Thanks! Given the curved lines, this is no surprise. But I've not find (in my old databooks, 1963, 1981) anything about this.
So without mentioned pinpointed references, where does these 'general' specs apply?
I have yet to see a datasheet that does not specify at least the points where they measure mu and gm.
For example, https://frank.pocnet.net/sheets/049/1/12AX7A.pdf, page 1 shows two working points, page 4 shows mu and gm curves.
Finally, you can always measure mu, gm or ra from the plate curves, but it can be a bit painful with low res small pictures.
For example, https://frank.pocnet.net/sheets/049/1/12AX7A.pdf, page 1 shows two working points, page 4 shows mu and gm curves.
Finally, you can always measure mu, gm or ra from the plate curves, but it can be a bit painful with low res small pictures.
And even if nowadays you buy a valve with a stated tested parameter, then be aware that such a measured value relates just to the valve tester instrument used, which doesn't imply that any other valve tester would then measure the same value.
In the long run, if you were concerned about a particular performance parameter for your particular application then its likely you would end up tube swapping from your available batch of a particular tube, and making particular measurements in your particular application. I've done that a few times, such as for example Williamson amps, but even then for example the screening for the best available distortion from a 6SN7 may have no tangible benefit as the influence of the screened parameter could be obscured by other parts of the amp (such as output stage valve distortion), so the screening ended up being to weed out valves with too divergent an overall set of measured values.
In the long run, if you were concerned about a particular performance parameter for your particular application then its likely you would end up tube swapping from your available batch of a particular tube, and making particular measurements in your particular application. I've done that a few times, such as for example Williamson amps, but even then for example the screening for the best available distortion from a 6SN7 may have no tangible benefit as the influence of the screened parameter could be obscured by other parts of the amp (such as output stage valve distortion), so the screening ended up being to weed out valves with too divergent an overall set of measured values.
You guys need a curve tracer.
Sometimes shifting the Vg2 a bit can bring two tubes into convergence. Some tubes however may have the grid/cathode mis-aligned with one side closer to the cathode or tilted, frame grid tubes for example. Bad triode curves then. Then there is mis-aligned grid1 and grid2 wires, which will cause high g2 current and big kinks.
Then there is low emission, which a tube tester can find. Back in the $1 tube days I always bought a 100 pcs for matching up (usually small tubes ). For output tubes a trim pot or two for adjusting Vg1 and Vg2 is a good idea. ( no UL here ) Be prepared to throw out a few tubes. Some tubes are hard to imagine what they did to them. I did find a sealed box of 200 6197 Mil Spec tubes which were so close as to be matched by random selection.
Good tube:
Kinked tube:
Nirvana:
Sometimes shifting the Vg2 a bit can bring two tubes into convergence. Some tubes however may have the grid/cathode mis-aligned with one side closer to the cathode or tilted, frame grid tubes for example. Bad triode curves then. Then there is mis-aligned grid1 and grid2 wires, which will cause high g2 current and big kinks.
Then there is low emission, which a tube tester can find. Back in the $1 tube days I always bought a 100 pcs for matching up (usually small tubes ). For output tubes a trim pot or two for adjusting Vg1 and Vg2 is a good idea. ( no UL here ) Be prepared to throw out a few tubes. Some tubes are hard to imagine what they did to them. I did find a sealed box of 200 6197 Mil Spec tubes which were so close as to be matched by random selection.
Good tube:
Kinked tube:
Nirvana:
Last edited:
Datasheets sometimes say something like "average tube". This could refer to some sort of a mean (arithmetic, geometric, whatever) or median, etc. I don't think it makes a big difference in practical use of the datasheet numbers.
A tracer is super helpful to understand how a particular tube works and performs (and a fun tool/toy for people like me).
However, I think the original question of this thread is aimed at making sense and interpreting datasheet numbers.
Thanks, I found the very spots and how to determine it for other tubes, very helpfull.
Also for the other contributions.
That's obvious - I prefer E12 (5%) resistors where possible.