[R-390] Cap Leakage

[Bob Camp]

Hi

There are a number of posts here in the archives on insulation 
resistance. Some of them relate to the fact that in order to fully know 
what is going on with the capacitors you need to put the full rated 
voltage (or close to it) on the caps. A megger (high voltage giga ohm 
meter) is often mentioned as the instrument of choice here.

Not to ignite a replay of previous debates on capacitors but I have 
never had a paper / black (or brown) beauty capacitor pass at any 
voltage ...

It should be noted that others apparently have had *very* different 
experiences with these capacitors.

	Take Care!

		Bob Camp
		KB8TQ



On Apr 3, 2005, at 5:27 PM, Dan Merz wrote:

> Hi Bob, Dave, Lee  et al.,  once again fine coverage here.  I was 
> thinking
> BAMA when someone mentioned T0-6a was used and that "insulation 
> resistance"
> was a standard measurement,  at least for the TO-6a instrument. I 
> didn't
> realize that - the only vintage capacitor checker I ever had left my 
> hands
> pretty quickly; it was a cheaper variety and the only plus I could see 
> for
> keeping it was that it actually put high voltage on the cap when 
> checking.
> My interest was renewed by the discussion.  So I had to take a look, 
> which
> was a 2 hour download of the BAMA TO-6a 8 meg file on 56k modem but it 
> was
> worth it.  It looks like insulation resistance with the TO-6a is 
> measured by
> matching instrument grid current to the capacitor leakage,  with some
> appropriate resistances  to scale the resultant current measurment in 
> the
> plate circuit of the instrument tube.  I would guess "nil" here is 
> still a
> pretty high resistance,  maybe in the 100 kohm range judging from the
> resistors in the circuit.  The manual indicates 100 Megohm to 
> 100KMegohm
> ranges so maybe the lower limit is even above 1 megohm.  Maybe someone 
> else
> knows the answer. I can't see the meter scale in the downloaded pic 
> clearly
> enough to tell. Thanks for the clarification (only downside is I'll 
> look
> twice at a T0-6a at some future swapmeet and be tempted),  Dan.
>
> -----Original Message-----
> From: Leland Bahr [mailto:pulsarxp at earthlink.net]
> Sent: Sunday, April 03, 2005 11:23 AM
> To: Bob Camp; Dan Merz; R-390A List
> Subject: Re: [R-390] Cap Leakage
>
> Nice post, Bob!  I enjoyed reading it a lot.
> Lee, w0vt
>
>>
>> Hi
>>
>> An ideal capacitor would pass no current at DC once it had been
>> charged. In fact it passes no DC current at all, it just stores it up.
>> That's another issue though.
>>
>> Back to insulation resistance ....
>>
>> So we put a DC voltage on a real capacitor and *suprise* it has
>> leakage current. It passes some current at DC even though an ideal
>> capacitor should not. This is simply because we can't make an ideal 
>> part.
>>
>> In order to decide just how good a real capacitor is we need to have a
>> measure of how much it leaks. The leakage behaves like a resistor the
>> easy way to describe it is as a resistor across a an ideal capacitor.
>> The bigger this resistor the better the capacitor.
>>
>> So far so good ...
>>
>> Unfortunately there are two ways to describe a resistor. One is in
>> ohms (zero is bad in this case) the other is in 1/ohms (conductance or
>> Siemens). In the case of conductance zero is good. You have to be
>> careful to be sure which your machine is calibrated in ...
>>
>> Just to make things even more complex ....
>>
>> The specification on capacitors is normally written in terms of
>> megaohm microfarads. The bigger the capacitor the more it leaks. The
>> more it leaks the lower the insulation resistance.
>>
>> The capacitors in question all should have insulation resistances in
>> the hundreds of megaohms if they are working right. Anything below
>> about 50 megaohms is likely to be a defective part. This of course
>> only applies to the parts that are not electrolytic.
>>
>> The electrolytic parts often are rated in leakage current. In an R-390
>> anything below a few miliamps is fine. A miliamp at 300 volts gives
>> you a third of a watt in heat in the capacitor. At ten mils you are up
>> to three watts and the capacitor can get a bit warm.
>>
>> The problem with leakage measurements is that they do not totally
>> represent the situation with the capacitors. On the black beauty caps
>> the leakage does not entirely behave like a resistance. It goes up and
>> down in bursts. The fancy term for this is popcorn noise. When you get
>> a burst of noise (or current) the voltage on the capacitor drops
>> fairly quickly in a typical R-390 circuit. This gives you an
>> intermittent snap crackle pop noise in the background on an otherwise
>> quiet signal. Since atmospheric noise can do the same thing it may
>> take a while to figure out what's going on.
>>
>> None of this is to suggest that you should replace otherwise good
>> capacitors. The issue is that bad capacitors can, but don't always do
>> cause real problems in a radio.
>>
>> About the only other point to make is that you do not want to replace
>> the paper or plastic insulated capacitors with electrolytic 
>> capacitors.
>> The leakage levels on all of the electrolytics are *much* higher than
>> the leakage of the plastic or paper parts.
>>
>> Hope that helps some.
>>
>> 	Take Care!
>>
>> 		Bob Camp
>> 		KB8TQ
>>
>>
>
>
>
>