[HBR] HR-10 to HBR project

[Walt Hutchens]

> I'm working now on replacing the 6BA6 1st IF stage with another 6LF8.

Done.  And that stage too needed a shield on the tube to keep it from
oscillating.  The 6LF8 was intended for use in TV color burst circuits and
-- though that's 3.58 Mcs -- I'm thinking that perhaps this tube doesn't
have the appropriate layout and/or shielding for high gain amplifier use.
Nothing is obvious from the basing diagram, though: It has an internal
shield connected to the cathode of the pentode and G1 is between K and G2 --
both standard provisions.

It does seem to work okay with a tube shield.

The most interesting thing about rewiring this stage -- if you can call a
mistake 'interesting' -- was the S-meter pegging when I turned it on.  Five
minutes of troubleshooting led back to an unsoldered B+ connection on one of
the lug strips I redid.   The good news is that because I don't wrap wires
(on lugs) there was no connection at all, rather than one that would be okay
until a little corrosion or dirt got in there, then stop working now and
then.  

I've never understood the obsession with wrapping wires around terminals.
The builder of this set did everything but tie 'em with square knots and the
result is that I sometimes have to destroy the lug in order to get the
connections off.   (Alternative:  When there's a bit of slack in a wire I
clip it against the lug, then clip/pry the wrap off a bit at a time.)

The military did this wrap stuff routinely but in the late 1950s the Navy
did tests that showed it did NOT make for more secure connections.  In fact,
it can conceal an unsoldered joint as happened once on this set.   Of course
most of my joints get taken apart anyhow, for one reason or another, and it
makes that far tougher.   I either stick the wires through a lug with a hole
or if that won't work, do a 180 degree hook in the end of the wire.  Then
when the joint is finished a tug on each wire/lead with the needlenose
confirms that it's actually soldered.

The next phase was tidying up loose ends before changing over the AGC
circuit.  The 3rd IFT had to be switched from a conventional winding and
single cap to a tapped configuration using two caps in series in order to
provide different (and reversed phase) voltages to the audio and AGC
detectors.   Unfortunately, I got the connections backwards so the AGC and
audio outputs crossed over each other -- bad idea if you want to keep the
BFO out of the AGC.  So the IFT had to be removed and pulled apart again (NO
fun with those clips),  then the caps removed from the very tight space, and
then reinstalled in the reverse positions.

It's the little stuff that eats you on these projects.

These are actually pretty good IFTs, though.   Litz wire, pie wound, and
looks like less than critically coupled:  They should do a good job (for
this frequency) of backing up the crystal filter.

Next was the realization that the 100 mmf filters on the detector output
were much smaller than could be used, leading to unnecessarily much BFO
signal traveling around the set.   Since the BFO was still adding about 2
S-units to the signal, I replaced these with 1000 mmf each.   Bingo ... AGC
leakage down to about 1/3 S-unit and no loss of highs on voice signals.

(The opposite phase of the AGC and audio outputs from the 3rd IFT allows for
neutralization of the BFO signal getting into the AGC by adding a cap of
about 1 mmf at the appropriate place in the circuit.  This fix too comes
from the Tempo ONE.   However I have found that it isn't always required if
you are sufficiently fanatical about controlling BFO coupling to other
circuits.

(You'll notice that in many sets with product detectors the BFO is in a
separate shield box with a coaxial cable to the detector.   That seems like
more work than the approach above and it doesn't retrofit very well to the
HR-10.)

When turning up the RF gain control there was a sudden scratching followed
by a hum right before maximum.  Signals, however, seemed okay ... bad
control?  No, wait -- that was a new control.   Checking the grid voltage on
the two stages controlled by that pot showed a sudden jump to -0.6V on the
grid of the 1st IF -- the new 6LM8.   Touching various things there with a
screwdriver showed this to be a VHF oscillation, fixed by adding a 100 ohm
resistor right at the grid.

Pretty much all the modern tubes will oscillate at VHF and require this fix.
The problem may be a bit worse in the HR-10 chassis because the tubes are
spaced about 1/2" to 1" more widely than would be normal practice so leads
are on the long side.

Lots of small chores to do.   Of course small means 'easy to troubleshoot
when it doesn't work ...'

1. Connect the 1st IF to the AGC line and change the AGC filtering to allow
for faster attack when the circuit is redone.   (The AGC time constants in
the HR-10 are based on AM operation; SSB and CW require quicker action.)

2. Delete the AGC switch.

3. Relocate the power switch out of the audio gain control to the already
installed toggle switch.   At some point the audio gain control itself needs
to be replaced.  

4. Redo the BFO switch to control the cathode rather than the HV so it can
also adjust the audio gain for equal volume on AM phone signals.

5. Provide separate HV filtering for the LO so its frequency won't be
affected by AGC action.

6. Wire the 100 kcs calibrator circuit.

7. Delete the existing RF gain control line; it'll be replaced by an AGC
threshold control when the new circuit is installed.   (I rarely use this
control: The AGC handles 95% of all situations.  However it's nice to have
for very noisy band conditions.

8. Study the remaining slight hum in the audio and decide what to do about
it.   The HV filtering is minimal ...

Many of these are messy because lug strips must be replaced (to get another
terminal) or existing wiring replaced because it goes the wrong place or is
too short.   

The set now plays beautifully with the RF gain full up and the volume
adjusted with the audio gain control.   When the above items are done it
needs the new (higher gain) AGC circuit and somewhat higher overall RF/IF
gain. 

Walt 
KJ4KV