[HBR] Yet Another HBR Project -- Chapter 6

[hbr@mailman.qth.net]

Maybe not that hard after all ... and very interesting.

Investigating the ARC-27 20-30 Mcs IF amplifier subassembly this 
morning I think there's a good chance that the synthesizer can be 
built there.   In addition to the crystal socket and switch (24.7 ... 33.7 
Mcs) the same shaft has a pair of cams which operate a slug rack 
with total of 8 slug tuned coils in four pairs.   One pair is in the plate 
of the Butler oscillator; that circuit and coils can be used unchanged 
except for retuning the coils slightly.  

(The reason for the retuning is that in the original the crystals run 
25.7 ... 34.7; I'm adding a new 24.7 at the bottom and removing the 
34.7 at the top, in order to improve the conversion scheme.)

A second pair of coils can be used in the plate of the 1st premixer -- 
the one that adds 7.55 or 8.05 Mcs.   The other two pairs can be 
switched in the plate of the 2nd premixer to subtract either 10 or 20 
Mcs.   The total range of frequencies is 12.25 to 41.75 Mcs so it 
shouldn't be hard to rewind coils that were designed to cover 20-30 
Mcs so they'll work.   Should it?   I mean, less than an octave either 
way ...

The switch has 14 positions in a full rotation, of which ten are used, 
corresponding to the ten crystals and the full range of the slug rack 
cams.  

There are two ways to go to control the thing.   It doesn't look too 
hard to extend the switch shaft in order to put a knob on it.   There's 
a gang that generates a four-bit code that could be used to switch 
front end coils on a per-Mc basis. There may or may not be enough 
friction to avoid having to make a detent.  Because of the strange 
number of switch positions, I can't just hook it to the shaft of a 
standard rotary switch.

The other route is using the autopositioner motor.   The code 
generating gang noted above feeds back the position, the motor 
makes it turn, and a slip clutch and pawl stop it when the code fed 
back matches the order.   Going this way I'd use a normal rotary 
switch to control the RF stage 1-Mcs steps and also generate the 
four-bit code for positioning of the synthesizer.   (Venn diagrams, 
anyone?   Now *there's* a rusty mental circuit!)   Conceptually clean, 
detenting is taken care of, and allows mounting the syn. anywhere I 
want, but the mechanical stuff -- gear alignment, rigidity, etc. -- could 
be messy and the motor/autopositioner assembly is heavy and 
bulky.   Do I want the receiver to go Rrrrrr-rrrrr--rrrrr ... CLACK each 
time I change the 1-Mc step?

A possible third way is to replace the 14-position switches with 
normal 12-position ones.   This may not work because of the need to 
operate the cams but in other respects it's appealing.

The rotary switch for the 10 Mcs and 0.5 Mcs increments will be off 
chassis; this will be marked 0 -- 0.5 -- 10 -- 10.5 -- 20 -- 20.5.  I may 
have to use a reed relay or two for switching on the chassis.   There 
are 7 tube sockets; 6 are 7 pin, which may be an issue -- though 
there's always the 6J6. There may or may not be room on the 
chassis for the three extra crystals -- 7.55, 8.05, and 10 Mcs.   

Anyhow, if I'm very lucky, except for the VFO everything up to the 
output to the signal mixer is going to wind up in one assembly for 
which someone else has done most of the mechanical work, 
including the most complicated rotary bandswitch and the tuned 
circuits.   It's a couple -- okay, a *few* -- months too early to 
celebrate, but I'm encouraged.

Walt
KJ4KV