____________________________________________
I have a Hickok 533a in storage, but I mainly use a
Russian L3-3 tube tester. I'm very confident the readings it delivers are absolute, but the set-up & use time is terribly long compared to Hickok or your Dyna-Quick. The whole point of most Hickok and B&K testers is to be more portable and quicker/easier to use ("
I need an ohmmeter to tell me now if my resistor is Open/Shorted/Good. I don't need a Wheatstone bridge and a galvanometer to measure the precise resistance to 8 significant digits...")
It is different from the bulk of American service tube testers in that there are separate, regulated power supplies for Plate, Screen, and Bias voltages, and there is an internal tuned-oscillator to feed the a.c. tester signal, along with a
selective voltmeter circuit to only see variational plate current resulting from the input signal. The initial setup procedure involves calibrating the supplies, transconductance measurement, and the meter's readings on a variety of scales (including the microammeter scales used in the leakage tests). The tester measures plate, screen & grid current as well as transconductance, and the first step is checking plate/screen current for the applied plate/screen/bias voltages (which can be compared directly to tube data sheet figures). And the tube can be tested with either fixed bias or with cathode resistors provided auto-bias.
The kicker is the plate current measurement. It's spelled out more in small-signal tubes (see page 4 of
this 12AX7 data sheet), but for all tubes the internal plate resistance falls and Gm rises when the tube is operated at higher plate current. By measuring plate current first, the L3-3 can identify tubes that tend to run at higher/lower plate current (than shown on the data sheet) for the given data sheet condition. I make a note of that actual plate current first. Then the fixed bias can be adjusted until the tube lands on the data sheet value of plate current, and a Gm measurement made. Now the Gm figure isn't artificially high/low just because the plate current was high/low.
- The maximum output of the L3-3's tuned oscillator is also only 450mV (for power tube types), and is further divided-down for small-signal types. The small a.c. signal means a closer approach to the true meaning of "Gm = ∆Ip / ∆Eg"
The plate current test also called out tubes as defective, that others might have called "better than NOS" on a typical Hickok. I was using the 250v plate & screen Class A Amplifier condition on page 2 of
this 6L6GC data sheet. Bias was set to -14v, and the sheet shows
ideal plate current of 72mA and Gm of 6000 micromhos (6mA/volt). I'd tested ~30 was American 6L6GCs and 5881s with that setting. Most landed at a plate current slightly less than 72mA, with a range from 60-88mA. Gm was slightly above & below 6mA/v, even before correcting for plate current & even if the tube's plate current was a bit lower-than-average.
But then I popped in some American-made 5881s from 1989, and with an off-brand label (though internally they looked like old Tung Sol 5881s). Plate current was 132mA, and Gm was off scale (above 7500 micromhos or 7.5mA/v for that setup).
Double the expected plate current couldn't possibly be right! I tried using another data sheet condition with 200v on the screen and -12.5v bias which should have given 48mA plate current & 5300 micromhos (5.3mA/volt), but the tube measured 98mA plate current and 6.9mA/volt!
My interpretation is the tube was defective and sold off by the manufacturer to the off-brand company way back in the 80s, who cheerily sold it as "better than new!" Someone testing that tube today on their Hickok or B&K will hopefully have tested a bunch of other 6L6 types and at least notice none of the other tubes pegged the Gm/English scale like that.
Topic: Advice on using a GM tube tester (Read 7276 times)