[R-390] killer and mica caps

[Jacques Fortin]

Larry, Chris....

I have seen so many film/foil caps that gone short in hi-end tube amplifiers (taking one of the output tubes and the HV fuse to hell at the same time) that I cannot tell them better or more immune to "discharge" incidents than self-healing metallised film caps. In fact they are worse.
All capacitors may suffer from a short between plates, and I tend to trust more the ones of vintage manufacture than the recently manufactured ones.
If basic care (clean room condition) is not taken when the capacitor plates are wound, airborne dust can came trapped in the winding.
With constant pressure and time, any single dust particle can pierce the dielectric and start a breakdown event.
This will cause a permanent failure in most film/foil types, and a self-healing effect in a metallised film one.
Remember that, for a metallized foil cap, most of the current involved in fusing the metallization layer(s) come from the capacitor charge itself and usually causes only a small voltage variation across.
This is very unlikely that the voltage across the cap can go to zero (then applying full B+ voltage to the input coil of the R-390A mechanical filter, as discussed here at the beginning) before the few atoms thick metallisation layer gives up and clear the short (a self-healing event last typically 10nS, so not enough time to fuse the coil winding either).
>From experience, metalized foil types are safer than the film/foil types.
Still afraid about the failure probability ? OK: let's go take a look at stacked metalized foil+foil types (Vishay/Siemens) or at floating-electrode types (WIMA MKP10, FKP1).
Or... take any good metalized film cap and using a hi-pot or leakage tester, apply two times it's maximum working voltage across for a full minute: if this will not clear any pending "self-healing" event, nothing will do, right ?

Just my 2 cents worth anyways...

73,
Jacques, VE2JFE 
Montreal

-----Message d'origine-----
De : R-390 [mailto:r-390-bounces at mailman.qth.net] De la part de kc9ieq via R-390
Envoyé : 10 octobre 2016 20:25
À : r-390 at mailman.qth.net
Objet : Re: [R-390] killer and mica caps

Edit, I meant metalized film, not film/foil. Chris 


Sent from my SMRTphone
-------- Original message --------From: kc9ieq via R-390 <r-390 at mailman.qth.net> Date: 10/10/16  7:21 PM  (GMT-06:00) To: r-390 at mailman.qth.net Subject: Re: [R-390] killer and mica caps 
Conversely, if a film/foil cap shorts, it stays that way and will DEFINITELY take out the filter.  I've contemplated placing a very small value ceramic or mica cap in series as an added safety measure. 
Thoughts?
Regards, Chris kc9ieq


Sent from my SMRTphone
-------- Original message --------From: Larry H <dinlarh at att.net> Date: 10/10/16  6:47 PM  (GMT-06:00) To: 'Chuck Collins' <chuckcollins at prodigy.net>, R-390 at mailman.qth.net Subject: Re: [R-390] killer and mica caps 
Chuck, There is a good type and poor type of C553 replacement - common manufacturing methods are 'film and foil' and 'metalized foil' (these are sometimes referred to as 'foil' caps and usually have a 'self healing' property).  The problem is that when a 'self healing' event occurs, an arc has occurred internally.  This arc (very high current pulse) may destroy the filters we are trying to protect.

Use a 'film and foil' type of cap and be safer.  An example of some are SBE 715P orange drops, but any high quality 'film and foil' type will work.  It should be 500v or higher.

Regards, Larry