[HBR] Hot-rodding an HR10

[Ian Wilson]

A few months ago I acquired a somewhat stripped Heathkit HR-10. It
actually turned
out to be operational, but since I wanted to donate the lattice filter
crystals to someone
whose had been stolen from a radio on display at Dayton(!), and the HR-10 has a
number of design flaws, I decided to use this as a testbed for developing a dual
conversion receiver.

The scheme I have in mind is close to that used by G2DAF: crystal-controlled
converters down to a tunable IF of 4.5-5MHz, second IF of 455kHz with a 455kHz
SSB filter preceding the 455kHz IF amplifier strip. (My choice of 1st
IF is more a
function of the crystals that I have on hand rather than a thorough
analysis of the
spurii. I wanted to see how the idea worked in practice).

The HR-10 has a 1.7MHz IF/BFO. Converting this to 455kHz proved surprisingly
easy; the Heathkit IFTs are the standard smallish square footprint ones held in
place my a spring clip through the chassis. The final IFT should be a
'diode' type
but since I don't have one, I used a regular types. I had a 455kHz BFO coil that
was pressed into service as a Hartley to replace the original Colpitts
oscillator
(since my coil had a feedback winding rather than a tap).

Had to add a 22k resistor across one of the IFTs to tame the IF strip
but now it's
working like a charm.

The original LO was removed and replaced with a Pierce crystal oscillator (using
the same triode half of the 6EA8 as the original). This works OK at 12MHz (the
converter frequency for 40m). I haven't worked out a good switching scheme yet;
the switches ground all the unused positions, which limits what you can do.

A dual-gang variable capacitor was added. This now becomes the preselector
(with the original input bandswitched coils unchanged). The original triple-gang
tuning capacitor now becomes the tunable IF + oscillator control (2 gangs used).

There was some spare space behind the coil screening box at the rear of the
chassis. Using ugly construction on a piece of PCB (is that cheating in a tube
design?), I built the following:
  - second mixer (5840 subminiature pentode)
  - cathode follower (5718 subminiature triode)
  - tunable oscillator (5840 subminiature pentode)
The coupling from first mixer to second mixer (4.5-5MHz) is a fixed-tuned LC,
small capacitor to variable-tuned LC in the second mixer grid.
(The cathode follower is present because of the need to match down to the low-
impedance 455kHz filter. The first 455kHz amplifier stage has been changed to
common-grid to provide a similar, low-impedance input for the filter).
At present I
am bypassing the filter to get the "easy" stuff working.

Although the individual pieces worked (and I had most of this working
with separate
modules on the testbench in a previous project), I had all kinds of
trouble getting
anything through the second mixer. In the end, I changed the mixer from cathode
injection to grid injection - if it works for the HBR, it should work
for me! - and
suddenly things started coming to life. A good omen was hearing a nice glowbug
CW signal.

A second problem - buzzy and 'thin' sound when using the BFO turned out to be
due to low BFO injection. The HR-10 relies on stray capacitance to
couple the BFO
to the second IF amplifier tube. Adding a 5pF capacitor from the BFO
cathode to the
IF amplifier grid cleared up this particular problem.

Next step will be to stabilize the power supply to the LO and BFO. I
will replace the
original 6X4 rectifier and use the socket for a 0B2 VR tube. This will
also free up
0.36A of filament current. The three subminiature tubes use a total of
0.45A; the two
#47 pilot lamps (not used) took 0.3A total; result: no additional stress on the
(probably marginally spec'd) Heathkit power transformer.

Once that's done I will insert the 455kHz filter (physically in place
but bypassed)
and see what happens.

The whole BFO injection/detection area of the HR-10 is pretty minimal
(an AM diode
detector with BFO injected into the IF chain, so you can't use AGC
when the BFO is
on). However, I am not going to rebuild this, rather, use the lessons
learned on a
future receiver.

This has been a very interesting learning experience. Even though I
started with a
working rig, and had a working prototype in module form, this has taken a lot of
effort to get working. This gives me even more respect for the HBR
design, and the
number of people who have successfully built one.

73, ian K3IMW