voltage divider (Steve Conner, please?)

[JC@]

Hi.
I have built something like a power supply to test capacitors for leakage. That is a transformer, a rectifier bridge, a little filtering and the components needed to test the caps.

I get 750 volts out of it which is ok to test 600 to 1000 volts coupling caps, but too much to test something like 200 volt ones.

Now, the problem I have is that there's quite a bit too much current to use a simple resistor voltage divider arrange to drop the voltage to other values like 200 and 400 volts, because to test the caps I guess I only need a little current and resistors would be huge wattage to handle so much a voltage drop.

Which could be a better approach to obtain approx. 200 volts and 400 volts from my 750 volts power supply?

Thanks in advance.

[bob p]

i can think of a few ways to attack the problem:

1. high power resistors. you can often find them very cheap on ebay.
2. variac
3. adjustable regulated power supply

[Enzo]

Too much current? The load on the power supply determines the current.

A 500k and a 250k resistor in series makes 750k. Put that across 750 volts, and it will draw 1ma. Put the 250k on the bottom and tap 250v. 250v across a 250k resistor will dissipate one quarter watt there.

Surely we can come up with common resistors for your divider without special parts.

[bob p]

i guess that one reason to worry about current draw is the likelihood that in using his cap testing rig, its inevitable that sooner or later he'll be testing a bad cap. as Enzo noted, the load will define the current draw, and a properly functioning circuit won't draw excessive current, so high wattage resistors aren't necessary. otoh, what happes if he's got a cap that's got a dead short to ground? current is going to be high, and components are going to start failing unless you recognize the problem or the fuse blows first. if that's the case, overbuilding isn't a bad idea, and maybe high power reistors aren't a bad idea either. maybe some sort of current limiting device would be handy, like a light bulb limiter used with a variac. of course, you could always go to BAMA and look at the schematics for the Heathkit cap testers. they might have some helpful ideas in their design that we haven't talked about.

[JC@]

Thank you!

Enzo, you are right and I am reading over these simple rules and formulas right now. But have to say that in my first attempt one of the resistors forming the divider went into smoke. My fault for sure..

My goal was to keep it as simple as possible but have to add a fuse and a few other things.

bob p, Great tip about looking to Heathkit's own testers. My first thought was to use a little variac which would avoid the issue of any switching.

[bob p]

if you google for "bama heath" you'll find the schematics. have fun.

[Steve Conner]

Someone call mah name! Every time I've tested electrolytic caps, I've just used a power supply the same as yours, hooked up to a variac, with something like a light bulb to limit the current if the cap shorts out. The variac is no problem since I have a variac and isolating transformer hooked up in my workshop permanently anyway.

Like you said, the power supply needs a transformer to boost the line voltage up. It's not so bad over here, since it's 240V which rectifies to about 320V. I usually just grab a transformer out of my junk hoard to suit the job. To test real high voltage parts I use one from a junk microwave oven (with a lot of respect!) to tickle them with up to 2000V.

I have one of those old tube regulated power supplies that I could use, but it only goes up to 300V, so it's not much good for testing 350V and 450V caps. And ironically, it's well overdue for a cap job itself.

The voltage divider thing would probably work too. You don't need to pass a huge current down the divider, since good capacitors will draw practically no current from it once they are done charging up. You could make a divider chain with taps for 100V, 160V, 200V, 250V, 300, 350, 400, 450 etc. and just hook the capacitor to the appropriate tap. Or you could use a large wirewound pot. I've seen 25 watt ones the size of a cupcake.

A simple method to calculate the divider is like Enzo says: make it so all the resistors add up to 750k Ohms. To get the above voltage taps, you would need resistors of 100k, 60k, 40k, 50k, 50k, 50k, 50k, 50k, and 250k ohms.

You can scale all these numbers as long as you do it by the same amount. So you could just as easily use 10k, 6k, 4k, 5k, 5k, 5k, 5k, 25k. This would dissipate 7.5 watts in total.

[Enzo]

Heathkit made some good old testers, and my Eico 950B had a variable voltage cap test as well. You should be able to find that schematic too.

[bob p]

On a related note, here's a link to a very useful DoD publication:

Storage Shelf Life and Reforming Procedures for Aluminum Electrolytic Fixed Capacitors.

To find the online link, I had to google for the DoD document number: mil-hdbk-1131A

[JC@]

Thank you all. I appreciate your help and comments so much.

[JC@]

What could be the current draw of a neon bulb like those used in tremolo circuits when operated from my 750 volts power supply? I'm using a 330K dropping resistor for it and it works.

I'm using one in my tester project..