We need to keep in mind that the dielectric materials are polycrystalline semiconductors and do not behave in a linear fashion especially when dielectric layers are measured in a couple of microns, that just adds to the fun of it all:-) and kept me gainfully employed for decades. 

John Maxwell

Sent from my iPhone

> On Jul 18, 2018, at 4:51 PM, John Burke <[log in to unmask]> wrote:
> 
> 
> 
> 
> 
> 
> 
> 
> 
>        
>        Lol,
> If your theory is correct the “fluid stream” would follow the laws of fluid dynamics which it doesn’t since for the flow (current) to increase for a given resistance the pressure (voltage) would have to increase which is exactly the opposite of Bernoulli’s laws.
> And yes an ageing capacitor does resemble a balloon in capacity terms.
> Best regards.
> John
>        
> 
>        Sent from my iPad Pro
> 
> 
> From: 30171501600n behalf of 
> Sent: Wednesday, July 18, 2018 1:41 PM
> To: [log in to unmask]
> Subject: Re: [TN] capacitor measurement Ahhh, those blank, blank electrical engineers again.   Why can't it be
> simple?  Current is a flowing fluid stream, just electrons not water.
> Resistors are valves that increase or decrease the flow.  Capacitors are
> tanks where the fluid flow is stored.   And now you are telling me the tank
> size varies with age and temperature.  What are you electrical guys using
> for a tank - some kind of water balloon or something?
> 
> On Wed, Jul 18, 2018 at 4:04 PM Wayne Showers <[log in to unmask]>
> wrote:
> 
>> What has been described is all good information, but here is the yeah but:
>> 
>> At design:
>> Voltage coefficient is the largest concern and for this reason, each cap
>> should be overtightened and then loosened.  General Rule I follow in RC
>> circuits is to try to keep the capacitance as low as possible to counteract
>> the effects of aging.
>> 
>> In production testing:
>> So long as the specifications are known, I use a variant of the old
>> Mil-R-39008 (Carbon Comp Resistors) which is group average of 2X tolerance
>> and for individuals, 3X tolerance.  This has been effective in limiting
>> both bad product and test failures.  I also target testing at 60Hz for
>> capacitors to limit capacitance depression from higher frequency and/or
>> source impedance.
>> These attributes will limit false rejections at Incoming Inspection and
>> ICT.
>> 
>> As for 'refreshing' capacitors:
>> Baking above the Curie point is not always possible or for that matter
>> even desirable, so I only ever did this on a limited basis.
>>