I have been recapping filters in boat anchors and test equipment for two years now. Sometimes I have had advice to parallel them to get a certain value. Today I was recapping an Eico 950B cap tester and needed a 4 uf e-cap. So I used two 10 uf in series, which read 5 uf. That's probably OK. But in the course of reading up on the subject, I think I found that paralleling e-caps halves the voltage.

Is this correct?

If it is correct, where I paralleled caps in the filter of boat anchors (probably at least a couple of times), should I expect capacitor failures as a result of exceeding the voltage? Should I go through and re-re-cap those boat anchors?

Another question: The Eico 950B needs three precision caps, apparently to make the cap bridge more accurate. What passed for "precision" when this tester was built? Would 1% be OK?

Thank youo for your attention and help,

Bill

Putting them in series doubles the overall voltage rating. Parallelling them doesn't change the voltage.

2% is more than sufficient, and actually 5% would do, provided all of the caps were off by the same amount, since you could shift the dial to compensate.

The voltage rating of the cap does not change regardless if they are in series or parallel.

In parallel, each cap will see the same voltage. Therefore, the applied voltage must not exceed that of the lowest rated cap.
Two in series, each cap will see one half of the total applied voltage.

Thanks for your answers, Alan & Dale. It makes sense now. I don't know where I got that idea and I can't find where I got it, didn't save it.

And thanks, Alan for the tolerance figures. I had no idea, I was going to try to find 1%.

Thanks again,

Bill

By shifting the dial you will throw your resistance readings off. It is not that difficult to match up some capacitors and resistors.

Dave

E caps in parallel, lesser of both . Series, sum of both if they are otherwise equal.

That's true, but who would ever use one of these to measure resistance? Anyhow the old resistors have probably drifted too.

When you wire capacitors in series you should put a 330K (more or less) resistor in parallel with each capacitor to equalize the voltage between them.

Sal

Sal, a very good point, I forgot about that and will do it. Thanks much for helping.

Bill

Electrolytics self equalize, resistors are not necessary.

Dave

+1

True - but you know many ARFers and Newbies that will take a cap out of a burnt out CFL and buy another one at Radio Shack of the same value and put them in series.
Best to use equalizing resistors unless you know that both are from the same 'batch'.

+1

I agree with using balancing resistors with series caps, and I'll tell you why.

I've never seen balancing resistors left out of commercial practice, particularly in heavy-duty amplifiers such as some of the larger and later Fender units (Showman and Dual Showman, Super Reverb, and some of the later Bassman series, all of which I owned at one point or another).

At a time when manufacturers were shaving out every possible part in the interest of saving every possible penny of production cost, balancing resistors were not abandoned, and I find that noteworthy.

On the other hand, I had become aware of the balancing resistor controversy by the mid-1970's, when I began building gear seriously. The ARRL handbooks, for example, stressed balancing resistors; Orr's Radio Handbooks took the opposite point of view. Then I found this:



This was Frederick Terman, Sc.D, Dean, School of Engineering, Stanford University, and Past President of Institute of Radio Engineers, in his Radio Engineers' Handbook (New York: McGraw-Hill, 1943), First Edition, p.611.

With all due respect to those who hold the opposing point of view, that clinched things for me (and still does). I build and repair this way; I have done so since the mid-1970's; and I have yet to experience anything to suggest that I took the wrong tack.

Hope this helps.

Larry

Cornell Dubilier's take on balancing resistors:

They recognize the equalizing effect of leakage changing with voltage, but still recommend the resistors. "Pencil whipping" their formulas for determining the resistor values, indicates to me anyways, that they are the largest the resistors should be. For example, if the formula yields 1M for the value, then using 470K, or 330K resistors would be better.

Agreed in full, and many thanks for the info .

I stick with 330K 1-watters for the balancing resistors when I'm working with voltages in the 450-600 range, and elevate the resistance proportionally with voltage above that range.

And by the way, I've "whipped" many a pencil on math over the years. They always wear out before I finish the math...

Larry

We will never have agreement on this subject. Agree leakage will increase with voltage and tend to cause equalization.

What happens it one cap has little leakage at 600 volts but is rated 450? Voltage could punch through insulation and cause it to short. This would put all voltage across the remaining cap.

Adding resistors doesn't cost much and they act as bleeders, discharging caps when the radio is turned off.

+1.

L

In a nutshell.

Get with the program.

There have been considerable improvements in electrolytic capacitor technology since the 1930s.

Read this:http://antiqueradios.com/forums/viewtopic.php?t=153805

Alan, I saw the thread at the time it was current, and while I appreciate your qualifications and effort, and mean no offense, I stand by my position.

Had I been privy to a test run involving multiple samples in comparison, including pairs with the balancing resistors and without, all tested over a considerable period of time, then I could be persuaded by the results. But a single shot over a few seconds simply doesn't hit the target for me.

As for technology, certainly, lytic cap technology has improved since Dr. Terman's day. I haven't overlooked that fact. However, the availability of this technology is, in itself, NO guarantee that it is being employed, or at least properly employed, let alone universally employed.

Frankly, I am far more suspicious of lytic caps in general today (and not just lytic caps!) than I was, thirty, twenty, or even ten years ago. In view of the hordes of suspect offshore caps running around in all sorts of drag these days, I'm sure my skepticism on this score is understandable.

I have to take my chances with the caps at hand like anyone else, and I prefer to play it as safe as possible. And that includes balancing resistors in series stacks.

Larry