[HBR] HR-10 to HBR project

[Walt Hutchens]

> Lots of small chores to do.

But before tackling them I discovered yet ANOTHER low level VHF oscillation,
this time in the 2nd IF.  This stage had a parasitic suppressor in the grid
but evidently that wasn't enough.  A 100 ohm resistor did the trick as
usual.

Connecting the 1st IF to the AGC led to an IF oscillation.   OOPS!  I had
used a metal film capacitor for AGC filtering and as I discovered on the
'Long SLOW ...' project, those have too much inductance for MF/HF bypass
service.  Replacing that with a disk ceramic eliminated the problem AND
further reduced the effect of the BFO on the S-meter reading -- now only
about a needle width.  That's tolerable, I think, so probably no need to
neutralize.

Other than having to run a wire from the S-meter all the way to the S-meter
zero pot on the back of the set, getting rid of the AGC switch was easy.
The AGC is now full time.

It's now possible to listen to the set.   The audio has just a bit too much
low frequency so that's a change in one or two capacitors.   It is also
slightly distorted: This is probably caused by my (rather casual) temporary
AGC circuit so I won't look into it until the permanent circuit is in place,
perhaps a week from now.

Transferring the power switch function of the volume control to the new
toggle switch was just the usual run-two-wires-in-an-awkward-place exercise.
The audio gain control may actually be okay -- the bad noises only happen at
very low settings and then only when you're actually turning it.

I replaced the BFO switch with a SPDT unit, grounding the center pole and
connecting the pole that's ON when up to the BFO cathode.   The other pole
goes to the feedback circuit in the audio amp where it shorts part of the
1st audio cathode resistor to raise the voltage gain from 10 (input grid to
voice coil winding) to 20 when the BFO is OFF.   The goal here is when you
zero beat an AM signal and switch off the BFO, the volume stays about the
same and this guess turned out to be roughly right.

I've been a remarkable long time coming up with this trick and will copy it
to the Long, SLOW project in the near future.

Giving the LO its own separate HV filter coming off the rectifier cathode
made no noticeable difference but since there's not enough AGC to cause much
change in the B+ load, that's not surprising.   When the AGC-controlled load
current (RF and two IF stages) is cut from ~45 mA to 5 mA on strong signals,
it'll be useful as that's a difference of 60 volts in the B+.

The filtering of the main HV in the HR-10 is just a 1.5k 10W resistor
between 20 mfd caps: There is no filter choke.   I could add the choke, but
the goal here is to keep things simple and require as few added parts as
possible.

Wiring the calibration oscillator was uneventful.   Well, except for a day
devoted to some kind of nasty stomach bug followed by Thanksgiving company.
I used the circuit of the Heath plug in accessory calibrator, HA-10-1 but
replaced the 6BA6 with a 6BH6 to save filament current.   I had to add 10
mmf in parallel with the 15 mmf trimmer I used across the crystal to get it
to zero beat WWV at 15 mcs but that's to be expected.   It's nice when
something works the first time.

Something about copying a well tested circuit ...

The first thing I discovered when I began looking at the full-volume hum was
that there was a loose connection somewhere in the HV filtering.  Gee ... it
took no time to notice that the filament winding ground shouldn't be via the
twist lug on the filter cap; wouldn't you think I'd also spot that the cap
itself wasn't well grounded simply by being twisted in place?   Wiring the
capacitor lug to a real ground got rid of the loose connection noise.

This appears to be a Heath design error and it would be surprising if it
didn't occur on some of the other kits.

The hum audible at full volume was coming from the detector stage.   I had
transferred the extra filtered B+ from the 1st audio (now inside a feedback
loop that takes care of a bit of hum) to the detector stage but completely
forgot that the BFO also needs pure DC.   The complication is that the BFO
is switched and draws about twice the place current as the detector.   Since
the detector operating point is somewhat critical the two do not share a
poorly regulated HV source very nicely.

There are two ways to go with this: Another separate RC filter for the BFO,
OR switching to a choke input filter with a second choke replacing the 1.5k
resistor.   I think either would be satisfactory: Adding another filter is
the quickest and cheapest but using filter chokes would greatly improve the
HV regulation (good for AGC action) and lower the power consumption
slightly.  There is enough space for a couple of the command set chokes.
For now, however, I'm going with quick and cheap!

On to increased gain/better AGC!

Walt 
KJ4KV