Sovtek MIG60 high plate voltage - filter caps rating?

[Sansteeth]

Hey everyone,
yesterday I decided to open up my old Sovtek MIG60 and rebias it.
It's a EL34 push-pull and it's one of those amps with real high plate voltage (around 600 to 630VDC), screen is at 325VDC
I have two questions:
Why would you go for such high plate voltage when you are going to have a screen voltage 300V lower? My understanding is that the screen voltage is more efficient to increase plate current (which is what we want of a power tube, I think?) than plate voltage, so why the insane plate voltage, eats more tubes, requires a beast of a power transformer, for a 50W power section, that doesn't sound logical to me, what am I missing?
The other question I have is: All my filter caps on B+ are 100uF rated at 500V, they are JJ's, not the original ones (schematic calls for 100+30uF at 350V). Should they be all rated at more than 600V? Or at least for the node that goes to the OT's central tap?

Thanks a lot!

[shooter]

G2 is limited by design, you can find the #'s in the datasheets.  I believe they can still "control" the plate well at 300vdc

look at the caps in the schematic, 2 in parallel, with 2 in series, so the "effective" cap vdc is 700v,

[Sansteeth]

Hey, thanks for your response,
my RCA Receiving Tube Manual tells me 425VDC max on the screen. I've mostly spent my energy understanding preamp circuitry and the workings of a triode so I'm pretty new to power section.
What would be the goal for such high plate voltages? My guess is that because it's (only) a 50watter, they try to get as much headroom as possible from the power section. But couldn't that be done with lower plate voltages and higher screen voltages? Your post seems to imply you would have to subject the screens to higher VDC, closer to max. rating, and surely we don't want that, right? 

look at the caps in the schematic, 2 in parallel, with 2 in series, so the "effective" cap vdc is 700v,

I did spot the parallel situation (that effectively does not increase the voltage rating) but I didn't notice the caps in series, thanks a lot shooter!

[shooter]

My guess is

 
spent my career talking n working with engineers, I just glaze over when they explain "why".  my job was to fix it, then go rock climbing 

[Sansteeth]

spent my career talking n working with engineers, I just glaze over when they explain "why".  my job was to fix it, then go rock climbing 

I can't blame you! 

[sluckey]

Mig 60 is a 60 watt amp. Mig 50 is a 50 watt amp. Big difference is the power supply. The Mig 60 runs much higher plate voltage to be able to squeeze out 60 watts. You still have to run the screens at a lower B+ (typically about 1/2 the plate B+) in order to not burn up the tubes.

     https://el34world.com/charts/Schematics/files/Sovtek/Sovtek_mig50_60.pdf

[PRR]

> My understanding is that the screen voltage is more efficient to increase plate current (which is what we want of a power tube, I think?) than plate voltage,

Power is V*I. If you have high V, you don't need high I.

Tube current is limited by cathode area and heater power. Both expensive. You get the most power for your tube dollar by raising the plate voltage way high, use just enough cathode and G2.

OTOH different plate and G2 voltages complicate the power supply. Yes, the "usual plan" in small (<100W) Audio amplifiers is to design a tube and amp so G2 can be cheaply dropped from plate power.

Look at 6550 data. Up to 50-60Watts they run plate and screen about the same, but for the 100W condition they take Plate way high and set G2 at about half that (the next most economical power scheme). Some EL34 sheets show a similar high-V condition.