Replacing a selenium rectifier in grid bias supply

[HighFidelityGuy]

Hi,

A few months ago I got a great deal on a Precision Model 650 valve tester on eBay. Despite being about 46 years old it's in excellent condition and seems to work fine. It even came with an instruction manual and other documentation.

However, after having a look at the supplied schematic I noticed that it uses a selenium rectifier to produce the negative grid bias voltage for the on-board 12AU7 valve. As these don't age well and can release toxic fumes on failure, I'd quite like to replace this for a modern silicon diode.

All the info I've found online about swapping a selenium rectifier for a silicon diode use the example of the SR in a standard positive high voltage (HT) supply. My valve tester has the SR connected the opposite way round so that a negative voltage is obtained.

Usually a diode such as a 1N4007 is recommended to replace the SR, along with a dropper resistor but I'm not sure if that still applies when the diode will be used backwards.

Does anyone know if a 1N4007 is still suitable and if I should still use a dropper resistor as is suggested HERE?

Here's the valve tester's shematic:



Cheers.

[Firebottle]

Does anyone know if a 1N4007 is still suitable and if I should still use a dropper resistor as is suggested HERE?

Yes and yes.

[HighFidelityGuy]

Thanks Alan!

[HighFidelityGuy]

I've been working on this today and after replacing the selenium rectifier with 1N4007 and then experimenting with different dropper resistor values I've found that I need to use a 3K resistor to get the -107V grid voltage I had when the SR was in place. That's 10 times more resistance than is usually recommended. So I think that the forward voltage drop on the SR had increased quite a bit over time and left the grid voltage lower (less negative) than it should be. The problem I have now is that I don't know what the grid voltage should be as this isn't mentioned on the schematic.

Can anyone tell from looking at the schematic I posted what the grid voltage on pin 2 of the 12AU7 should be in this application?

Thanks.

[Firebottle]

Dave I would say that's about right.
Don't be concerned that your series resistor appears large, selenium rectifiers are quite high impedance devices.

If you can get the required adjustment on the VTVM and cal controls then all is good.
Damned useful having the circuit

Alan

[HighFidelityGuy]

Thanks Alan, that's a good point. I still need to order the resistor as I've currently just strung together some smaller values for testing. So I think I'll get a 2K7 and a 3K to try out. I'll replace the 10uF smoothing cap at the same time. Actually, seen as I'm adding an extra resistor, I guess I could add an extra cap to make a pi filter to reduce the ripple further. It might be worthwhile seen as the rectifier is only half wave.

It really has been handy having the schematic. The tester also came with a full copy of the manual and an original tube substitution book. So I got a great deal.

Do you think I'll need to do much in the way of calibration once I've got it up and running again?