[HBR] Good GRIEF -- Talk about a homebrew general coverage receiver!

[Walt Hutchens]

This is probably one that some of us already know about but it was new
to me when I happened to see it in the November 1970 issue of QST.  One
John E. Pitts Jr., W6BD, built what must be close to the ultimate vacuum
tube general coverage receiver in his 24 tube GCR-100A.

The receiver is triple conversion starting with a 7360 mixer.  It covers
  3.5-30 Mcs in 500 kcs bands.  The 1st conversion oscillator is crystal
controlled using 31 separate crystals.  Since 51 crystal converted bands
would be needed, it sounds like 2nd harmonics are used where possible.

The IFs are 4-3.5 Mcs (so on 80, the receiver tunes backwards and the
1st mixer is straight through), 2215 kcs (LO tunes 5715-6215 kcs), and
455 kcs.  The final conversion was done to eliminate problems with BFO
feedthrough in the IF.  There are three crystal filters in the 2215 kcs
IF, a Lamb noise silencer in the 1st (broadband) IF, a Q-multiplier in
the front end, and a Select-O-Ject circuit in the audio.

The IF tubes are 6GM6's -- the runner up to the 6EH7 as an IF stage.

The mechanical work is elaborate to say the least: The VFO is built of
milled 1/4" aluminum stock, the Eddystone 898 is geared down slightly so
the full dial range rotates the Miller 2101 LO tuning cap only 154
degrees, thus eliminating end effects, and he linearized the oscillator
so the dial reads approximately directly in kcs.  Everything is done as
subassemblies, mounted on a 13x17" chassis and connected by RG-58 as needed.

Sensitivity is claimed to be 0.1 uV for 10 dB S/N, he doesn't give
numbers for crossmodulation performance but gives anecdotal evidence
that it's extremely good.  Long term drift is 30 CPS/hour; the LO runs
continuously whether the switch is on or off.

It's hard to find any fault with this, but when did that ever stop me?
30 cps/hour drift isn't remarkably good for a very solidly-built
oscillator that runs continuously.  He used a 7587 Nuvistor pentode in a
Colpitts circuit.  I think a push-pull oscillator would have done
somewhat better -- more gain makes possible less coupling between tube
and tank.  Use of back-biased diodes (1N64's) to shift a VFO for
calibration and sideband switching and an RF choke in the cathode of the
tube are significant compromises with stability -- these components are
never stable enough for such use.

The cathode of a stable oscillator should be at the same RF potential as
the filaments because the filament-cathode interface is basically a
crummy diode shunted by a small capacitor of unstable characteristics.
The 7587 is probably better than larger tubes but the filament wires
tend to squirm inside the cathode sleeve so this cannot be an
electrically stable component.  You can somewhat 'fudge' this by biasing
the cathode negative with respect to the filament to take the diode out
of the circuit (that gets rid of the hum modulation you pick up from
having AC on a diode that carries part of the oscillator RF) but you've
still got that capacitor.

A good check on a local oscillator is to listen to a harmonic -- say
something like the 5th or so.  If you can hear line frequency FM, it
needs more work because even if you CAN'T hear anything on the
fundamental, there'll be enough phase shift to cause audible distortion
when receiving SSB signals.

The same test should be applied to the BFO and any crystal oscillators
that are involved in the conversion scheme.

However those comments are just quibbles.  This is an outstanding job of
building a fairly straightforward but elaborate design.  I wouldn't call
any of the design 'inspired' -- it was all taken from other receivers of 
the day -- but the selection of the circuits and overall execution is 
outstanding.  This is basically an exceptionally skilled home 
constructor's answer to the R-390A.

W6BD appears to have been reassigned as a vanity call.  Google finds a
John E. Pitts, Jr. but he was an Air Force General who passed away in
'77 and does not appear to ever have been stationed in California -- the
address given in the QST article.  Nor did Brig. Gen. Pitts have the
kind of career suggesting this level of electronic skill -- he was a
fighter pilot's fighter pilot from 1946 until his retirement, 30 years
later.

W6BD had plenty of radio design and construction experience, he had
machine shop access, and either plenty of money or a source of crystals
-- 30-some custom HC-6/U crystals at the time would have cost more than
week's pay for an engineer, would they not?

I guess that starting with the 80 meter band as the 1st IF it would have
been possible to use at least some crystals from the R-390 or -392
receivers.  That's a possibility I didn't think of when doing the GC-HBR
design.  In fact, by switching the 1st IF so it covered a whole Mc, you
might do most of it with those crystals.

The project took Mr. Pitts about four years -- 1965-69.  He concludes:
"Deep appreciation is given to the XYL for her patience during the
construction of the receiver.  She had only one request after it was
finished -- 'Please don't build anything else for a long time.'"  Sharyn
got a good laugh out of that!

Walt Hutchens
KJ4KV