[HBR] HBR -- Part 14

[Walt Hutchens]

Having considerably raised the gain by using a pi-wound coil for the
2nd IFT, I decided to do the same for the 1st. The complication there
is that it has to be center tapped to drive the half-lattice filter.
In the original version I had a primary wound over the secondary, but
with the pi wound coil that isn't convenient.

When switching to the pi-wound coil I used more inductance, thus
reducing the resonating capacitors from 300 to 100 mmf each.

I used the scheme from the 2x4+ Superhet design in the '65 Handbook:
a capacitive center tap with one end fed through a high impedance and
the other going to the mixer plate. I found that a 15k resistor worked
fine -- a simpler answer than the self-resonant choke in the Handbook
circuit.

The high Q coil also makes the top of the filter approximately flat,
rather than it having two peaks.

I returned the mixer cathode to -50 VDC, thus eliminating the problems
with the parasitic diode between the (hot) mixer cathode and the
filament and making up for the plate voltage lost in the 15k resistor.
There's still some hum on very strong DC carriers (like my battery
operated GDO); crude experiments indicate that this is about half on
the AGC line and the rest somewhere at the end of the IF stages, most
likely in the plate detector circuit.

The gain is now ample.  It'll copy strong signals without an antenna;
there's enough pick up in the partially shielded antenna coil.

The high gain exposes poor far out skirt selectivity but I think that
can be improved with better shielding in the crystal filter area.

The far out skirts probably won't be as good as in a 'real' HBR
design, with double conversion to a multi-transformer 85 or 100 kcs
2nd IF, but for so few tubes and such a simple crystal filter, it's
not bad.  Any of the various multi-crystal filters that are out there
or a home brew full lattice would get it on the first team.

I have figured out how to get the design back down to 8 tubes. I'm
using half of a 12AU7 for 2nd audio; the other half was to become the
3.5 Mcs band edge marker. The RF stage (currently a 6BH6) can become
another 19JN8 giving a triode for the band edge marker AND the same
pentode as in the two IF stages for RF service -- an advantage because
it'll have the identical AGC characteristics. Deleting the 12AU7 will
provide the 12 additional filament volts needed for the RF stage swap.

I'm using a 117Z6 with the two diodes in parallel for a HV rectifier
but a 117N7 rectifier diode-power pentode will easily handle the ~40
mA load AND allow more audio output.

These tubes were used in battery-or-line radios. On battery, they had
1.5 volt filament tubes in series, operated from an appropriate
battery pack that also furnished something like 90 volts for plates.
When switched to line operation the diode section of a 117P7 (or one
of the equivalents) provided the B+ while the pentode took over audio
duty from the 3Q5 (3S4, etc.) The rest of the battery tube series
filament string formed the cathode resistor of the power pentode.

The slightly bassy audio is going to get fixed soon.  It's good enough
to listen to but would be much clearer with less lows.

Walt
KJ4KV