[Amps] 2 x 3-500Z filament choke + 30/60/80A fil chokes

[Jim Thomson]

Date: Fri, 2 Dec 2011 09:07:32 -0000
From: "Iain" <iain at g4dpf.co.uk>
Subject: [Amps] 2 x 3-500Z filament choke

Hi All

 

I am gathering the parts to build a two valve 3-500Z linear amp.

All the suggestions for the filament choke seem to be to purchase from RF
parts or surplus sales in the USA.

I haven't been able to find any info on winding a homebrew one.

Is there something really special about the core or will any ferrite rod do
?

Any info on wire size and the number of turns would be very useful

 

Thanks Iain G4DPF

##  I bought just the 8"  long x 1/2"  diam rods [ type 43]  from SSON. They were
wound with 10 ga polyimide magnet wire.  SSON also had the plastic, 1/2" ID
clamp down nylon ends, that slide over the ends of the rod..and pinch it tight.
The nylon clamps ate then machine screwed to a stand off.  10 ga magnet wire is
 required to handle the 30A CCS of current.  Watch out, even with solid connections,
you will still get a V drop  across em. 

## Those rods are brittle.   The 10 ga solid magnet wire is stiff, and tough to work with.
The procedure is to BIFILAR wind both windings at the same time...onto some 1/2" OD 
Aluminum tubing, with the AL tubing firmly clamped into the vise.  Use  a 12-16" long
AL tube.  As you get 1/2 to 3/4 of a turn onto the tubing, simply grab the starting end, and rotate
it 180 degs, then wind another 1/2 turn onto the rod, then rotate the windings  backwards 
another 180 deg, etc.  Leave loads of slack/pigtails  at both ends !! 

## Using 8" long AL tubes [1/2" OD],  slide the finished windings  off the 12-16" tube,
and on to the temp 8"  AL tube,  complete  with nylon clamps + standoffs on each end.
Then mount the standoffs  into the underside of the RF deck.  OK, then form each of the
4 x wires very carefully.  Then cut off all excess wire, and also allow enough length  to 
install the crimped lugs, or machine screw box lugs, or solder em to the socket at socket
end, and terminate the 'cold' ends to some type of cross connect terminal strip.  Of course,
the low V wires from the fil xfmr  will also have to terminate to the same terminal strip. 
You will also have to install a .01uf @ 1 kv  disc ceramic cap   from each terminal on the same strip,
to chassis. [ 2 x  caps required, one per leg].    The caps will kill any residual RF that gets past the
chokes. 

##  Ok, now  you can test the tubes out, and check for V drops in all the wiring.  When that's  done,
remove the entire mess, and  remove the AL tubes....and  THEN replace with the 8"  long  ferrite rods.
Then re-install the assy.   The reason the AL temp tubes are used is so there is NO chance of breaking 
the brittle rod while  forming in the stiff 10 ga wires, etc.  IF using lugs, or crimp lugs at the terminal strip,
they have to be installed such that  you align them properly  [rotated into proper position]. You don't want
to have to twist the lug, in order for the hole ion the lug to line up with the machine screw in the terminal
block.  It should all just drop in. 

##  Mine ended up being aprx 52uh  per winding.  Forget how many turns  but 10 ga magnet wire is aprx 
.108"   Those nylon clamps on each end will eat up exactly 1/2" .... so the useable portion of the rod is
7".   When I measured the SSON type 43 rods, they were exactly  7.875" long  [ 7 + 7/8"] 
7" / .108" = 64 turns of wire required, [ 32 PER  each winding]   .5" x  3.14" =  1.57"  per winding. 
1.57" x 64 turns = 100" wire required in total [ 50" per winding]   Then also allow 6-10"  for each of the
4 x pigtails so it's really  124-140" wire required. 

##  You can also use 1/2" wood dowelling for the former, etc,  but make damn sure the 1/2"  dowelling is 
actually  .5"  OD..and not  < .5"    To be safe, measure the OD of the Ferrite rod, and also any AL/wood
former you use.  My completed BIfilar winding simply slid off the 14" AL tube...and onto the 8" AL tube
[both from the same scrap piece of 1/2" AL tubing, some scrap ele tip off an old wilson yagi]. 
The assy also slid onto the ferrite rod easily.   You don't require the expensive polyimide 10 ga magnet
wire either [15 kv rated expensive stuff], any brand of magnet wire will do.   Check for shorts between windings,
nicks, uh [ if u have a LCR meter], before you slam a 30A load on it. 

## I have made several of these BIFILAR fil chokes, using the above method.  We have also done the same choke,
but using bigger 8 ga magnet wire, [ .134"]  in which case the inductance will drop down to  aprx  30 uh per
winding.   

##  for folks who require a 60A or 80A fil choke,  you parallel 2 x of the above choke assy's.   For a 3CX-3000A7
fil [ 50A],  2  of the 10 ga Bifilar choke assy's  are built.  Then any one winding of choke 'A'   is paralleled with
any one winding of choke 'B' ..and vice versa.  This ensures opposing  current flow in each rod assy = zero rod
saturation.  With opposing current flow [ just like a regular bifilar], the 60 hz fields in each winding will cancel out.
Saturation is not caused by drive RF.    For a 3CX-6000A7 [79A], 2  of the 8 ga Bifilar assy's are built.    They are
paralleled  in the same fashion as the  10 ga assy.  When the windings are paralleled in this fashion the uh does
not drop in half, like you would think.  The 10ga assy is  38 uh....and the  8 ga assy is 22 uh.   Both have been used
from 160-10m.  1 kw of drive  will not put a dent on them. 

##  The RF parts 30A choke has real short pigtails, and is made for PC board mount.   I have no clue how long they
are, nor  what type of ferrite was used, or what the uh is per winding. 

Hope this all helps. 

later....... Jim  VE7RF