Query: Replacing a rectifier tube with diodes and another preamp tube in the socket

[Mark Hammer]

I have a head made for a 5Y3, 2 x 12AX7, 2 x 6V6 tube complement. I have no idea whether the power transformer is up to the task of powering the filament of a 6th tube. I also have a pair of nice Fender tweeds so it's not like I need a 3rd amp to have sag as one of its traits.

If I replaced the socket for the 5Y3 with a 9-pin for a dual triode, and replaced the 5Y3 with a SS rectifier, the extra two triode sections would open up some possibilities.

1) Is this advisable?

2) What diodes should I use for replacing the 5Y3?

3) Is there any risk that the resulting plate voltage would be higher than acceptable because of this change?

I'm open to suggestions and ideas, including "If you have to ask these questions, you shouldn't be working on a tube amp".

[jakehop]

Hi Mark,

1) If that is what you want, and the amplifier is not otherwise useable for you in it's current form - and selling it doesn't make sense.

2) 1N4007 is the common one. I usually use some kind of ultra-fast diode with a snubbing capacitor, as it doesn't cost a whole lot more, and sounds good to me.

3) Yes, this is a possibility and it will probably do something to the sound as well. You can drop this voltage with a MOSFET-regulator or a large resistor.

Was it nessesary to remove the 5Y3? An extra preamp tube isn't drawing much current from the HT-supply. The problem is usually the heater-wiring. My math tells me, that a 300mA load on the 6,3V winding will correspond to a 21 ohms resistor. Measure the heater-voltage with all the tubes in, and a 20ohm resistor across the heater wires, and if it's 6,3V AC +/- 10% and the transformer isn't buzzing or getting hot, chances are you can get away with it.

Kind regards, Jake

[R.G.]

1) Is this advisable?

I have no clue.

2) What diodes should I use for replacing the 5Y3?

You can probably get away with 1N4007s in this circuit. The full wave CT power circuit puts 2x B+ on the diodes for a reverse voltage. If it's well under 500V, then a 1kv diode works. 1A is plenty of current rating. It's the reverse voltage that is the problem. You could use four 1N4007s with 2.2M resistors paralleling each one in two sets of two in series. The resistors force the reverse voltage to divide equally and don't let unequal junction leakages or capacitances pop a junction for chain failure. With 1200V diodes or above, I wouldn't bother with series diodes.

3) Is there any risk that the resulting plate voltage would be higher than acceptable because of this change?

Yes. It'll be about 50V higher maybe. You can fix that with a power resistor in series from the junction of the diode cathodes to the first filter cap. This introduces a good enough emulation of the 5Y3 forward resistance to get the voltage back down. It's what is done in Weber's copper cap rectifiers.

I'm open to suggestions and ideas, including "If you have to ask these questions, you shouldn't be working on a tube amp".

Nah, you'll do fine. Just remember to stand still until the lion is nearly to you before you raise the spear, and keep your heel on the butt of the spear on the ground. Don't want that slipping.

Another 12AX7 only adds a couple of ma of plate current, insignificant compared to the output stage. The extra drain is all in the added heater current. By not using the 5V heater to the rectifier, you have more of the power transformer's power available, so the total power out is fine. The only real worry is whether the 6.3V heater winding itself is right on the edge with the existing current load. I'd bet it would be fine.

'Course, I'm betting with your money...