Filament choke

[wd8das]

I'm building a 160m grounded-grid amp around the GI-7B tube
(it worked so well in my SB-230 retrofit project:
http://www.wd8das.net/SB230.html    )

Anyone know of a source for a filament choke appropriate for this tube?  The filament draws only 2 amps, so the 30-amp ones I see available online are really overkill (and awfully large).  But for 160m I'm thinking that a somewhat higher inductance than might normally be used for 80-10m amps might be in order.  Say >50 uH?.  Bifilar-wound would be best, but two separate chokes would probably work too.

Should I not find a ready-made source, then I guess I could make one if I find some ferrite rod suitable for use at 1.9 MHz.

Thanks...

Steve WD8DAS

[K1JJ]

Steve,

Slip a stack of ferrite beads over both wires.  ie, Feed both fil wires thru the same beads.  No need for turns.

Get the appropriate permability beads for the 160-10M band.   This choke shud be at least 10X the filament input impedance for the lowest band.


If the amp works on the lowest band, then the choke is OK.

I've done this on seven linears here in the shack and works FB. 

Use standard bypass caps like any bifilar choke.

I posted a way to test the choke using a sig gen and a load in another thread, if you're interested.

T

[WA1GFZ]

Type 77 /73 would be good for 160 but type 43 will be fine just need more of them. Size the pair of wires for the current then find a bead that slips over the pair.  You can test it with a low lwvel signal source. Say your input Z is 100 ohms. Take the choke and put a 100 ohm resistor (noninductive) in series. Put it across a signal source and measure the RF voltage across the resistor and then across the choke. If most of the voltage is across the resistor then the choke inductance is too low. When most of the voltage appears across the choke you have enough beads. Teflon insulated wire will work the best.

[wd8das]

Tonight I ransacked my parts bins and found a 3/8" dia ferrite rod about 4 inches long.  I covered it in two layers of quality electrical tape, and wound a bunch of pairs of turns from two spools of identical enameled magnet wire about #18 or#20.  Covered about 3 inches with close turns, with a half inch open on each end.  Secured the wire at the ends with superglue, then covered it all in a sleeve of heatshrink tubing.

Each winding measured about 135 uH on my LC meter, with about 0.2 ohms of DC resistance.  I then put it across the high-Z output of my URM-25 signal generator and tuned the spectrum.  I saw no dips anywhere, and the coil seemed to load the output down about the same amount as a 2k resistor.

I also tried putting the coil in series with various resistance values, and at 1.9 MHz it appeared that the voltage division was even when the resistance was around 1.5k ohms.  The calculator says that a 135 uH coil would have a reactance of 1611 ohms, so this data all hangs together.   

I think I've got a bifilar filament choke suitable for my untuned input 160m amp.

Steve WD8DAS
 

[Opcom]

Somewhere around here is a coaxial filament choke made of 1/4" copper tubing with a #10 through it, insulated. It's in the shape of a coil about 2"D by 6"L and has many taps. Sick, I tell you!! Who would create such an abomination? Might be commercial, I don't see hammy hambone making something like it. Not sure if it is going to be used or not, saved it from a box of junk going into the recycler.

[w4bfs]

Patrick .... sounds like the input coil described by Bill Orr in his initial 3-400 or 3-500 amp .... the coil was resonated with a 3 section bc variable to both match/resonate the input and provide dc return ...gnarly ...73...john

[WA1GFZ]

I have the choke from the GPT 10K. 3/8 over 1/4  inch copper tubing air wound teflon tube insulated.

[Opcom]

3/8" ? Bet that was not easy to wind. The smaller one I have also - -likely was made on a machine. I need to find that thing and shelve it, I am sure it is in a pile of junk in the garage. The Bill Orr article - have to look that one up.