> the actual mho's ... .. with assembly language and ...microcontroller boards
Oh, geez. Make a mountain out of a molehill. (OK, it was for school.)
Do you know how to test Gm?
A "dead" tube has no Gm. You have to bring it up to a happy Operating Condition. In general, the OpCon in the Tube Manual.
So you start with DC power supplies. For guitar-amp tubes we will almost always be working for the 250V plate voltage condition (seems the Old Guys left their lab-supply set at 250V), and -1V to -15V on the grid. So build a steady +250V and -15V supply. The grid eats very little current and a 9V battery (or two) may be convenienter. And of course a Heater supply.
Wire like Fig 1 below. An Amp-meter tells the plate current.
Already this is VERY useful. If you set-up a 12AX7 with 250 Vp and -2 Vg you should find about 1.2mA plate current; otherwise it aint healthy. Recall that all tube numbers are +/-20%, so "1.2mA" could be 1.0 or 1.44mA and it isn't a dud.
The definition of Gm implies causing a small change of grid voltage and measuring the change of plate current. We could add say 1V to the grid bias and observe the change of plate current. However 1V change on a 12AX7's 2V bias is not "small". Also the change will be very different if the 1V is added or subtracted. And you have to subtract two current readings to find the change.
(In fact I think the B&K et al did use a 1V change. Many old Gm plots show Gm falling-off as bias crosses 1V. This was fine for 1930s tubes which used large biases, less-fine for 1940s tubes of high gain. However larger test signals are easier to read with old passive meters; DMMs are game-changers.)
It is just as valid, and maybe easier, to insert a small AC voltage on the grid bias and measure the *AC* current output. If 0.1VAC at grid causes 0.16mA AC current in plate loop, then Gm is 0.16mA/0.1V = 0.000,160A/0.1V = 0.001,6 Mho. Because heavy-zero fractions are ugly and typo-prone, old-US practice was to multiply by a million and get uMho. 1,600 uMho. Many European sheets favor milli-Mhos (and yes, we now honor Siemens).
Figure 2 shows adjustable grid bias, AC injection, and a (small!) plate resistor to read plate current, both DC and AC.
Figure 2 will work. But having the metering up at the 250VDC point is a safety disadvantage.
Since the plate resistor must be small, and the 250V DC supply must be very solid, we can re-arrange it all to read plate-cathode current on the cathode side of the battery. Figure 3. The cathode resistor should be "small", in the sense that its DC voltage drop at test current should be much less than the DC grid bias. 12AX7 at 1.2mA for -2V bias, 2V/1.2mA is 1,667 Ohms, so we want much less than that. We also want round-numbers if possible, for easy math. 100 Ohms may be convenient.
The grid DC feed resistor must be less than the datasheet rated max grid resistor, but not so low it is hard to drive with AC. That's not a tough problem.
There should probably be a series grid resistor so we don't measure grid-current operating points which are not generally useful in amplifiers.
The signal frequency must be OK for the tube and our AC meters. The tubes are generally good from DC to far-far beyond the audio band. Many inexpensive AC meters are dubious at 20Hz and above 1KHz. The most obvious choice is to tap a 50/60Hz winding on a power transformer and divide it down to a suitable AC Volts.
Topic: Tube Testor thread (Read 3727 times)