[HBR] 40m osc coil

[Walt Hutchens]

Ian said:

> ... the tuned winding is ugly and misspaced. This appears to be due
> to the Cu wire I used being too springy. It has lots of memory for
> kinks, etc. I plan to remove the winding and try rewinding it.
> However, I am not clear how a 6.5-turn winding occupying 3/4" can
> retain even spacing without some help - and it can't be anchored
> until the calibration is done. Hmm.

First, straighten the wire by sliding it across something like a
toothbrush handle under hand tension and with a bend at the handle.
This will put a uniform 'kink' in the wire, end to end. You'll find
you can get the wire into one long gentle curve and when you rewind
the coil you'll have no problem.

This may sound harder than it really is. Once you try it, it'll be
second nature when winding coils.

Rewind the coil as tightly as you can and with the turns uniformly
spaced since that's the configuration that will take the minimum of
wire and will thus tend to hold its place.

I use narrow strips of masking tape to hold turns when doing an
initial adjustment. When you have it as close as you can get it that
way, use narrow strips of clear nail polish to hold all but one turn.
Let this dry for a couple of days as the drying will significantly
affect calibration.

(The solvent is a dielectric so it increases the distributed
capacitance. Since the nail polish itself is also a dielectric the
final result will also be low in frequency -- but not as much as when
the polish was wet.)

When you think it's dry, try the frequency a couple of times, say
morning and evening: If they're practically the same you're good to
go.

Finally, adjust the one unglued turn and fasten it with JUST DABS of
clear nail polish.

> I would like to fit the new coil to the old scale. I realize that
> this may be too difficult, though.

You should be VERY close if the original builder followed the same
directions you're using for the coil and you have the original tuning
cap. The difference will be due to different distributed capacitance
-- that nail polish, the type of coil form (polystyrene, Bakelite, or
? Since you are restoring a receiver you have the original wiring.

Certainly if the original builder's technique for the 20M coils you
have was okay you should copy it as closely as possible since he
probably did the same on 40.

Hopefully you haven't had to change any caps in the oscillator circuit
since all of them will affect the calibration in ways that can't be
perfectly corrected with the coil trimmer.

I assume you're adjusting the coil inductance at the low end of the
band and the trimmer at the high end. Because there's so much
similarity to the original, you'll probably be very close. To get
(almost) spot on, you'll need a three (rather than just a two) point
fit to the dial scale.

There are many ways to get a three-point fit. The one that's built
into the HBR 11 and 13 circuit (and the others, if I remember
correctly) is the fact that the coil is tapped for the trimmer cap. If
you move that tap and readjust everything for a fit at the same two
points, you'll find the middle of the scale has moved one way or the
other.

That is, the new tuning curve passes through the same low and high
points but is a different shape.

Re-tapping a coil is a nuisance but you can effectively move the tap
by squeezing turns on one side and spreading them on the other. Of
course to do this the turns must not be glued down. Wind tightly and
use a couple of strips of tape. Given the fact that you're not
changing much at all this method may give you enough range to get the
middle right -- or it may not.

When doing a three point fit you have the problem of the two point,
namely that the adjustments interact -- but there are three of
them!

Three point fits are challenge but not beyond any methodical worker.
Do it when you won't be distracted and make notes on your adjustments
and the results. Do the ends, check the middle. Change whatever you're
adjusting to move the middle, then set the ends again and recheck the
middle ...

If you dig through the old literature you'll find mathematical methods
but they are of little practical use because they apply to specific
circuits and things like distributed capacitance and inductance are
unknown.

I've done three point fits -- one was for an HBR-takeoff (different
circuit, same concept, including plug in coils) for which I had a
nicely calibrated Eddystone 898 dial from an unknown receiver. Same
deal for another that was built around major parts of a Heath HR-10
but with plug in coils and a new circuit -- I wanted to keep the dial
calibration.

Even more interesting was the high band on a navy SRR-13 receiver
which was designed as 16-28 Mcs. but due to a last minute spec change
was calibrated as 16-32 Mcs on the dial, just by adjusting the coil
and parallel trimmer. Needless to say the middle of the band was off
by a bunch. The circuit allowed a three point fit by changing a fixed
capacitor and I did that experimentally. RCA ... I was not surprised
when they were early casualties of the radio supplier wars.

It just takes time.  

Walt
KJ4KV