Hello,
It is most certainly true that this is a complex matter.
Maybe the most appropriate action now is to find out what capacitance the
capacitors referenced to
in the original post have. If they are too low capacitance for blocking
capacitor service in the intended circuit,
the questions will have to be reformulated.
My own practical experience of interpreting the RF performance of "doorknobs"
comes from selection and use of Thomson-LCC (now Kyocera TPC) 3000 pF 20 kV AC
rated units, 60 mm diameter (rated at about 5 kVAr at 8 MHz), as replacement
plate blocking capacitors in Granger and Marconi HF transmitters in the 5 to 10
kW power class. No known failures did occur during the remaining lifetime of
the transmitters.
Design curves from this manufacturer, albeit somewhat difficult to interpret,
do show an approximate 1/f relationship for maximum RF current vs frequency.
Also, 2000 pF 10 kV ceramic blocking capacitors (Centralab manufacture) of
about the same size
(ca 80 mm) held up nicely in Collins 208U-10A amplifiers even when, somewhat
carelessly,
tuning up at > 15 kW levels.
I am looking forward to sharing your findings in this matter.
73/
Karl-Arne Markstrom
SM0AOM
----- Original Message -----
From: <MorgusMagnificen@aol.com>
To: <sm0aom@telia.com>; <amps@contesting.com>
Sent: Sunday, January 05, 2003 10:23 PM
Subject: Re: [Amps] KVAR Caps Rating
> You better take another look at the data. You will find that the KVAR ratings
> for the large doorknobs is more like .5-1KVAR at RF. The maximum current
> rating at 1MHZ is only approx. 2A. How well is this going to play on 10M?
> The assumption of a 1/f maximum current dependence is not quite true, as I am
> finding out as I study this issue. The 1/f rule is fine for skin-effect
> related metallic conductors, but the dielectric losses are considerably more
> complex.
>
> This whole issue of high-power capacitor ratings has bothered me for some
> time and I hope to have some new information to share in the near future.
> Manufacturers data is quite scant, leaving us to do a lot of guessing.
>
> Eric von Valtier K8LV
>
>
> --- StripMime Report -- processed MIME parts ---
> multipart/alternative
> text/plain (text body -- kept)
> text/html
> ---
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
|