Hello All,
Looking for the sweet spot.
I borrowed this process from Nelson Pass.
Somewhere I have a plywood breadboard with a 9-pin tube socket and a couple of rows of euro type screw mount connectors.
Also I have a Keysight N5772A 600VDC variable power supply.
After more than a few late nights I found that you can dial in the supply voltage and swap in and out cathode resistors to adjust the bias current and grid voltage.
Tubes most definitely do have a sweet spot (or grid voltage, plate current range) where the noise and distortion is minimized.
I found that the JAN GE 5760’s were fairly consistent. The 6NP3 is the Russian version or nearly equivalent to the 5670.
I spent more time with the 12AU7 and found that even if the 2nd and 3rd H’s were fairly consistent the 5th, 6th H’s and on up could be all over the place.
The harmonics beginning with the 2nd have the typical stair steps of decreasing HD’s.
Yes most often the 2nd H is the highest step, however the 3rd, 4th, and on up are stair stepping down not far behind.
Tube distortion is rarely if ever just 2nd Harmonics.
Also keep in mind that with the APx555 or any other Analyzer large sample size and averaging only filters out the appearance of noise on the FFT plots. In real time the noise swamps or masks most of the lower magnitude Harmonic Distortions that we see on the FFT plots.
Thanks DT
https://www.keysight.com/us/en/assets/7018-01235/data-sheets/5989-1330.pdf