[Amps] Toroidal filament choke and DC current

[Jim Thomson]

Date: Wed, 3 Jan 2018 14:51:23 -0500
From: Peter Bertini <radioconnection at gmail.com>
To: Amps at contesting.com
Subject: [Amps] Toroidal filament choke and DC current

<I'm building a four tube 813 amplifier and I am planning to use a 240-43
toroid core for the filament choke.

I've seen some concerns about core saturation due to the DC cathode current
involved in this application.

I was planning to add a third winding (trifiliar) to allow bringing the
filament transformer CT back to the top of the windings, thus cancelling
the DC field.  I'd have to add a second ferrite rod choke to fully decouple
the third winding, but it would be a trivial task.

I was  also going use the core to provide a bifiliar winding for
neutralization, if needed.  All windings are 7 turns, which yield about 50
uH and over 500 ohms Z at 160 meters.  Core is wound and ready to go.

Any advice, encouragement, or warnings?

Pete k1zjh

##  A looong time ago I had a hb  4 x 813 GG amp.  But it used a .5 inch
x 7.5 long ferrite rod for the bifilar.   Wound with 12 ga magnet wire to handle
the 20 A CCS  of fil current.  Then a 10 vac  at  20A  fil xfmr, with CT.   Zero issues
with rod saturation with the dc component.  Of course there will be zero issues with the
20A component, since current is always  flowing in opposite directions  due to the bifilar
design.   The magnetic fields cancel. 

##  What bands are u using it on ?    50 uh is more than ample  even for 160m. 
A buddy has a 3x6 metal tube in GG..and 24 uh used for the fil choke...zero issues
on 160m.   The 3 Amp dc current component is also a non issue.    But in this case,  2 x 
identical bifilar rods are used.  Any one winding of rod A..is in parallel with any one winding
of rod B..  This is done at the tops and bottoms  of each bifilar rod assy.    8 gauge magnet
wire used.   Uh  does not drop exactly in half either, but a little over one half..go figure.  Done
this way, since  8 gauge magnet wire will handle 40A  CCS..and  2 of em in parallel will handle 
the 80A  CCS required by  that tube. 

##  I use the identical setup on my hb  3CX-3000A7 fil  assy, except 10 ga magnet wire used. 
10 gauge will handle 30A  CCS, and 2 in parallel will handle 60A.  Tube fil  sucks  50A  CCS. 
Plate current is between  1.5 to 2.5 A.   I have also seen similar setups used on even bigger tubes,
with 5+  amps of plate current

##  IMO,  I cant see even a single  240-43 core saturating from the plate current of  4 x 813s, but dont
quote me on that, as I have no experience using a 240-43 core as a bifilar.   A buddy used type 43
torroid  cores on his hb 3CX-3000A7 fil setup, using 6 gauge magnet wire.  Dunno how big in diam the
core was, or how many used.   Dunno if it was used on 160m or not.

##  You could always use  2 x 240-43 cores in a stack, and even less than your 7 turns, to get the same uh.  
provided you have the room for the 2 x stacked cores.

Jim    VE7RF