# Home Brewed Oil Capacitor Rectification

Discussion in 'General Electronics Discussion' started by Proschuno, May 28, 2013.

1. ### Proschuno

94
1
Aug 1, 2011
I'm attempting to build a 80-1000V variable power supply, and I'm stuck on the choice of capacitor; any capacitor with this voltage rating is over \$50, which I do not want to spend so I'm attempting to build my own, which instructions are easily found by googling.

1. Will an oil filled be good for this application?

2. If so, what are good values to use?

2. ### Proschuno

94
1
Aug 1, 2011
And for anyone curious, this will be used to power tubes, so current output shouldn't exceed 500 milliamps.

3. ### Raven Luni

798
8
Oct 15, 2011
Not sure you need to go to all that trouble for something as low power as that. One of the basic voltage multiplier designs should suit your needs - the stages are in series so component voltage ratings only need to be suitable for a single stage.

4. ### Proschuno

94
1
Aug 1, 2011
Great idea; considering the transformers I was going to use can supply at least 600 watts each, the size, weight and current consumption of the power supply would have been terrifying. But would current supply at the output be at 500 mA? I assume the multiplier behaves sort of like a transformer right?

5. ### Proschuno

94
1
Aug 1, 2011
So I should also explain that one of the transformers steps down to a 20 VAC winding; using two sextupler circuits will achieve the desired 1000 volts.

So using a voltage multiplier, does it multiply the input RMS or peak to peak?

Would the calculated output voltage = 20 * 1.414* 6^2 = 1018V, or just 20* 6^2?

And also, I have a bunch of capacitors of varying capacitances, most of them being similar though. I also have a bunch of diodes of varying voltage drops (i will have to count, i might have 12 of one kind.). Could I build a multiplier with varying capacitances and voltage drops, considering at such high voltages the drops wouldn't matter?

Last edited: May 28, 2013
6. ### duke37

5,364
772
Jan 9, 2011
1000V at 500mA is 500W, this supply will be big and heavy.

The trouble with voltage multipliers is that the capacitors are in series so that big capacitors are required to get the capacity. To get high voltages for electron microscopes, high frequencies are used to help this problem.

You should use a transformer with the required voltage and a string of capacitors to get the voltage capability. There will need to be equalising resistors along the string.
You should calculate the capacity required from the ripple that you can tolerate.

Feeding the capacitors directly will give Vrms * 1.414 peak.
A choke input filter will give Vrms * 0.9. This is much kinder on the components but adds a lot of weight.

Oil capacitor. Have you worked out the area of the plates and the separation to get the capacity you need? What happens if the plates touch?

7. ### Proschuno

94
1
Aug 1, 2011
Ok, so I have a step down transformer that has several windings of 4, 20, and 40 volts, then I have another that can step up 120 to about 2400 Volts (this second one was taken out of an industrial microwave). My plan was to tie them back to back, then rectify, filter, then regulate this voltage.

I plan on making it able to switch between the windings that get fed into the step-up transformer to make things more efficient.

Then to filter it I have a bunch of electrolytics rated for 400V then some for 200V which will be tied in series. I didn't realize that putting these in series both gave me the voltage rating I wanted, and the decrease in capacitance far beats those hefty oil capacitors.

The next question is what would be a good voltage rating for a rectifier, 1200V? 1400V , or higher maybe?

8. ### duke37

5,364
772
Jan 9, 2011
The microwave transformers that I have seen do not have an isolated winding which you will need.

Rectifiers are cheap so go for eight or ten 1000V 1A diodes in series. In the early days, balancing resistors and capacitors were used but I think they caused more problems than solutions. Avalanche diodes came in to do the job without the extra components, I do not see this term used with modern diodes.

Here is a pictue of 4 + 4 diodes to replace a 5R4GY tube

File size:
162.6 KB
Views:
126
9. ### Proschuno

94
1
Aug 1, 2011
So that totally clarifies the rectifier question. But question about isolation, I'm using a conventional step down transform which is isolated, would stepping it back up with the microwave transformer break that isolation? The microwave transformer does use it's own body as a conductive path now that I realize, and so I assume bolting this to the supply chassis which would be grounded ,would eliminate any isolation?

Last edited: May 28, 2013