"Ian White, G3SEK" wrote:
> Oleg Skydan wrote:
> >OK. Consider I have a 40 uF capacitor in the PS charged to 3000V
> >(one that uses in the pulse laser PS). Will it be safe for electronic
> >crowbar to discharge it ?
> >
> Good call - you have to think about the maximum current that the crowbar
> will have to handle. The OE5JFL circuit is quite carefully designed to
> limit that peak current, using a combination of a small inductor to
> limit the rate of rise, and also a 10R sense resistor. (The inductor is
> only 30ft of wire, random-wound on a bobbin, but it makes a very
> significant difference. It also means that the 10R sense resistor
> doesn't have to be built like a conventional glitch resistor.)
>
> Bottom line: if you're going to copy OE5JFL's circuit, copy *all* of it.
>
> http://www.qsl.net/oe5jfl/flashover.htm
The Harris RF-110A uses a similar circuit for protecting the screen circuit
from tube flash over discharge (R series resistor at the bottom of the
screen's
regulating string of zener which act a SCR; all of the tubes screens - two
8122 drivers and two 4CX1500B finals- are OR connected by diode at
the SCR anode). It saved me the entire power zener string once, when
I had to verify a 4CX1500B (the tube was verified ok at hi pot before,
but the big bang occurred anyway and destroyed the RF124's PS HV fuse).
>
>
> >My idea was having a glitch resistor and vacuum relay in series in HV.
> >In case of overcurrent the glitch resistor will limit current until relay
> >will
> >switch off (~10ms). This will prevent discharging PS capacitor through
> >glitch resistor and tube.
> >
> This would work, but the current that you'd need to break may be 40-50A,
> at however many kV. It's always difficult to open a contact reliably
> under these conditions, and much easier to build a crowbar based on
> closing a contact.
>
If the tube needed to be protected is a $ 300 value, then a $ 15
russian V2V vacuum relay (35 amps at RF) may worth the money,
what ever its connection (crowbar or series switch) or reliability,
even if it will need to be replaced after let say only two "big bang"
(after a second flash over, I would have big doubts about the tube,
amp design, antenna, etc. or the operator!).
If one want a cheap BIG vacuum relay, there is the V9V, but
its speed may be not so good:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=36328&item=2568153347
Actually, if the protected tube is such expensive, a cheap russian
thyratron used as a crowbar or even a cheap surplus/used big
triode/tetrode used as a switch may worth the $, this as an
extreme solution (it may not even need the socket or cooling also!).
But as always, the real problem is the safety/total cost
(both $ and effort) ratio, which obviously will tend to avoid
the extreme solutions.
Anyway, good design may prevent the use of high value filtering
capacitor, which some don't consider, despite the good practice.
I can't understand why some insist for using very high value caps
for modern high gain tubes, as these tubes have good "straight"
anode characteristics. They may doing this for being sure that the
B+ "is stiff" and/or that in the case of a flash over the tetrode's
screen will be melt enough and nicely deposited over the other
tube's parts or the bang to be heard by the DX even louder
than their received signal?!
100 uF oil caps are often used for B+ filtering, when as low as
30 uF may suffice for the 1.5k amp at 3.5 kV and may have
other advantages also (as less rectifier peak current and less
stress for the transformer). The drawback may be not so
important, unless the last tenth of % for the amp efficiency is
the prime goal, which shall never be the case for a good linear
amp (those using 4CX10k for 1.5k out may already know this...).
All the Best,
73,
Traian,
YO9FZS
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|