regular guitar amps. .

Probably not commercial ones. But several people here and in other fora have used a MOSFET follower for harder drive of an output tube grid.

[Chuck H]

I've used a cathodyne followed by a pair of shunt NFB triodes capable of about half the gain of a LTP but you can adjust the feedback amount to favor lower impedance or gain depending on what works best for a given amp. It worked pretty well to eliminate crossover distortion in an amp designed to drive the piss out of a pair of el84's Three triodes vs. four for an LTP and cathode followers. Of course MOSFETs are probably cheaper and easier. But if one were interested in the "all tube" thing...

[loudthud]

Chuck, the time has come for you to obtain and use an oscilloscope. Things that seem mysterious now will become clear to you. Soon you won't know how you ever got along without it.

[Chuck H]

I have one! I was using it to monitor the output for clipping and using my 87 to measure voltages. Had I put the scope on the grids, well...

[pdf64]

It's great to have those eureka moments, when everything clicks into place and the fog lifts

[catalin gramada]

and one disclaimer. cannot pick corect voltage mesurement into big impedance node with a regular cheap voltmeter due to usually low internal resistance means how much dmm internal resistance afect the circuit you measure. sometimes analogs works better

[g1]

I have one! I was using it to monitor the output for clipping and using my 87 to measure voltages. Had I put the scope on the grids, well...

I thought you had one, but figured maybe you were just being stubborn.