Max Grid Resistance - Preamp Tubes

[potatofarmer]

Hey, longtime lurker, finally got off my butt and joined up.

I have a quick question about a 6C4 that I'm using as a cathodyne.

I put in a 470k grid stopper but I'm still having trouble with the frequency-doubling effect, so I'm considering increasing that to 1M. My only problem is that the datasheet lists 1M as the maximum grid circuit resistance. 6C4s are cheap (now and probably for a long time) but they aren't in current production, but on the whole I'd rather not burn up tubes if I can avoid it.

Should I just stop worrying? I'm thinking about doing the math and seeing how low I can go for the master volume - the input impedance is around 1.5M currently so I'm thinking a 250k or even 100k pot might not load down the previous stage too much.

I've attached the schematic as it's currently built, though there are some tweaks I still have to do (cut bass, tweak attenuation, etc.). The voltages on the schematic are inaccurate; they were just estimated when I was designing the power supply. My maximum plate dissipation is probably in the neighborhood of 7 mA (iirc the B+ is somewhere in the 350V - 400V range) so I'm a little more concerned than if the 6C4 were used as a typical gain stage.

Thanks!

[kleuck]

In this configuration, the input impedance of the 6C4 is boostrapped, so it should not be an issue to use 1M grid stopper or even higher.
A lot of people report that 1M5 is just fine with 12AX7 concertinas.

[potatofarmer]

Yeah, high frequency loss isn't my concern, though I should double check the math anyway. I'm more worried about what happens as this tube ages and starts to get gassy, what with the ions and slamming into the grid and the thermal runaway and whatnot.

Also I believe 12AX7 datasheets usually specify something absurdly large for maximum grid leak resistance when cathode biased. 22 megohms or something?

A 6C4 is basically one triode of a 12AU7 in a 7-pin envelope; it can push around more current than a 12AX7.

[exclamationmark]

Thermal runaway shouldn't be a problem with pre-amp tubes as they have both a plate resistor and cathode resistor to combat the effect - they're constrained to a load line, so huge currents can't develop past what your resistor says it should. Also, in all likelihood your bias point is probably sitting near it's maximum dissipation point for that particular load line anyway, so you can't actually increase dissipation any further in the case that the tube does run away (I haven't actually heard of any pre-amp tubes undergoing run-away, though).

Power tubes on the other hand really only have the DC resistance of the transformer as their load (assuming their not cathode biased), which is minuscule, so they can easily get into situations where they can melt themselves to death. Cathode biased amps sort of alleviate the problem. As more current flows through the valve, the negative bias will increase because of the extra current across the cathode resistor, hopefully negating the reverse grid current across the grid leak resistor. Not to say that cathode biased things can't undergo runaway, it's just that they tolerate it better.

[loudthud]

What's the cathode voltage on the 6C4? It needs to be about 20 to 25% of the B+ for that tube to maximize headroom. Are you using NOS or current production 7591's?

[Chuck H]

Is your mid pot scratchy? Even if it's not, you may want to move the .047 coupling cap after the mid control to before the mid control. Right now you have HV on the pot. And pots fail all the time. That would leave you with HV shorted to chassis...Or much worse.

[potatofarmer]

Is your mid pot scratchy? Even if it's not, you may want to move the .047 coupling cap after the mid control to before the mid control. Right now you have HV on the pot. And pots fail all the time. That would leave you with HV shorted to chassis...Or much worse.

That's a very good point. I figured it wouldn't be a problem because there isn't a DC path through the mid control, but I didn't think about pot failure. It isn't scratchy, no.

Hey, Chuck H since you're here - have you ever had a problem with zener diode failure on your 'zener diode cathode bias' trick? Really interested in trying it, but I've seen people note on ax84.com that when zeners fail they usually fail short so it's a good idea to put some resistance or fusing in series with the zener. I'm not sure if it's worth the hassle, or if that's even true about zeners. I think it'd be a good fit for this amp, considering how much power 7591s lose in cathode bias and how far from ideal the power tube overdrive tone is currently.

