[HBR] HBR -- Part 13

[Walt Hutchens]

With changes in the plate and cathode resistors, the linearity of the
plate detector (and sound of the receiver) is now okay.  It probably
can be a bit better, with more study.  The acid test is listening:
I'll know a more in a few days.

There doesn't seem to be a crossmodulation problem because it sounds
just as good during peak evening hours as it does in the middle of the
night.

The HBR-type coils are impressive: The Q's are WAY up there, and that
means that off-frequency signals are attenuated before they can cause
crossmod trouble in the front end.  I did use #22 wire for the RF and
mixer coils rather than the #26 (I think?) recommended in the
instructions.  The 160M coils will have to go to the smaller size,
though.

Further revisions to the detector stage will be needed when the BFO is
wired, since there's no good place for BFO injection in the current
circuit.

It needed more gain. I figured the IFTs (remember, 100 turns of #40
wire on a 9/32" form?) were a bit Q-challenged, so that was the first
thing to be looked at. I had a pi wound coil that was self-resonant in
the BC band: By removing about 2/3 of the turns I was able to adapt
that to use with a 50 mmf shunt cap at 1700 kcs as a replacement for
the 2nd IFT. In the original, the shunt cap was 150 mmf. Higher Z =
more gain, even without the higher Q.

That gave a good boost but considering that higher bands typically
have less front end gain, even more would be useful. I'll try the
same again, with either or both of the 1st IFT (which drives the
half-lattice filter) and the 3rd IFT.  I can also optimize the mixer
cathode resistor for slightly higher gain.

For an HF receiver, you want no more gain in the RF stage than needed
to cover up mixer noise -- very little when using reasonably modern
tubes. (More of an issue if you're into using 6K8s ...) The big
problem on these bands is overloading the mixer and/or RF stage,
causing crossmodulation. Necessary total gain can be made up further
along, usually in the IF stages.

The ceramic socket I was planning to use for the oscillator coil
turned out not to quite fit the hole -- the mounting screw holes are
too far apart.  I found an Amphenol MIP socket that fits AND holds the
coil solidly.

When pulling the shield plates in the 1st IF grid and crystal filter
area I discovered a joint that had never been soldered ... no wonder
it was sometimes noisy if tapped. However, I wired the filaments of
the three new tube sockets without a single mistake ...

With the two crystals 1.5 Mcs apart (1700 kcs +/- 750 cps) the
bandwidth is just about an ideal compromise for a receiver of this
type.  It'll copy AM okay (though typical AM round tables require
occasional retuning) and isn't much too wide for SSB.

Total receiver plate current is around 40 mA.  The IF stages are about
7.5 mA plate+screen.  A few more mA will be needed when the BFO and
marker oscillator are wired up.

Walt
KJ4KV