[AMPS] PS diode equalizing

[Steve Thompson]

-----Original Message-----
From: Vic Rosenthal <rakefet@rakefet.com>
To: Peter Chadwick <Peter_Chadwick@mitel.com>;
amps@contesting.com <amps@contesting.com>
Date: 05 July 1999 15:38
Subject: Re: [AMPS] PS diode equalizing


>
>Peter Chadwick wrote:
>>
>> I use resistors across the diodes, with no capacitors.
With 1kV diodes, I
>> use two 500volt resistors in series, although that's
probably 'belt and
>> braces'. If I was using capacitors, they'd be .01
microfarad 1kV working
>> micas or plastic film, not disc ceramic, because of temp
co and tolerance.
>> I've never had hash caused by the diodes switching
abruptly, either. With
>> 1N5400 series, they're relatively slow, so I wouldn't
expect it.
>
>30-odd years ago, I worked at a BC station.  We had a 5kw
FM transmitter whose
>power supply contained boards made of many, many 400 piv
diodes.  They were
>paralled with (I think 470K) resistors.  We lost quite a
number of boards
>(sounded like a machine gun when they went).  We tried to
match resistor values,
>etc. to no avail.  At some point, someone suggested that we
use .001 uf
>capacitors instead, and after that we had no further
problems.
>
>So far, the advice I've received has included 1) don't put
anything across the
>diodes, 2) put only resistors across the diodes, 3) put
only capacitors across
>the diodes.  I'm leaning to the capacitors, because of my
ancient experience.
>
>73,
>Vic, K2VCO
>Fresno CA
It's my guess that diodes in those days had a catastrophic
failure mode once they entered reverse breakdown. Modern
diodes tend to be much more controlled, sort of like a zener
to a limited extent.

I tested some 1N5408s which started leaking a few uA at just
under 3kV. This made me think - without added resistors or
capacitors, the voltage sharing along a string in reverse
bias will likely rely on the capacitances of the diodes.
Lower capacitance takes higher voltage, and the varactor
effect pushes the capacitance lower and so on (and the
reverse on the rest of the string) - you could end up with
some VERY uneven sharing. If the diodes have catastrophic
reverse breakdown you get domino effect. Controlled
breakdown alleviates the effect, shunt components avoid it
but deal you a different set of potential problems.

Steve


--
FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm