Stromberg-Carlson AU-36B -- 5k bleeder resistor necessary?

[potatofarmer]

Hey all,

I've been poking at an old project again, a Stromberg-Carlson AU-36B from ~1952. Here's a schematic:

http://4.bp.blogspot.com/-fpBxYORLvg...orrections.gif

My amp is slightly different from the only image I can find online so I noted the differences in red. There is a schematic stuck to the inside of the amp, but getting a good picture of it is really tricky.

I'm in the process of recapping it - the PT is putting out 410-0-410 with modern line voltages so unloaded B+ is roughly 540V after the paralleled 5U4Gs. The schematic shows a plate voltage of 420V, so I'm expecting my loaded B+ to be around 470V. It's back-biased, so there's plenty of series resistance to keep the rectifier tubes happy. I have put protection diodes on each 5U4G; fusing to come soon. This iron's been around for half a century; I'd like to keep it for another!

Now my question is about that 5k/40W resistor to ground right at the screen supply. This schematic specifies 6L6 power tubes, and the schematic inside my amp specifies 6L6G tubes. Pins 1 and 8 have been tied together, so I'm thinking they were anticipating the possibility of a customer installing metal-envelope 6L6 tubes.

Given that, it's understandable why they'd want to bring the screen voltage down. (schematic in my amp calls for 350V on the screens). But does this still make sense with a quad of 6L6GCs? I figure constantly dumping a fair amount of current will help stabilize the bias, and I certainly don't need to make any more power, but if I don't need to be baking a cake inside the chassis I'd rather not do so. The preamp dropping chain would have to be adjusted, of course.

Oh yeah, and the OT appears to have a primary impedance of 4.4k. The photofact for the "AU-36" (not "-36B") specifies a 2.7k Raa.

So: let sleeping dogs lie, or remove the space heater?

[MarkusBass]

What is the voltage rating of C2 and C3 capacitors? If this is 500V or less, you have the answer. There should be nothing wrong with 540V supply assuming there are appropriate capacitors and you have correct bias voltage.
But most probably the capacitors have lower voltage rating than 600V, right? You would need to change them to stacked capacitors, which is not that complex.

Mark

[potatofarmer]

Yeah, I should've mentioned that I'm going to be stacking caps either way. I've got a pile of 22uF/450V 105C radial caps, so it's not going to be pretty but that should get me a nice long time before the next recap.

Also that's 540V with just the rectifier tubes in. It's also back biased so that's already knocking the plate supply to ~500V. Add in the heater draw and I'd wager the PT is going to start sagging. The earlier AU-36 photofact specifies a PT that's 365-0-365 at 385 mA.

Reduced supply voltages is a clear benefit though. I am worried a little about taking that current from the bleeder resistor and letting it pass through the output transformer instead. It is big, sure, but it's also got a million taps on it. And I am going to be adding individual screen stoppers, probably 1k.

[MarkusBass]

Yeah, I should've mentioned that I'm going to be stacking caps either way. I've got a pile of 22uF/450V 105C radial caps, so it's not going to be pretty but that should get me a nice long time before the next recap.

I'm afraid that you need slightly bigger caps. Similar amp Hiwatt DR103 has two 220uF caps stacked.

Also that's 540V with just the rectifier tubes in. It's also back biased so that's already knocking the plate supply to ~500V. Add in the heater draw and I'd wager the PT is going to start sagging. Reduced supply voltages is a clear benefit though. I am worried a little about taking that current from the bleeder resistor and letting it pass through the output transformer instead. It is big, sure, but it's also got a million taps on it.

I haven't noticed that 540V is without tubes. With tubes it will be less. The current depends on the bias so you don't need to worry that the increased current will pass through the OT. Just set the bias lower. You can easily test it with a variac (if you can get one).
If the voltage with tubes is below 500V, you could use 2x100uF caps. This may help you if you have space constraints.

Mark

[potatofarmer]

I'm afraid that you need slightly bigger caps. Similar amp Hiwatt DR103 has two 220uF caps stacked.

Ooh, the 5U4Gs wouldn't like that very much! I'll probably stick close to the existing 30uF just to keep those happy. I should go through the math in Merlin's power supply book to make sure the existing reservoir is OK and if I could bump it a little... but that'll probably be for a later day.

[MarkusBass]

Two stacked 220uF caps give 110uF total. Two stacked 22uF caps give 11uF. You most probably will not be happy with such a filter for the power tubes.

[potatofarmer]

Two stacked 220uF caps give 110uF total. Two stacked 22uF caps give 11uF. You most probably will not be happy with such a filter for the power tubes.

Very true! The reservoir needs to be kept tiny so the charging pulses don't pop my rectifier tubes though. The typical max for a 5U4G is 40 uF, and if I assume those rectifiers are going to share the current evenly then intuitively I would guess the maximum I could push it to would be 80 uF.

But if I take three 22 uF caps in parallel and run those in series with three more, then I get roughly 33 uF, which is within tolerance of what is on the original schematics. And of course with ballast resistors to ensure equal voltage sharing. Far from ideal, I admit, as I'll have to use 4 for each filter node.

But anyway.

For now, I'll leave the huge bleeder in the circuit but if no one sees any obvious reason why I shouldn't take it out and drop the preamp voltages to the original values, I'll give it a shot at some point and report back.