Plate voltage and tube health

[Larpy]

I was rolling 5670s in my Kaiju today and noted that when I did, the newly inserted tubes measured low--around 150 volts on the plate rather than the specified 175 volts I had adjusted the previous tubes to.  Am I right in thinking that tubes that measure lower in this situation are actually healthier, in better shape in terms of performance?

Or is it the opposite?  Are such tubes closer to being worn out?

In the context of the Kaiju, should I be concerned about a tube that measures 150v at the plate when the previous occupant measured 175?

[Paul Joppa]

You should re-adjust the plate voltage using the trimmer as described in the manual.

The lower voltage indicates the new tube has more emission, either because the old one is a little worn out, or just variation between individual tubes. But the voltage is set to operate the tube in its most linear region and avoid clipping on signal peaks. It's a good practice to check (and adjust if necessary) that voltage whenever a tube is changed, or once a year if the amp is in daily use, to maintain performance.

A deviation of +/-5 volts is inconsequential.

[Paul Birkeland]

It's also worth mentioning that the more you anchor down different dimensions of a tube's operation, the more variation you'll get on the parameter(s) that can move.  In the Kaiju driver, the plate current is fixed by the CCS and the cathode bias voltage is fixed by the 431/trim pot combo, so the plate voltage is going to wander a fair amount.  With simple resistor bias and resistor plate loading, the plate voltage stays in a much narrower range, but of course that doesn't sound quite as good. 

[Larpy]

Thank you to both Pauls.

Yes, I immediately readjusted the bias trim pot to get the newly installed tubes to 175v each.  But I was curious about whether a tube whose plate measures 25 volts lower at the same spot on the bias trim pot as the previous tube installed there is a healthier tube or a more tired tube.  Sounds like the answer is "maybe."  It could mean the newer tube has higher emission, or it could simply be a reflection of normal tube variation.

Is it fair for me to conclude that it's preferable to have to raise the bias of a newly installed tube to get it to its specified plate voltage than to have to decrease it?  Or is it not that simple?

I'm asking because I have multiple pairs of GE 5670s that sound more or less the same to me, and I'm thinking I should choose the "healthiest" pair.  But I don't have access to a tube tester.

[Paul Joppa]

Variation of the emissivity is mostly due to the geometry, and does not much affect the longevity. Transconductance (gm) is most closely connected to the cathode to grid spacing, which is very small; gain (mu) is more connected to the grid to plate spacing and the grid wire diameter. Exhaustion of the cathode coating is pretty complicated, but the purity of the underling cathode metal plays a big role, along with the vacuum. (The 5670 has a high-purity nickel cathode structure.)