[Amps] Commander HF 2500

[Jim Thomson]

Date: Thu, 25 Jan 2018 07:15:40 +0800
From: Alek Petkovic <vk6apk at bigpond.com>
To: Jay Sturtevant <k2ztdx at gmail.com>
Cc: amps at contesting.com
Subject: Re: [Amps] Commander HF 2500

<No. Not necessarily but what is critically important, is to check the 
<value of the resistors across each of the filter caps.

<My friend, K4HB, had a filter cap erupt and burn out the section of 
<board it was mounted on. He ordered a new set of caps and I advised him 
<to check the resistors.

<Not a single one was in spec. The one across the faulty cap was open 
<circuit.

<With new caps and resistors and a repaired board, he is back on the air 
<with the 2500.

<73, Alek.
<VK6APK

##  This is good advice.  The V across any given cap will be proportional to the 
equalizing resistance value.  Caps  that happen to have  higher than normal  values of
resistance  will also get the highest value of dc voltage across em.  The actual  value of 
eq resistor is not the issue..... what Is important is  to make  sure  ALL the  eq resistors
are the SAME value.   Worse case scenario is when a resistor goes open... or connection
at either end of the resistor is lousy.   Infinite resistance across  a  given cap  means that cap 
gets the FULL  B+ applied to it !!   IE:  a 2800 vdc  supply  will  apply  2800 vdc across that
one cap, which of course blows into a million pieces, damaging everything around it. 

##  I use Vishay brand  100 K  at  3 watt  MOF  eq resistors across  any 450 vdc  electrolytics. 
These are the   1 %  Tolerance resistors.....available  from Mouser.   100 per box.  I bought
several boxes of em.. and tested ALL  400  of em.. they ALL measured the same 99,800 ohms. 
Tolerance was  zero. 

##  Vdrop across  each lytic was identical !    The eq resistor  does not have to be wired directly 
across the cap terminals.   They can be several inches away if you like.   BUT the connections to
the cap have to be  100%  rock solid. 

##  Vishay PN  CPF3100K00FHB14 .      Customer PN 71-CPF3-F-100K-T2    Then no more
need to play matchup.  Just grab  XXX  amount of resistors to do the job at hand.   My lytics have 
machine screw connections and use an internal lockwasher.   I use sta-kon crimp connectors, closed
round eye types.  Crimped, then also soldered.   Then crank em all down tight.   I also wire a 1N5408
or  6A10 in parallel  with each cap.. diode is wired RVS connected. 
If any leg in the FWB or  FWD shorts, you will end up with raw AC, on half cycles, across the entire series 
string of lytics.  Lytics dont like raw AC.  Diodes are cheap insurance, esp on expensive lytics like the types
I use.  2100 uf,  2300uf, 3600 uf, 10,000 uf.   Diodes get their own sta-kon crimped + soldered ring
connectors. 

##  100K  at 3 watt  MOF is a good value.   Even with as many as  24 x caps  in series, the resulting
eq bleeder current swamps the internal leakage current of each cap  by well over 20 X. 
Internal leakage current in any lytic drops like a rock as soon as the applied vdc across that cap 
is below its stamped Vdc rating.   IE:   dont run a 450 vdc rated cap at any more than  400 vdc.
I usually run mine at no more than   70-75 % of their  450 vdc rating.  There is NO requirement
to be using lower values of eq resistance. 

##  max  plate current you can safely run is typ way less than the cap max ripple current  rating. 
Take the ripple current rating...and divide by 2.56  to get the max plate current you can safely use.
If the amp  runs at say  1.5A....  MIN ripple current rating should be   1.5  X 2.56 =  3.84 A. 
Typ the bigger uf caps  also have higher ripple current ratings.  Like  10 A,  13A  etc.   You can also
parallel lytics to increase the ripple current rating. 

later....  Jim   VE7RF