[HBR] Yet Another HBR Project -- Chapter 5

[hbr@mailman.qth.net]

We have a decision ...

I'm gonna try a general coverage design.   I have worked out a 
synthesis scheme using a series of crystals from the ARC-27 that 
should do the trick.  

The synthesizer:

1.  Add 7.55 or 8.05 Mcs (500 kcs steps -- that's the tuning range of 
the prototype VFO using 1U4's) to 24.7, 25.7, 26.7 ... 33.7 Mcs (1 
Mcs steps) to give 32.25 to 41.75 in 500 kcs steps.  Note that all 
desired products are above the 10 meter ham band and can be 
selected with a single bandpass filter.   The 24.7 ... crystals are 
overtone units and the ARC-27 uses a Butler oscillator; I will look at 
recycling the tuned circuit -- the oscillator, tuned circuit, crystal 
switch and socket are in one end of an assembly and it is 
conceivable that I could whack off that end.  But probably not -- there 
are other circuits similarly ganged and chances are the 'easy' way is 
to provide ten tuned circuits.   

(Slight possibility which will not be further discussed right now -- use 
the mixer and post mixer amp from this ARC-27 module too.   That 
would solve the mixer tuned circuit problems too ... although my 
problems are simpler than those of the ARC-27 because of the use of 
'sort of doubly balanced' mixers instead of conventional single ended 
ones and the narrower output tuning range.   Going this way I would 
have to wind some cam operated slug tuned coils ... gee, that ought 
to be fun.) 

The ARC-27 is a Collins set -- there's a slug rack, and jillions of 
gears drive this stuff.  Fewer cams than in the R-390 but a 
comparable number of gears, and in less space.
.
2. Subtract 20, 10, 0 Mcs.   This requires one oscillator which can be 
a pentode ECO with a choice of two tuned circuits in the plate.   This 
gives 12.25 to 41.75 Mcs in 500 kcs steps -- the sixty 500 kcs 
bands for full coverage of 0-30 Mcs.   One of three bandpass filters 
must be selected in the plate of this mixer to pick the right output 
frequency.

3.  Subtract the VFO -- 5.7 to 5.2 Mcs.   This gives 6.55 to 36.55 
Mcs which is 6.55 (the IF for which I have clock crystals to make a 
ladder filter) above 0-30 Mcs.   Doubly balanced mixer using a pair of 
6BN6's; output single ended and will be gang tuned with the receiver 
input circuits which are push-pull -- total of 5 gangs required on that 
capacitor.   

The details of the RF stage and synthesizer output tuning are 
undecided.   One way is to use a wide range tuning cap and cover 
several switch bands with one front end band -- say 1.5-3.5; 3.5-10; 
10-22; and 22-30 Mcs.   Fewer RF parts but the switching's very 
complicated; basically you have to build logic circuits through the 
two bandswitches and probably use relays to switch the coils.   And 
you get Q (hence gain) issues with such wide tuning ranges.   

The other way is individual 1 Mcs bands over the whole range.   
Small tuning cap, somewhat less switching logic (I think) but lots of 
parts -- say eight slug tuned coils and most of 30 (60 for RF stage) 
fixed capacitors per circuit.   The choice may be forced by the tuning 
cap --I'm not sure I can find a five gang wide range unit and if I do, it's 
apt to be a very large one.   (Update: no wide-range 5 gang in the 
archives.)   

Basically all this adds up to a 10-tube front end (eight for the 'local 
oscillator') with two bandswitches, a five-gang tuning cap, and a good 
double handful of coils.   It'll have to be built and tested on a separate 
chassis which can be mounted on or under the main receiver 
chassis.    

Well, maybe I can at least get the front end done this winter.   The 
VFO wasn't as hard as I thought ... maybe the same will be true of 
the synthesizer.

Walt
KJ4KV