Gotta a power transformer, can't seem to decided what to do with it - any thoughts ???

[jmaf]

You mentioned that, and now I can't put the idea out of my head, so...


HV for B+, probably doesn't need too much filtering?
bias supply, variable over a wide range?
I have a couple 0..100mA meters for real-time comparison testing of plate or cathode current. Or maybe both plate and cathode, and have a single socket.
I'm still trying to imagine a jig to compare the current differential between two tubes. Two 1R resistors on the cathodes and measure mV difference at the cathode pins? I guess that could be done on any amp with bias test points.

Who's done this before?

edit: of course I should be looking at bias current too, but my assumption is that plate dissipation is what tells how closely the tubes match at any given bias voltage.

Filtering: it depends on what you want to spend on the project. If you've got the components and $ then there's no reason not to build a hard PS. Remember : testing with class A bias will set the tube to 100% power consumption, so you'll get ripple unless it's hard PS. Sag isn't important, ripple probably is depending on how you plan on measuring currents.

Bias: 0 to -50 maybe?

Bias current? You plan on matching using AC signals? If matching in DC at a fixed point then you won't need to measure any bias current. Unless I understood what you meant incorrectly.

Differential current: it'd be neat if you had an analog needle centered galvanometer and the needle moved to the side of the tube with more current. Matched tubes would show the needle exactly on center.

If you build a regular output stage with 1R resistors on cathode, all you need really is a dual channel oscilloscope to match any pair of tubes. Just tap the reading off the cathodes and view the difference right on screen.

[eschertron]

Bias current? You plan on matching using AC signals? If matching in DC at a fixed point then you won't need to measure any bias current. Unless I understood what you meant incorrectly.

I think I wrote what I meant incorrectly. Slow brainwave day on Monday. Upon reflection, I would expect exactly zero current through the control grid for any of this bias voltage only applied to grid testing.

And yes, dual-channel scope on 1R sense resistors allows me to read individual and differential currents with a flick of a switch. I was simply (again, Monday) thinking of a more Golden-Age-Appropriate type of test rig. Kind of 50s sci-fi with gauges and dials.

[jmaf]

I think I wrote what I meant incorrectly. Slow brainwave day on Monday. Upon reflection, I would expect exactly zero current through the control grid for any of this bias voltage only applied to grid testing.

And yes, dual-channel scope on 1R sense resistors allows me to read individual and differential currents with a flick of a switch. I was simply (again, Monday) thinking of a more Golden-Age-Appropriate type of test rig. Kind of 50s sci-fi with gauges and dials.

Since the last post I got my mind set on one of these:



I'd drill the test tube sockets holes to each side of it on the chassis. The needle would point to the hotter tube. A perfectly centered needle means matched tubes.

[pdf64]

...dual-channel scope on 1R sense resistors allows me to read individual and differential currents with a flick of a switch. I was simply (again, Monday) thinking of a more Golden-Age-Appropriate type of test rig. Kind of 50s sci-fi with gauges and dials.

Yes, monitoring the signal at cathode current sensing resistors is invaluable for assessing weaker tubes in a p-p set; for a 4 power tube amp, it massively simplifies the task.

[eschertron]

Yes, monitoring the signal at cathode current sensing resistors is invaluable for assessing weaker tubes in a p-p set; for a 4 power tube amp, it massively simplifies the task.

If building a dedicated test jig, is there an advantage to sensing current at the anode versus at the cathode? Short of 'proving out' how much current flows at anode and screen, would it be desirable (or even possible) for a tech to identify tubes with wonky/damaged screens by the ratio of anode/screen current that might not be detected by simply looking at the current sum at the cathode?

[pdf64]

I suspect that bad screens generally identify themselves pretty well with fault current, arcs etc. Trying to assess screen to plate current ratio and quantify a 'go/no go' point is probably taking things further than manufacturer's / vendors QA bother with.