[Amps] HV rectifier strings - to bypass or not?

Jim Thomson jim.thom at telus.net
Sat Oct 30 23:15:58 PDT 2010


Date: Sat, 30 Oct 2010 20:51:55 -0500
From: "Gary Schafer" <garyschafer at comcast.net>
Subject: Re: [Amps] HV rectifier strings - to bypass or not?

Why not? The output of the rectifiers provide some mighty spikes of their
own. The capacitors handle it just fine.

73
Gary  K4FMX

## per PSUD, plus putting a  scope across the diodes,  there is no big spikes...
provided there is no big spikes  coming in off the street. In the case where transients 
and spike come in off the street/drop wire, the electrolytics will absorb the spikes 
that get through the rectifiers'.  Now whether the lytics can handle freq spikes
etc, is another question.  Some of this may well  depend on the total V rating of
the string of lytics.  If you operate the caps at 90-96% of their rating, you have 
no wiggle room left.   That 'surge' rating you see stamped on caps is only for a few secs
max, per the cap makers. If you operate the caps at 70-80% of their V rating, they may
fare better.  Total cap value may well play into this too.  8 x  200 uf caps with low 2 A 
 ripple current rating is one thing.  8 x 2300uf caps with a 10A  ripple current rating
is another thing.   Either way, one would think the caps would look like a dead short to any
spikes?  Oil caps  just eat the spikes. 

## Some super fast transient suppressors  would probably do the trick. The Joslyn giant
mov's that get installed under your main 200A panel have a claimed  response time 
of < 1 microsecond. The dime and quarter size mov's you buy typ have a lot slower
 response times...like 50-400 microseconds.  Typ on the quarter sized mov's, the response
time is slow enough, that they don't clip off the spike, till  the peak line V is double....now that
I have seen a lot of times.  That's usually the reason given to size the qty  of diodes per each of the
4 x legs at double or triple the  no load B+ voltage.   You see that all the time in commercial
broadcast rectifier assy's. [ piv = triple the no load B+].  The same broadcast rect assy's  will
also include a MOV  across each diode as well.

##  If u draw out Carl's  idea of 1 x cap across each leg......when the spike comes along, the caps
will simply short out the spike..and none of it gets through to the HV caps.  

later... Jim   VE7RF 




More information about the Amps mailing list