Amps
[Top] [All Lists]

[Amps] Sensible filter Cap sizing

To: <amps@contesting.com>
Subject: [Amps] Sensible filter Cap sizing
From: "Jim Thomson" <jim.thom@telus.net>
Date: Sat, 26 May 2012 01:54:43 -0700
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Date: Thu, 24 May 2012 11:33:19 +0000 (UTC)
From: Paul Decker <kg7hf@comcast.net>
Subject: Re: [Amps] Sensible filter Cap sizing



Hi Gang,

I've had a couple comments about 5x500V not being enough for 2488V and that I 
should use 6 or 7 caps. When I was first building my ps, I was thinking along 
the same lines, and if I had 6 or 7, I likely would have put them in just for 
good measure. On the other hand I also rationalized it by thinking those 
capicator designers have already put in a safety margin to meet their MTBF 
stats, and I know that my p-p best case is 2488, even exceeding 2248 it 
woudln't be a normal or prolonged use case, so why should I add yet another 
safety margin to the system. So far, so good. I guess I could replace them with 
the 8 3900uf 450V caps I got for free the other day.

73, 
Paul, kg7hf

##  Its gonna  blow up in your face !   And you can forget about any “ 525 vdc 
surge rating”   either.   Any surge rating is for less than 2-3 secs max.   
With the caps maxed out, your leakage current will be sky high. 
Did  you even measure it.     How do you know the voltage is dividing equally 
between ALL the caps.     Did u measure the V drop across EACH cap.   What 
exact value is EACH of the eq resistors across each cap.   The value is not 
important..but 
they all gotta be the same !     IE:  all 90 K.... or all 100 K.   I use the 
vishay brand  100K 3watt  from mouser. These are 1% tolerance type..so no more 
playing matchup by hand.   100 per box... and I bought several boxs of em..and 
3 different 
lot numbers too.   I checked all 300 of em... and they are all within just a 
few ohms of each other     OK, now the v drop across each ca will be the same.  
I instead, you use say 5%  tol eq resistors..watch out.    the caps are in 
series..and so are the
eq resistors.  The bledder current will be the  SAME through all the resistors. 
  BUT the V drop  will be ...  bleeder current X resistor value.  The resistor 
with the HIGHEST  value  will also get the HIGHEST V drop..which is gonna be 
way over 500 Vdc. 
IF your line  240 vac line voltage   from the street happens to wander to the 
high side one day.... =  kablamo.   Loads of folks with AL-1500’s  have had 
that happen already. 

Ideally, the  bleeder current will be at least 15 x  greater than the leakage 
current of the caps.   The leakage current on the caps will be largely 
dependent on the  V drop across each cap.   Look up the eng specs and notes 
yourself.  Run the cap at 100% of 
its V rating, and its leakage current is high.   Run the same cap at 90% of its 
V rating..and leakage current drops like a rock.   I wouldn’t run caps at more 
than 75-80% of their rating.   2488 div by .75 =   3317.     3317 div by 450 =  
7.37     
You need a min of 7 x caps... end of story.      BTW... I have used those 3900 
uf caps.   They work damn good. 

later... Jim   VE7RF

_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
<Prev in Thread] Current Thread [Next in Thread>
  • [Amps] Sensible filter Cap sizing, Jim Thomson <=