[Top] [All Lists]

[AMPS] Plate RFC choke design

To: <>
Subject: [AMPS] Plate RFC choke design
From: (Rich Measures)
Date: Fri, 27 Mar 98 05:05:48 -0800
>>Does anybody have a proven design for a plate choke?  
>>I'm building a 160-10 (no WARC) amp using a pair of 3CX800A7 tubes, (1.5A max
>>Ip).  I've seen lots of plate RFC designs that use anywhere from 80 uH 
>>(Ten Tec Titan), and on up.
>>All the homebrew designers treat winding a choke as a black art.  I've got
>>various lengths of 1" and 1.25" Delrin rod to work with, and would like to 
>>use #20 wire.  It can't be that hard, can it?
>Plate chokes are an art as I have learned.  The biggest problem is the 
>doggone resonances.  

yea, verily, unless one is trying to build a charcoal-lighter for the 

>My PA had a Hammond 1521 in it and I couldn't figure 
>out why it kept blowing up on 10m.  Well, after coupling a dip meter to 
>it, I found the choke to be series resonant at 29 MHz!  

indeed, and 'tis mo' betta to dip the hv rfc wired in place.  

>If you have room, 
>the ARRL handbook (at least my 1989 version) has  a really nice example 
>of a choke in it in one of the chapters.   It's not in the PA chapter.  
>The choke is really long and has several sections to it all with 
>different amounts of windings.  I think it is something like a foot long. 
> The different windings effectively act as inductances in series and help 
>to break up the resonances.

That's the popular, traditional, cherished, old theory, at least.  
However, orienting two coils end-to-end seemed to me like it would 
enhance coupling, not decoupling.  .  I built various hv rf chokes to see 
check the cherished theory out.   I observed that chokes wound with the 
supposed anti-resonance gaps have pretty much the same resonances as a 
same-L choke wound without gaps.  
-  Terman is right.  To minimize interaction between two coils, they need 
to be oriented at right-angles.  To maximize interaction between two 
coils, they need to be oriented end to end.  .  .  When will the ARRL 
Handbook get it's hv rfc info straightened out?  My guess is not likely 
in the next five years.  I asked an ex-staffer of QST why the org. clings 
so tightly to bum dope.  He explained that there are certain ARRL 
employees who eventually achieve guru status.  Once guru status is 
achieved, true, as well as not true, techno-utterances become as if they 
are cast in bronze.  
>A lot of literature says to pick a resonant spot somewhere that's not in 
>a ham band, like 11 MHz or 17 MHz.  These are problems for WARC bands, 
>but you aren't doing WARC.  

Maybe not.  5 or 6 percent clearance seems to be enough to avoid choke 

>After it goes into resonances, the choke will 
>then begin to look inductive again and present a larger and larger 
>impedance as you go up in frequency until you reach a peak and then you 
>start going down again.  Go high enough and you'll reach another 
>resonance and so on.  The problem is that each succeeding peak after a 
>resonance is a smaller maximum impedance.>
>I wanted to avoid any kind of resonance in the ham bands.  So I wound a 
>choke on a ceramic form from an old Johnson Viking Invader.  It's about 
>3/4" in diameter and has 6" of winding space.  It's also grooved to help 
>winding.  I used #22 AWG wire.  I made this choke to be resonant around 
>34 MHz (about 105 windings).  

a safe parking place indeed

>Then I tried it in the amp.  Good 
>peformance on 10m, BAD on 80m.  Problem:  Impedance at 80m was too low 
>(about maybe 1 KOhm).  I had another identical ceramic form and so I 
>wound a second choke with about 115 windings or so.  I put the two chokes 
>in series and tried everything out.  80 meters now worked fine (about 2.5 
>KOhms now).  I have no resonances anywhere from 3.5 to 30 MHz.
>I would have liked to put a few more windings on the second form and 
>boost the impedance up at low frequencies, but frankly, I ran out of room 
>on the form!  You probably could get by with 26 ga wire, but I wanted to 
>use 22 ga just to be safe.
In hv rf choke service, using wire with silicone-varnish insulation 
(rated at approx. 200 deg C) #22 Cu is sufficient for a tube with an 
anode current rating of 4a. (8171).  

>Your biggest problem is that you want to go all the way to 1.6 MHz.  I 
>would definitely use 2 coils in series.  You might wish to check out Rich 
>Measures' (AG6K) home page (  Regardless of your 
>views on his parasitic theories, Rich has some very good information on 
>adding 160 meters to amplifiers.  
>Personally, I would use a two choke system and try to wind the second 
>choke with a large number of windings.  Who cares if it is resonant in a 
>ham band, the first choke (resonant about 34 or 35 MHz) will still 
>function properly.  

Surely, mounting the two chokes at right-angles reduces coupling.  
However, it is still a reasonable precaution to park large-choke 
resonances in unused areas.  

>Try to get a large diameter core (>1") as the 
>inductance is related to the square of the diameter.  You'll get much 
>more inductance with 3 inches of windings on a 1.5" core than with 3" of 
>windings on a .75" core.

True, however, the larger the diameter, the greater the distributed C.  
>I hope this helps.  People might think the two choke approach is goofy, 
>but let me tell you, it works.
amen, Jon, and it's easier than bandswitching different hv rf chokes with 
vacuum relays.  


R. L. Measures, 805-386-3734, AG6K   

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>