Hello jazbo8, DIYBras and All,
No I have not tested any of these valves. I have tested and matched a pile of JFET’s in the past.
When we are accustomed to 1% resistors and 5% capacitors 20% valves are all over the place. jazbo8 what is your experience, how much to the tubes you have vary. Is it like cow plop bingo?
DIYBras I went to bed thinking about test procedures. Perhaps as you say; a constant B+ voltage, no anode resistor and a fixed resistor bias to see where they bias up in terms of grid voltage. this may be close enough.
Maybe with the same setup, switch in and out a parallel cathode resistor. This would select for original bias and provide a second bias point for comparison. If a pair of valve bias up at the same points they are close to being matched?
Throw in a plate resistor to allow monitoring changing anode voltage then calculate gain.
DT
No I have not tested any of these valves. I have tested and matched a pile of JFET’s in the past.
When we are accustomed to 1% resistors and 5% capacitors 20% valves are all over the place. jazbo8 what is your experience, how much to the tubes you have vary. Is it like cow plop bingo?
DIYBras I went to bed thinking about test procedures. Perhaps as you say; a constant B+ voltage, no anode resistor and a fixed resistor bias to see where they bias up in terms of grid voltage. this may be close enough.
Maybe with the same setup, switch in and out a parallel cathode resistor. This would select for original bias and provide a second bias point for comparison. If a pair of valve bias up at the same points they are close to being matched?
Throw in a plate resistor to allow monitoring changing anode voltage then calculate gain.
DT
My results were not as wide as the extreme cases shown by smoking-amp above, but probably varied +-25% with a small lot of 10, which is quite wide IMO, may be I am just un-lucky...
"cow plop bingo?"
😀
Maybe if the tubes are all from the same manufacturing lot they will be better behaved.
😀
Maybe if the tubes are all from the same manufacturing lot they will be better behaved.
The two bias points is a good idea! The tubes with most close results in the two cathode bias points certainly will have the most similar gain. And this will increase only a little the test difficulty. *Maybe* is not necessary to further test in anode, I think.
Last edited:
jazbo8,
I am assuming that you used a tube tester to test your valves. If I remember correctly a tube tester tests transconductance or current change vs grid voltage change.
I am without a tube tester. If I were to do this at home, perhaps a constant voltage at the grid and pick an externally supplied bias voltage, record the current, adjust the grid voltage maybe 0.1 volts and record the current change then calculate the transconductance.
I will try something like this and report the results.
Yes you may be very unlucky or maybe you have a normal sample. Let’s find out!
DT
I am assuming that you used a tube tester to test your valves. If I remember correctly a tube tester tests transconductance or current change vs grid voltage change.
I am without a tube tester. If I were to do this at home, perhaps a constant voltage at the grid and pick an externally supplied bias voltage, record the current, adjust the grid voltage maybe 0.1 volts and record the current change then calculate the transconductance.
I will try something like this and report the results.
Yes you may be very unlucky or maybe you have a normal sample. Let’s find out!
DT
In some old designs I made, by the time I don't had μTracer, I tested the valves only in one point, one that would apply to the project. When I achieve a reasonable number of valves with good proximity of these measurements, I finished the project, and I measured the tube gain during normal application (with audio test signal, multimeter and FFT), with subsequent change for some tubes that deliver gain out of specification. In PP amps I always measured the residual 2H for matching tubes.
No, I used a tube tracer, with the exact same settings for all the tubes, then I compared the curves from the pictures I took - a bit tedious, but you see everything, the grid curve spacing and the curvature differences, etc.
- Status
- Not open for further replies.