----- Original Message -----
From: "Jim Thomson" <jim.thom@telus.net>
To: <amps@contesting.com>
Sent: Friday, December 06, 2013 8:47 AM
Subject: Re: [Amps] PARALLEL CAPS IN OUTPUT
Date: Thu, 5 Dec 2013 03:03:44 +0100
From: "Peter Voelpel" <dj7ww@t-online.de>
To: <amps@contesting.com>
Subject: Re: [Amps] PARALLEL CAPS IN OUTPUT
It?s quite simple without measurements.
Lets assume the amp is tuned for full legal output into a low SWR on 80m
which needed 1000pf of load capacitance.
Then suddenly the SWR goes sky high because the coax connector went open.
How can the voltage across the load capacitor become 5KV when the
reactance
of the load capacitor is just 45 ohms on 3.5MHz?
That would require about 110 Amperes of RF current through that capacitor.
And where is that big current coming from in a 1.5KW amp, not in a 100KW?
73
Peter
## back in the late 1970s.....with my single 4-1000 GG amp....it did aprx
1900w pep out.
It used a clunky 3PDT mech relay. One day the output side didn’t quite
make contact, from
all the hot switching etc. 130w of drive applied.....and NO ant connected
for a split second.
The result was the 1500 pf cardwell air variable load cap was welded
shut. It had aprx 1500 V
spacing. I used a flat ignition file to grind the weld free. It
happened a 2nd time months later..same deal.
## I found the safest way to ensure the expensive vac load cap never
opens up is to install an adjustable
spark gap across it. Then adjust it for a break down V that is less than
what the peak V rating is on the
vac load cap. In normal operation, the amp will shut down when ant SWR
exceeds 2:1 or whatever
threshold I desire. The spark gap is adjusted for the eq of say a 4:1
SWR. If the ant fails...and coax goes open,
and IF the high swr protection doesn’t shut down the amp 1st.... the
simple adjustable spark gap will ensure
the vac load cap will survive.
## If you have seen how the mech screw adjustments are on any vibroplex
keyer, the adjustable spark gap
is very similar. Just take a pair of 10-32 machine screws....and grind
the ends to a fine point. Make one fixed,
and the other adjustable, complete with lock nuts to maintain a precise
gap. Some bar stock etc, can serve to
support the machine screws.....just tap threads in em.
## For folks with PI-L networks in their amps.... the load cap is typ at
a 200-400 ohm point already..and peak
V across the load cap is already a helluva lot higher than the load cap
in a PI net.
## For folks with a PI net, make sure you use a piece of 50 ohm coax
between the load cap and the TR relay.
And with shield bonded to chassis at BOTH ends. If wire or cu strap is
used instead...then u end up with a huge Z bump.
** The coax isnt long enough to act as a transmission line at HF....in most
amps anyway.
That alone will put the load cap at something higher than a 50 ohm
point... increasing the peak V. On 10m band, the
stray L of a piece of wire instead of coax will make the PI now appear
as a PI-L.... putting the load cap at a much
higher Z point. same deal peak V increases. Now with a wide open on
the ant, coax cable etc, the load cap
will weld shut.
** All the coax does is act as a shielded cable with some value of C.
At 10M even 1/10 wave of cable is 1.6' so it is invisible as coax.
## as far as the current ratings on vac caps go.... read the various
graphs very carefully. They only handle max
current when FULLY meshed. Most graphs will depict a peak at some
freq..like 4 mhz. Below 4 mhz, the current ratings
drop way off. At lower freqs, the XC will increase for a given cap
value. Current drops off because of either the
cap is being used at less than max C. and or the freq is below 4 mhz.
Both cases will cause an increase in XC.
Below 4 mhz the cap is said to be V limited. IE: the V would have to be
increased well above max V ratings to
increase the current. Above 4 mhz, the vac cap is now said to be
current limited...again when fully meshed.
IE: you cant run max rated V on freqs higher than the peak depicted in
the various charts.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1432 / Virus Database: 3658/6396 - Release Date: 12/06/13
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|