Need to help to understand using 6x5 as half wave rect

[Mooly]

It all sounds like a plan

I was trying to think of suitable diodes... and it seems many of the one I remember have disappeared from the catalogues although the BY 255GP seems cheaply available (1.3kV).

VISHAY - BY255GP-E3/54 - DIODE, STANDARD, 3A, 1300V | CPC UK

 

[Mooly]

… off track … but I prefer your “2 or 3 solid state rectifier diodes in series, with PIV 'sharing' resistors in parallel with each.”

They're easy to make, and to test before committing to circuit. And they're remarkably cheap. And safe. "Over-engineered to perfection." Personally, I like 1 watt 1 MΩ PIV balancing resistors. They do the job, and they don't leak enough to even marginally affect the power supply ripple. And the rectifiers last forever - even with power line zapping with local lighting storms.

GoatGuy

Posted together.

The series diodes should certainly work although the resistors would need to be 'voltage rated' as well as just being safe in terms of power dissipation.

It all seems very do-able.

 

[hpeter]

I got in touch with another hobbyist who restores this kind of equipments namely, Eico, Heathkit, Paco, Knight etc.

First he told me was that there was no way I could get 720VDC from 560VAC!!!

 

[pchw]

Excellent, I will prototype with the 1N4007's that I have in the bin along with
the big military 2M 1% for the balancing resistors string. As these are cheap and I have voltage to spare, I can go with a string of 3 just to be safe ....

 

[pchw]

Success, sort of
1st pic is the readings across C1 and C9, close enough to the 520 / 200 split.
2nd pic is the flying wires and SS rect string.
The biggest surprise was that the SS rect only gave me additional 40VDC or so, not anywhere near 800VDC. FWIW, the secondary was 578 VAC. Regardless, it is a happy result.

Edit, forgot the attachments.

 

[BinaryMike]

I was trying to think of suitable diodes... and it seems many of the one I remember have disappeared from the catalogues although the BY 255GP seems cheaply available (1.3kV).

Mouser stocks the Rectron R5000F (5KV, 200mA) at $0.75 quantity one.

 

[Mooly]

That's looking good. I wonder if the ripple across the caps is swaying the DVM reading. In the simulation there was around 20 volts pk/pk on the positive rail and 10 volts on the negative.

Mouser stocks the Rectron R5000F (5KV, 200mA) at $0.75 quantity one.

Not one I'm familiar with but it sounds as though it could suitable as long as the cap charging currents were within limits. They are only small caps but the peak diode current could be quite high. Worth looking at though.

 

[pchw]

There is no doubt that the ripple is very high. The chance is good that the DVM reading is the average, not the peak to peak. Since I have SS diode now, I don't have the tube rect's input cap size constraint. I probably can increase the capacitance considerably to reduce the ripple .......

On the other hand, the small capacitance from the C1 and C9 stack is relatively small; therefore, the reversed voltage and the reversed current is not too big either.

 

[Mooly]

This is what the simulation shows for a ss diode and the peak diode currents. Adding some series resistance to the diode would tame this dramatically, as will the reality of real world components (the caps and their real world series impedance).