[HBR] Fw: Building a HB-65.... (sort of)

[hbr@mailman.qth.net]

Hugo observes:

> ...TWO I.F. stages.

Arrrgghhhh!   That's what I get for taking a quick look at a circuit.   
Well, I was only off by 50%.   

Having looked again and counted more accurately, I don't think you'll 
have a problem if you replace the 6BA6's with 6EH7's.  These tubes 
not only have higher gain but also a much better characteristic with 
large signals -- they were about the last tubes developed for TV IF 
service and they are *good* -- as well as cheap because they were 
never much used by designers.   A 9-pin socket is required and the 
plate current will be a few mA higher.   

There are even higher gain IF tubes available but I think you won't 
need them and they're more likely to lead to stability problems.  
Should more gain be needed, the 1st mixer is the place to get it -- 
there are higher gain versions of the 6U8 as well.   But I wouldn't start 
there.   

Using toroids for the front end coils will give you more gain there -- 
because higher Q (than the standard Miller coils recommended) will 
give you more voltage step-up from the antenna.   You also get better 
rejection of unwanted signals which will make the mixer happier.   
Downside -- I'd plan for a vernier drive on that knob, as it will tune 
*very* sharply.   The only ways to beat that are to reduce the Q 
(back to Miller) or use a bandswitch to change fixed caps (and/or 
coils) and a much smaller two-gang tuning cap.

With a little more chassis space the easy way to get even more gain 
if needed would be an IF stage at 1700kcs, stuck between the two 
1700 kcs IFTs.

The one remaining issue is the 1st oscillator.   I'm pretty near certain 
that you won't be satisfied with the SSB performance unless you 
provide a buffer.   An unbuffered 5200-5800 kcs oscillator just isn't 
going to be stable enough to avoid pulling on signal peaks.   I think it 
will be okay receiving AM or CW but on SSB you'll have noticible 
distortion.   Of course SSB may not be an important design goal.

I built a somewhat similar receiver, about 11 years ago.   No RF 
stage, 12AT7 1st mixer, 12AT7 VFO/buffer, 2x12BA6 IF stages, 1/2 
12AX7+1/2 12AU7 plate detector (complicated 2-tube circuit -- 
overkill), 1/2 12AX7 1st audio, 1/2 12AU7 AGC, 35C5 2nd audio, 1/2 
12AT7 BFO.   No power transformer -- series filaments (voltages add 
to 117), halfwave rectifier from the AC line for plate voltage.   I did go 
with an IF of 5550 kcs and the VFO thus tunes 1750-1550 kcs, giving 
a receiver tuning range 3800-4000 and 7100-7300 kcs.   

Using the 5550 kcs IF you pretty much have to go with a crystal filter 
for selectivity.   I used a home made ladder crystal filter.   And 
custom crystals are required -- last time I looked I didn't see a 
standard crystal anywhere near the frequency.   $100 or so worth of 
crystals.   IF transformers, though, are smaller -- I used toroids, 
unsheilded, single tuned.   The 1550-1750 kcs VFO is a heck of a lot 
easier to stabilize than one operating 5200-5800 kcs.   So that is a 
design trade off.

The plate detector is set up to provide about 20 db gain.  That's how I 
got by with 6BA6 IF's.   Front panel mounted 2"x3" oval speaker -- a 
one-off part that I'd never use again.   I didn't provide a threshold 
setting for the AGC -- that was another mistake, as the AGC lets the 
noise come up between words.

Dial was from a command set control box tuning mechanism with 
home made face and escutchion.   Audio derived AGC as in the HB-
65 design.   The BFO -- stability critical for receiving SSB -- is a 5550 
crystal with a front panel knob to allow some tuning.   

I had a heck of a time with IF instabilty but this was my first serious 
receiver, 5550 kcs is harder than a lower frequency, and series 
filaments/line HV is a more difficult design problem.  It works fine but 
took weeks to get there!

I just pulled this receiver out, dusted it off and plugged it in.   It works 
fine, copying ECARS on 7255, solid as a rock.  Compared to the 
HBR2K, it sounds tinny but the filter is narrower and there's that tiny 
speaker.   The preselector tuning knob has an 8:1 ratio and it's none 
too slow.

Okay, so I was wondering what to do when the HBR2K project wraps 
up:  I'll take the R8040 receiver (as described above) and redo it from 
scratch.   The original was designed for minimum size (7"x7" 
chassis) and weight; if reoptimized for ease of duplication it would be 
a lot more useful.   For starters a more reasonable chassis size -- 
maybe 8"x12".  

My HBR2K pictures have been taken down; I'll try to get them put 
back up (probably have to take new ones) within a couple of days.

Walt Hutchens
KJ4KV