tube matching

[TUBESMAN]

Lets say you have a bunch of single triodes, 6J5 for example. You want to match them for use in a CCS driven Differential stage. What would you measure to establish that tube #3 is the matched pair of tube #81?

I assume that the best place to start is to place them in the circuit you plan to use at the intended operating point and measure the voltages across the tubes.

How equal is equal?

[SY]

Actually, if it's a CCS-loaded long tail diff stage, the matching isn't at all critical.

If you do want to match, I'd set up a jig to run the tube at the designed operating point (plate current and grid-to-cathode voltage), then match for plate voltage and gain.

[Wavebourn]

Trace curves on transparent films then match them visually.

There was a project related to some geological research. One Russian computer center got an order to calculate percentage of some spots on maps of the entire Earth. Instead of loading the data into the computer a programmer took scissors and cut maps, then weighted results on precision scales. He got a premium for fast and accurate results.

[TUBESMAN]

Thanks Sy, that is about what I figured would be the case. Sometimes I ask questions so I can be sure I am as crazy as I think I am, by seeing if someone agrees with me. As far as curve tracing, I remember a few years ago an Italian (I think) company sold a microprocessor controlled curve tracer. I think they stopped making it, I think, is there a similar unit from someone else?

[TUBESMAN]

[QUOTE]Originally posted by SY
[B]Actually, if it's a CCS-loaded long tail diff stage, the matching isn't at all critical.

Is it also possible that in a CCS LTP that one tube, in an extreme example, could draw 15mA and the other 5mA in a circuit set for a total of 20mA? That is why I figured I'd like to match the tubes.

[Wavebourn]

You cam feed your target stage with a sawtooth and save oscillograms to compare.

[ErikdeBest]

Hi tubesman

For a differential stage with CCS in the tail and matched plate resistors (which are essential conditions for a differential stage anyway) Morgan Jones recommends to match triodes by comparing plate voltages: same voltage = matched.

Erik

[SY]

Is it also possible that in a CCS LTP that one tube, in an extreme example, could draw 15mA and the other 5mA in a circuit set for a total of 20mA? That is why I figured I'd like to match the tubes.

Yes, but unless you're direct coupling to the next stage, it doesn't matter much; remember, the swing is constrained to be equal and opposite. As a practical matter, it's tough to find two tubes of nominally the same type that are THAT far apart.

[TUBESMAN]

Originally posted by SY


Yes, but unless you're direct coupling to the next stage, it doesn't matter much; remember, the swing is constrained to be equal and opposite. As a practical matter, it's tough to find two tubes of nominally the same type that are THAT far apart.

I did say it was an extreem case! I realize that unless the makers of the tubes really don't give a rats a**, they will be closer than that! I think I have an adequate group of answers to my question. Now all I have to do is buy the 100 triodes that I want to stockpile!

[Gordy]

Originally posted by TUBESMAN
...I remember a few years ago an Italian (I think) company sold a microprocessor controlled curve tracer. I think they stopped making it, I think, is there a similar unit from someone else?

I think the old company was Sophia. For a unit nowdays look to Mr. Hagerman: http://www.haglabs.com/vacutrace.html