315 volts on a 12ax7?

[camnicklaus]

I modified my 5e7 towards a 5f6a in a few places including dropping less B+ to the preamp stages. I wondered about having that much voltage on the plate of a 12ax7 when I looked up a data sheet it reads 300V max on the plate, but then I double checked the old fender scheme and sure enough V2b's plate is at B+ with no plate resistor, they have 325 written there. Is this just another case of amp builders running things above spec and waiting to see how well it works?

[Twist]

I think what you should be concerned about is the voltage across the cathode to the plate. As you can see part of that 325 volts is at the cathode.

[camnicklaus]

hmmmmm, that makes too much sense. I've got 230 on the cathode which is apparently 30 volts over max...
Is that enough to worry about?
could I change v2a's plate resistor and V2b's cathode resistor to reign it in while maintaining the CF balance?

[Twist]

Are they both 100K? That would be a common value to use. I don't think your going to damage your tube. It would be nice if you could scope a sine wave at the second cathode to see if it's clipping or nice and clean.
I'm always concerned about clean pre amp signals up to the output section. It gets ugly sounding if it just clips the top or bottom of the signal. Although some people like that sound.

[Mandopicker]

Twist,

Is there an optimal voltage that would be more likely to reduce this ugly clipping? (I realize there may be a number of factors that play a role.)

And..after achieving the optimal voltage, is it necessary to re-bias the tube? (12ax7) ie. adjust the cathode bypass cap and resistor combination.

Thanks in advance.

[loudthud]

Voltages on tube data sheets are almost always specified with the cathode grounded. In amps with the 5F6 type tone stack driver you almost always see the cathode resistor of the gain stage is 820 or 1K. This is done to make the stage clip more symmetrically.

[martin manning]

Are they both 100K? That would be a common value to use. I don't think your going to damage your tube. It would be nice if you could scope a sine wave at the second cathode to see if it's clipping or nice and clean. I'm always concerned about clean pre amp signals up to the output section. It gets ugly sounding if it just clips the top or bottom of the signal. Although some people like that sound.

Twist, Is there an optimal voltage that would be more likely to reduce this ugly clipping? (I realize there may be a number of factors that play a role.). And..after achieving the optimal voltage, is it necessary to re-bias the tube? (12ax7) ie. adjust the cathode bypass cap and resistor combination.

For gain stages, this is about where the operating point is placed on the plate characteristic of the tube.

Generally higher B+ (supply) voltage will increase the available headroom (grid voltage swing before clipping) as long as the idle point remains well centered. Centering is achieved by the combination of plate and cathode resistors used. The cathode bypass cap doesn't affect this, it is only important for frequency response.

For a 12AX7, a cathode resistor of about 1.5% of the plate resistor value will pretty much center the operating point for any typical supply voltage. So, combinations like 100K plate, 1.5K cathode; 150K plate, 2.2K cathode, etc., will put you in the right ballpark.

A higher cathode resistor value will lower the idle current, move the operating point (plate-to-cathode voltage) to the right, towards B+ and cut-off, producing more compression and distortion on the negative swing of the grid signal. Note the 100K plate and 10K cathode resistors on V1b in a Marshall 2204, for example. Reverse this for lower cathode resistor values, except now it is grid current limiting that clips the waveform on the positive swing of the signal.

For cathode followers (with no plate load resistor), it is the filament to cathode voltage that becomes a problem. For DC coupled cathode followers like V2b in the 5F6-A circuit the cathode is going to be within a volt or two of the preceding stage's plate, so you have something close to that voltage from filament to cathode on the cathode follower (it's filament is close to ground potential). This is pushing the limits of the insulating coatings, and I believe some current 12AX7's are known to fail quickly in this location.

MPM

[martin manning]

...In amps with the 5F6 type tone stack driver you almost always see the cathode resistor of the gain stage is 820 or 1K. This is done to make the stage clip more symmetrically.

Given a 100K Ra on the gain stage before the CF, a 0.82K Rk (like in the 5F6-A) will lower the plate voltage on that stage and help out the CF by letting it operate at a lower Vk. But that actually un-centers the operating point for the gain stage. A higher Ra and Rk (180K/1.8K) would bring down Va and keep the operating point centered, but maybe that lower headroom is part of the sauce!

MPM

[Enzo]

That plate rating is not an absolute voltage, it refers to the voltage across the tube, which is basically what Loudthud was saying. In other words if you put 500v on the plate and 400v on the cathode, the tube would have only 100v across it. If you put 500v on the plate and ground the cathode, well then you are way over the limit. So it is not just the 500v, it is 500v with respect to the cathode. The cathode can operate at high voltage as long as it is reasonable with respect to the other elements in the tube. There is a heater to cathode max voltage. THAT is what you need to concern about. Otherwise the high voltage on the tubes in this CF are fine.

[camnicklaus]

ah ha. that makes sense. It's like the voltage ratings for most components...