Vacuum Rectifier Faking (for shane...)

[R.G.]

I got off into the weeds wondering whether people could really hear the difference in SS rectifiers and vacuum rectifiers. It had always seemed to me that any differences in rectifiers would get mashed into submission by filter caps, but I decided to see if there is a repeatable basis for the claim. Turns out, there is.

Bare SS rectifiers produce a different ripple voltage waveform, and the implied different set of harmonics than a vacuum rectifier. So to the extent that the ripple worms its way into the audio path, you could pick out a difference, maybe.

I then wondered how closely a set of SS components could approach a vacuum rectifier, so I tried to make a diode-zener-resistor circuit that would come as close to a 5AR4 (for instance) as possible. After some tinkering and simulation runs, I got it pretty close, not only in the V-I curve matching, but the spectrum of the ripple voltage. Here's a picture:

Fake 5AR4 model.png

The resistor-capacitor nets on the transformer snub diode and leakage inductance ringing for standard slam-off SS diodes; using UF.... type diodes or FREDs might make this unnecessary.

Each type of vacuum rectifier will have a different V-I curve, and so the circuit for a 5U4 type will be different from the circuit for a 5AR4, and both will be different from a baby 6CA4/EZ81. The zener values and resistor values will have to change to fake other curves.

The setup has approximately all of the advantages and nearly all the disadvantages of a real 5AR4/GZ34 rectifier including the heat wasted in the resistances. But it sidesteps the need for heater current and the inevitable wear-out of a real vacuum cathode and plate.

[Stevem]

Interesting, thanks for sharing RG!

[shane]

That's awesome RG, thank you so much for posting that I will read it most carefully.