Making a solid state rectifier

[Bob M.]

I have a broken rectifier tube (the glass has been broken) and I would like to use the base to put some solid state diodes inside of it. Assuming an octal pinout, plates on pins 4 and 6, is there a preference for the diodes to connect to pin 2 or pin 8? I am reading about 5.5 VAC between pins 2 and 8 from the transformer of the device in question. Pin 8 to ground is about 4.5 VAC and pin 2 to ground is just under 1 volt VAC. In an older Fender amp, pin 8 is usually used as the rectifier output, but why?

[Old Tele man]

...the two different voltage reading could simply be the filament winding tap not being perfectly "centered" electrically (hence the 2 x 100-ohm resistor "phantom"-ground technique).

...as for pin 8, that's so that different rectifier types can be used without having to rewire the socket...a standardized output pin for both filamentary and cathode rectifiers.

[Robert M. Martinelli]

Hi Bob, Hi OTM,
I would advise Bob, in order to emulate the performance of the old rectifier tube, to add a series resistor right after the two diodes' cathodes. Tube rectifiers have a greater series ( or forward ) resistance than SS ones, and this causes an appreciable voltage drop, so, by using the diodes alone, the +B ( + HT ) is going to be at least some 20-30 V higher due to the SS diodes' very little forward resistance.

Bob, can you tell us what type the broken rectifier tube was?

( e.g. if it was a 5AR4, the voltage drop @ 300 mA is some 20 V, so a 68 Ohm 10W series resistor will do just fine ).

Hope this helps

Best regards

Bob

[Old Tele man]

...excellent point.

...also, maybe put a suitably "sized" (voltage-wise) switch across (parallel) that resistor and you'll be able to switch between "tight" SS-rectifier and "saggy" tube-rectifier capabilities, however, you'll need to rebias with each setting if the output tubes are not cathode biased.

...and, remember the resistor will get VERY HOT, so a heatsink or external mounting (to the socket) should be considered.

...note, however, that even with this added voltage 'dropping' resistor, the "sag" characteristics will not be identical to a tube-rectifier because the resistor has a "linear" voltage drop while the tube-rectifier has an "exponential" voltage drop that starts increasing slowly then gets faster and faster as current increases, ie: Vp = (Ip/K)^(2/3), where K is Perveance, a tube constant determined by the tubes physical geometries.

[Robert M. Martinelli]

Originally Posted by Old Tele man

...note, however, that even with this added voltage 'dropping' resistor, the "sag" characteristics will not be identical to a tube-rectifier because the resistor has a "linear" voltage drop while the tube-rectifier has an "exponential" voltage drop that starts increasing slowly then gets faster and faster as current increases, ie: Vp = (Ip/K)^(2/3), where K is Perveance, a tube constant determined by the tubes physical geometries.

Correct!
That's why I used the word "emulate"....
( Perveance.....Hmmm....You're taking me back through the years, when I was younger, things were easier and Van der Bijl, Childs, Langmuir together with a few others, were my only masters..... ).

OTOH, I think that, by choosing an appropriate resistor, the "emulation" should be pretty close, because the curve's deviation from linearity in the 150-300 mA range is not all that pronounced, and the difference in the voltage dropped amounts to a few volts, so I don't think our less-than-perfect ears could really detect the difference.

Cheers

Bob

[Gibsonman63]

I have a broken rectifier tube (the glass has been broken) and I would like to use the base to put some solid state diodes inside of it.

That is a lot of voltage to have dangling on the outside of the chassis. It seems like it would be much safer to solder the components internally or just stick one of these in there to see if you like the sound.

solid state rectifier

Or if you really want to roll your own, try some of these parts.

Tube Amplifiers and building supplies for the do it yourself type. We specialize in vintage circuits.

[Steve Conner]

Originally Posted by Old Tele man

...and, remember the resistor will get VERY HOT, so a heatsink or external mounting (to the socket) should be considered.

I presume this is the reason for the copper cap on a Weber Copper Cap, it's a heatsink for the sag resistor. But there are safety issues with exposed metalwork so close to B+, that can't be grounded because the rectifier socket lacks a ground pin. Same applies to a sag switch built into the assembly, it had probably better be a plastic one.

I agree that nobody's going to hear the difference between the VI curve of a real tube rectifier and the "best fit straight line" of a sag resistor.