Man, I even have a scope but it's just sitting there because I don't have an appropriate dummy load.

What's the cathode voltage on the 6C4? It needs to be about 20 to 25% of the B+ for that tube to maximize headroom. Are you using NOS or current production 7591's?

I'll check that tonight; according to the load line it should be pretty close to center biased.

As for the tubes, they're used old stock that came out of an HH Scott amp (which is happy with JJ 7591s). I'll probably replace the ones in this amp with EH when the time comes but I've read so many threads about new production 7591s that I'll probably hold off as long as possible. I know one is a Sylvania, the other one doesn't have any ink left on the envelope apart from the etched '7591A USA' on the top.

I also tossed a Mullard long-plate ECC83 in V1 because it was lying around. I'm not a tube roller so maybe I should sell it instead and put the money towards backup 7591s.

Thermal runaway shouldn't be a problem with pre-amp tubes as they have both a plate resistor and cathode resistor to combat the effect - they're constrained to a load line, so huge currents can't develop past what your resistor says it should. Also, in all likelihood your bias point is probably sitting near it's maximum dissipation point for that particular load line anyway, so you can't actually increase dissipation any further in the case that the tube does run away (I haven't actually heard of any pre-amp tubes undergoing run-away, though).

Thanks; that makes a lot of sense.

[Chuck H]

I don't have a fuse on any of my amps. No failures yet. I had a thread here once about component failure mode. Just a short discussion about what things short and what things open in components and circuits. I can't recall if diodes were discussed. But if they were I must have been asleep since if they typically, or even sometimes fail short I think a fuse would be an important addition to the cathode zener circuit.

If you decide to try it you should dust off that scope. I don't use them often for design. But when dealing in the fine details of bias as it relates to power amp clipping I surely use mine. It's good to see the symmetry, spikes, dips, recoveries, crossover notch and just general behavior of the power tubes if your going to be beating the piss out of them while trying to make them sound good. A dummy load is pretty easy. Here's the one I'm using. I used an aluminum chassis and those finned chassis mount resistors. The 50 waters were the best value at just over three dollars. It even has a low level output for monitoring through a speaker if you want to.

[potatofarmer]

What's the cathode voltage on the 6C4? It needs to be about 20 to 25% of the B+ for that tube to maximize headroom. Are you using NOS or current production 7591's?

Yup, this was it right here. I'd grounded the master volume instead of returning it to the cathode resistor junction. Stupid terminal strips. Now the master volume is less useful and there's a noticeable hum but the distortion tone is markedly improved. No need for a bigger grid stopper. The hum I expected because I got lazy on a couple grounds and was pleasantly surprised that it hadn't caught up to me.

Anyway, the 6C4 is now at:
plate: 316V
grid: 51V
cathode: 71.4V

So, colder than I'd planned but close enough. One more thing to tweak later.

[potatofarmer]

Turns out I never wired it that way, just drew it that way for some strange reason. The coupling cap is before the mid control and it's a 100n... probably just what I had on hand.

[loudthud]

The tone controls look unusual. How well do they work?

[potatofarmer]

The tone controls look unusual. How well do they work?

The bass and treble still need to be tweaked (as well as the bass throughout the preamp) but they work really well, at least at the frequencies at which they're effective. I might convert that treble control to a Princeton tone control as the biggest complaint I've heard is that it's too interactive with the volume control it's attached to but V2b has a fixed voltage divider. Not that I need any more treble, but who knows.

The mid control is incredibly effective - the whole sweep of the pot is useful and it can go from tweed to more scooped than blackface. I actually did the math to design a bridged-t filter like the one in the Gibson GA30RVT, but in Merlin's preamp book he showed that the taper on the control wasn't very useful so I used this version instead. It does make the amp a little tricker to dial in but that might just be because I'm not used to it. I've labelled the mid control "character" because it changes the sound, feel, and gain of the preamp so much, and anyone who plays this thinking it's a traditional FMV stack is going to be in for a surprise!

Oh, and the bass control needs to be able to cut way more. That 1n cap might have to go as low as 100p... or I might just wind up copying Merlin's 10n||470p.