[Amps] Choke input filters, design criteria?

[Jim Thomson]

Date: Mon, 17 Dec 2012 17:36:26 +0000
From: Chris Wilson <chris at chriswilson.tv>
To: Amps <amps at contesting.com>
Subject: [Amps] Choke input filters, design criteria?

Having built a big 5kV 1.8 amp supply that seems to work safely and
well, I want to build a lesser one for a single GS35B 2 meter amp. I
have a suitable transformer with a good range of taps to be able to
vary the output voltage through a fair range, and have acquired a 4
Henry 2 Ampere choke (not a swinging choke, a "normal" one). I am
wondering if I should attempt to use it rather than a capacitive input
filter? What design criteria can I follow to check if it's suitable,
and if so, to choose a suitable cap and bleeders to go with this? It's
from a dismantled and parted out Hirschmann 3500V DC supply. My supply
will be run off single phase, 50Hz UK 240 volt mains. Thanks.

-- 
Best Regards,
Chris Wilson.

##  After building 21 HV supplies over the years.... I would suggest   not using the choke. 
And that’s in either the normal choke input scheme..... or using the resonant choke scheme like
henry radio did. 

##  You say you have a xfmr with a good range of taps.....so why not use a lower tap....and a C input filter. 
Most of the GS35B HF and 6m amps  that I have been involved with work good with aprx 4000-4400 vdc
no load. 

##  heck, you don’t even need to build a new supply.  All you really need is to use a HV tap switch, ( or a spst vac relay)
to change the HV taps.... and route the lower B+  to the  GS35B amp.   You could also use a 2nd spst vac relay in series with the
B+  feeding the GS35B amp. Then with both vac relays  operated, the 5 kv is fed to the big amp.  With relay’s de-energized, the
4 kv is fed to the GS35B amp.   You see what I’m getting at.  Do it in such a way that it’s fail safe...and the GS35B  never gets 
5 kv. 

##  5 KV  no load will not blow up a GS35B.... not on RX anyway.   On TX, it would be over the top. 

##  Bottom line is... chokes are a pita.  The dynamic V regulation even on a resonant choke set up like henry used is nothing to
write home about.  A big plate xfmr + high C filter, with big ga wire back to main panel is the ticket these days. 

##  Lose a bleeder on any choke config..and the no load V will skyrocket to dangerous levels, creating more problems. 
Instead of the usual .9 x sec V..... it becomes 1.414 x sec V if a bleeder open up. 

##  IMO, your real NO load B+  is what you have..when drawing idle current, not what you have on RX. IE: hit the ptt,
and let the amp draw normal ZSAC.  The tube has to draw normal ZSAC b4 you can drive it.   With any big C filter config, it
will not drop down very much on ssb/cw. 

later.........  Jim  VE7RF