[R-390] CAPACITOR FIGHT
David Wise
David_Wise at Phoenix.com
Tue Aug 5 12:42:19 EDT 2014
For C553, TC, VC, DF, DA, even IR and piezoelectricity, are irrelevant if held to real-world limits.
As long as it doesn't short-circuit, you're golden. (Or open, but most failures are shorts.)
I've read that about self-heals too. Unless the material and construction are closely controlled, each self-heal event contaminates the surrounding dielectric. If you keep cranking the juice, eventually they'll cascade to a short. There are papers about specific metallization alloys, how thick a dielectric to use, and how tight to wrap it, all looking for the sweet spot where this happens the least.
Dave Wise
-----Original Message-----
From: R-390 [mailto:r-390-bounces at mailman.qth.net] On Behalf Of Charles Steinmetz
Sent: Tuesday, August 05, 2014 6:21 AM
To: 390 list
Subject: Re: [R-390] CAPACITOR FIGHT
Dave wrote:
>I tried to calculate the reliability of film/foil vs. ceramic for
>C553 and got unclear results
There are other important factors to consider besides
reliability. Unless the ceramic is type C0G (NP0), the voltage
coefficient of capacitance, dielectric absorption, and dissipation
factor can cause unwanted distortion in the IF. 0.01uF, 1000v C0G
caps are available (e.g., AVX SV01AA103JAA), so this is an option.
Maybe it's just tradition, but I think I'd still rather use a film-and-foil.
>I also simulated a metalized-film self-heal event in SPICE, and got
>an 80mA current peak that lasted 10us. I don't think this will heat
>the wire enough to fuse it, but who wants to test it when film/foils
>are still readily available?
80mA is a lot of current for the tiny wire in a mechanical filter
drive coil. Also, once a metalized film cap starts having
self-healing events, they tend to become more and more frequent. No, thanks!
Best regards,
Charles
More information about the R-390
mailing list