[HBR] 160 Meter Coils

[amargosaent at iscweb.com]

Hopperdhh at aol.com wrote:
> In a message dated 12/25/2006 9:37:19 AM Eastern Standard Time, 
> Hopperdhh at aol.com writes:
>
> The oscillator would be much easier because it tunes 3.4 to 3.62 MHz.  If you 
> don't tap the coil like Ted did, you would need 99.74 uH and a padder 
> (including strays) of 56.26 pF.  Exact calculations for the tapped coil 
> arrangement 
> are not possible as far as I know.  I'm not sure what the design criteria was 
> that Ted used.  Does anyone know the answer?
You are correct, there are no calculations for the tapped coil according 
to Ted Crosby himself.  I asked that very question of him during a visit 
with him in 1967, and the essence of his response was that his 30-years 
of receiver design gave him the knowledge to estimate a starting point 
for the coil configuration, including the tap location.  Ted liked the 
Hartley oscillator because of its inherent stability, particularly when 
the LC component was biased with negative temperature coefficient 
capacitors.

Notwithstanding Ted's response, Ted's oscillator design looks most like 
that in Fig 5-13A of the 1956 ARRL Handbook, plus a couple of Ted's 
inevitable modifications (for which I have no real  explanation).  
However, the Handbook narration for the figure gives a hint with regard 
to the tap location on the coil  - /"...it is necessary to adjust 
feedback to get optimum results.  Too much feedback may cause 
"squegging" of the oscillator and generation of  several signals 
simultaneously; too little feedback will cause the output to be low.  In 
the tapped coil circuits, the feedback is increased by moving the tap 
toward the grid end of the coil."./

It being a rainy day, I chose to do a little reverse engineering to see 
if anything jumped out from Ted's specification for coil sets.  The 
cathode taps are all well within the physical bottom half of the coil 
winding in every case.  If we divide Ted's specified number of turns 
bottom to tap point and divide it by the total number of turns in the 
coil, the resulting percentages may give us a rough idea of where his 
30-year experience took him.  My results are as follows -

     80 m      6.25/18.5  = 33.78%
     40 m      4.25/13.5   = 31.48
     20 m      2.43/8.5     = 28.67
     15 m      2.125/5.5   = 38.63
     10 m      1.4375/4.5 = 31.94

     40-17m  2.4375/7.5 = 32.5

     Range is 31.48% to 38.63%      Ham band average is 32.9%, say just 
about 1/3

In mulling over these numbers, remember they allude to an overall LC 
assembly (coil end-to-end, plus fixed padder cap, plus variable tuning 
cap, plus fix neg temp comp cap) of resonance on the high-side 
fundamental frequency for 80m and 14-17m, low-side fundamental frequency 
for 40 m, and low-side 1/2 fundamental frequency for 20m, 15m, and 10m.

I suspect Ted did some rough LC resonance calculation, wound a trial 
coil, and only then used his grid-dip meter to verify he had gotten the 
band spread needed.  I saw him use the grid dipper (which he had 
painstakingly calibrated), but never caught him with pencil and paper in 
hand!  I know he rather enjoyed the mystique of being the sorcerer, but 
I alsothink he would have shared with me his "magic potion" if such 
existed - I was half his age at the time, and he seemed to view me (to 
some extent) as "the sorcerer's apprentice".

When time permits, I will have another go at the actual number of turns 
on my final coil sets (which were checked and trimmed out by Ted).  At 
the same time I will look through notes and letters to see if I have any 
nominal oscillator injection voltage numbers for the 1st mixer.  Stay 
tuned, as they say.

Hope these faint recollections help.

73 de W6HHT, Jay