[HBR] HB-67

[hbr@mailman.qth.net]

Jeff wonders:

> Has anyone built the HB-67 from the ARRL 1967 handbook? If so what
> did you think of the performance ...

Haven't built it, but have certainly stared hard at it a few times.

The beam deflection mixer has four potential advantages:  (1) Very 
high gain.   (2) Outstanding large signal handling ability.   (3) With 
the input signal on the grid and a balanced VFO feeding the 
deflection plates, isolation between the VFO and the signal is 
outstanding, meaning no trouble with pulling of the VFO.   (4) With a 
balanced mixer output circuit, you get outstanding rejection of 
signals at the IF arriving from the antenna.   

In the HB-67 design, advantages (3) and (4) are not realized because 
the VFO feed and mixer output are unbalanced.   (2) is realized in 
the mixer tube itself, but it feeds a mechanical filter -- they're *not* 
good large signal handling devices and neither an RF gain control or 
AGC on a pre-filter stage is there to help matters.   

The one remaining advantage is the high gain ... but on 75 meters 
gain's not the problem -- the ability to handle large unwanted signals 
without non-linearity (which leads to trashing nearby weaker sigs 
you want to hear) is more important.   

To this situation is added a 6BA6 RF-6D10 mixer converter to cover 
the bands above 75.   As a second mixer the 7360 has all the issues 
it had in the first mixer job, plus it's handling *even larger* signals 
because of the preceeding gain.

In my opinion this design is noteworthy mainly as an attempt to 
apply the 7360 to an amateur receiver.   It is not, however, a well 
conceived application. I believe that the later W6TC HBR designs will 
give better performance.   

For a comparaison, take a look at the 'Junior Miser's Dream' receiver, 
also in the '67 handbook.   A 7360 mixer is used without an RF stage 
on all bands although a signal-frequency Q-multiplier provides some 
gain when needed.   (The Q-multiplier makes pretty good sense -- if 
you're listening to a strong signal the Q multiplier can be backed 
down but for pulling out a weak one, higher gain for the desired signal 
with better rejection of off-frequency sigs is the ideal combination.)   
The crystal filter was used, first because a higher 1st IF is needed on 
the higher bands (3300 kcs) but also it has better large signal 
handling ability than a mechanical filter.   Only one IF stage is used --
 though a 6AU6 (sharp cutoff pentode) is a strikingly poor choice.   
Overall, the design has only a reasonable amount of gain; it should 
do much better than the HB-67 at handling conditions on the modern 
ham bands.   

Operating the VFO directly on each band probably means the JMD is 
less than SSB-stable on the highest bands, but the same was true of 
most ham receivers of the time.   

If I were going to do something like this I'd build something like the 
JMD but use the plug-in coil scheme of the HBR-series -- higher Q 
that way, and easier construction.   Or if bandswitching is desired, 
then use toroids in the front end.   Leave out the Q multiplier.  (Or 
leave room to add it later if desired.)  

You can either use the crystal filter scheme of the JMD or go with 
conventional IFTs and double conversion as in the HBR series.   Do 
balance the mixer output, don't provide too much IF gain.

(I don't know why the JMD designer didn't balance the mixer output 
since he already had an output coil in the circuit and just had to give 
it a center tap.)

Use a 6EH7 IF tube -- more gain than a 6AU6 and much better large 
signal handling.   (The AGC voltage will need to be higher -- a higher 
ratio for the transformer would do it or you could use a full wave 
bridge to rectify the AGC.)  Build a push-pull VFO -- except for 
needing a split stator tuning cap it's no harder than singled ended; a 
given level of design care will give you better stability (because the 
tube capacitance is effectively halved); and with attention to balance, 
you'll greatly reduce oscillator pulling.   

And use a later tube than the 7360.   The 6JH8 and 6ME8 (I think) 
are both improved and certainly a bunch cheaper and easier to 
locate.   The voltages and pinouts are different, but as far as I know, 
they do the same job and do it better.

Walt 
KJ4KV