[AMRadio] High Voltage Power Supplies

Jim Candela jcandela at prodigy.net
Sun Oct 30 09:42:16 EST 2005


    It might be helpful if you describe that power
supply topology. Is it full wave with transformer
center tap, full wave bridge, and is the filter a pi
type, choke input filter, etc. If this is an option
for you, make a sketch, and scan it into a JPEG file,
and upload it to the net soemwhere. I have net space
if you email it to me. This way we can all look at
your schematic, and comment better without guessing.

   In general diodes fail from two transient factors.
The first is current surge, and the second is
avalanche reverse voltage breakdown. Since you added a
step/start circuit already, the current surge issue
should be contained. That leaves reverse breakdown. As
Brian mentioned, a transformer primary varistor (like
V130LA10A, for 115 volts, or V250LA10A, for 220 volts)
might help from power line transients.

    If you look at the Bill Orr handbooks around 1970,
Bill goes into detail describing diode failure modes,
and ways to protect them. This includes a custom
series R-C across the transformer secondary, and
across the filter choke (if choke input). Today's
diodes are tougher, and this precaution is often
unnecessary so long as good diodes are used, and the
diode PIV rating is at least 2X what the formulas
state you need. However diodes like the 1N4007 (1A 100
PIV) need protection, whereas diodes like the 1N5408
(3A 100 PIV) are a lot more rugged. In researching
diodes, look for the term "controlled avalanche".
These are the best because they can take repetitive
PIV spikes beyond rating without turning into a piece
of wire. The 1N4001 to 1N4007 series are not in this
catagory. Unfortunately the data sheets sometimes omit
the "controlled avalanche" term.

Jim Candela

--- Ed Swynar <gswynar at durham.net> wrote:

> 'Morning All,
> I'm curious as to why I have to replace the silicon
> diode strings in my 2500 VDC power supply about
> every 8-10 years, or so...I just went through the
> exercise again early this morning --- turned on the
> B+ to my 2 x 813 linear, & got nothing back for my
> trouble but smoke from inside the enclosed p.s.
> unit...
> Opened it up, & the resistor across the relay in my
> time delay circuit --- transformer primary side ---
> was fried. I checked the conductivity of both legs
> of my diode string (it's a full-wave set-up), & sure
> enough, one leg was A-OK, but the other displayed
> conductivity on BOTH sides of each diode...not good.
> Is this the reult of transients / voltage spikes
> somehow "overwhelming" the diodes...? 
> It used to be more frequent prior to the
> incorporation of the delay circuitry --- but,
> despite its presence, this HAS to be at least the
> second time that a re-build was in order.
> Any thoughts re. a dose(s) of possible "preventative
> maintenance" for the future...? I'm at a loss as to
> anything else here that I might try...
> Thanks in advance...
> ~73!~ Eddy VE3CUI - VE3XZ
> AMRadio mailing list
> Home:
> http://mailman.qth.net/mailman/listinfo/amradio
> Help: http://mailman.qth.net/mmfaq.html
> Post: mailto:AMRadio at mailman.qth.net
> AMfone Website: http://www.amfone.net
> AM List Admin: Brian Sherrod/w5ami

More information about the AMRadio mailing list

This page last updated 24 Feb 2018.