voltage regulator for tube tester

[Alex R]

I'm thinking of putting together a simple tube testing rig with meter test points that will run up to 4 output valves at a time, providing amp-level voltages, currents and signals.

What do you think would be the best way of providing regulated HT and bias supplies that would produce the same current readings whether one or four tubes were plugged in?

[Enzo]

Why do they need to be regulated? Isn't just variable good enough? Set the current each time.

[Alex R]

... and classify them on how much bias voltage they need to achieve the standard current?... ok, thanks. I was a bit interested in how you might go about regulating HT, but am quite happy not to know!

Is the simplest way to provide a plate load a big fat resistor per tube?

[Steve Conner]

You can pick up tube regulated power supplies quite cheaply.

I personally have one that gives dual 0-300V outputs (one 200mA and one 100mA iirc) plus a fixed -200V and a few 6.3V AC outputs. The variable outputs have voltage and current meters.

I bought it second-hand years ago, and am looking to get rid of it, since I've just scored a more modern switchmode one that cranks out 0-600V at a more-than-adequate 1.7 amps.

The simplest plate load is none at all, just connect it straight to the supply. This gives the same DC conditions as a tube amp, where the OT also has negligible resistance at DC, so it's arguably more relevant to tube matching than a circuit with a load resistor.

[Alex R]

Cool but I was thinking of running a signal through to look at headroom and breakup, and maybe get an objective idea of frequency response. Wouldn't I need a plate load then?

Well, pity Enzo pointed out I don't need one, otherwise I might have made you an offer on your PSU . But I will ask after all, even though I don't really need one - how does a tube regulated power supply work? By causing a tube to pass current when a certain potential is reached?

[Enzo]

Think of a single ended amplifier like a Champ. The power tube controls the current through the load. In this case the load is the OT primary. But that load could just as well be a resistor or something else.

VOltage regulation is a matter of having control over a voltage and a feedback/sensing arrangement that keeps the adjustment stable. So in my Champ, I could eliminate the OT winding and wire my output tube plate right to B+. Now the cathode resistor is my load. The bias would have to be fixed/adjustable now. The lower the bias voltage the higher the current through the tube - and so the higher the voltage across the load. Add a tube to sense the output and use that to control the power tube, and you have a regulator. There needs to be some sort of reference in most cases, so we use a glow tube just like a zener. Glow tube - gas regulator tube.

See chapter 33 of the RDH http://headfonz.rutgers.edu/RDH4/CHAPTR33.PDF

[Steve Conner]

Yes, you will need a load to test it under signal conditions. A big resistor, or a resistor shunted by a choke to simulate an OT, or whatever.

Tubes don't have a frequency response as such. They have transconductance and plate resistance, and those don't change from DC to MHz. It's the non-idealness of the OT that determines the frequency response of a tube power amp.

Hence, the frequency response you see in your test rig will be determined by whatever test load you choose, rather than the tube you're testing.

Different power tubes do give different frequency responses in tube amps, but only because of their different distortion harmonics, and different gm and rp values interacting with the OT and NFB loop. For instance, tubes with low rp will wring more bass out of an undersized OT, and damp the speaker too. Tubes with high gm will make the presence control seem to work better. And so on.

Like Jon Blackstone said, it's not the tube, it's the rest of the circuit.