[HBR] Re: HBR -- Part 5

[Walt Hutchens]

The AGC plate detector is wired and seems to be working. That included
the two controls (back panel for AGC delay and front for AGC
bias/threshold, effectively an RF/IF gain control) and the bias supply. This
circuit has a BUNCH of parts; I'm glad it's done.

Rather than use a voltage divider on the bias output to get down to
around -50 to -70 volts for the AGC detector cathode, I hooked the
diode-connected triode bias rectifier to the filament of the same
tube, letting the filament string serve as the voltage divider. Seems
to work fine, minimizes heater-cathode voltage, and with the filaments
arranged properly, delivers the right bias.

With the detectors seeming to be okay, the next thing was winding the
3rd IFT. I have some shielded single-slug tuned coils with four base
pins that were used in the 1.5-2.0 Mcs range, so they were ideal --
but they had only one winding. The 3rd IF to detector transition
corresponds to the switch from IF signals that reference chassis
ground to audio signals referencing the common buss; by far the
easiest way to do that is with a transformer. Primary signal ground is
the chassis, secondary signal ground is the common buss: What could be
easier?

Also the secondary must be center tapped to provide signals of both
polarities so the BFO signal can be cancelled out at the grid of the
AGC detector.  Yep, Yaesu (makers of the Tempo ONE) had smart people
even in the 60's -- I'd never have figured this circuit out for
myself!

Anyway I stripped off the single winding and replaced it with 100
turns of #40, an experience I'm not looking forward to repeating for
the other two transformers. It can be done, though -- in this case,
after breaking the wire halfway through on the first two tries. Two
300 mmf caps in series give a center tap and it tunes about 1.5-1.9
Mcs in the can. It wasn't quite pure luck: I stuck one of the caps on
the original winding (looked like about 50 turns of #34 or so) and
dipped that, to guide my guesswork.

Over that winding a layer of paper and on top of that, 30 turns of #30
as a primary. Everything stuck together with clear nail polish. Should
work ...

I expect about 50 or 60 turns of #30 scramble-wound in a pile would
also have worked, although there wouldn't be much room for the primary
if it were done that way. The two windings must either be insulated
from each other or spaced because the primary is at plate voltage.

With the transformer installed and the output side of the 2nd IF stage
wired, I stuck a temporary grid resistor to ground, draped a clip lead
antenna over my GDO coil and got a 'thump' as it tuned past the IF.
A DVM attached to the AGC output showed a peak to several volts.  I
stuck one of the filter crystals in a test oscillator, draped the clip
lead antenna over that, and peaked the coil at the correct frequency.

It's silent, either with or without a carrier signal, meaning no
mistakes in grounding (that would pick up the trash between chassis
ground and the power line common) ... yet. Everything is working so
far!

Today I hope to wire up the LO and start figuring out that problem --
particularly getting it to track the dial. In periods of frustration
I'll wind the next IF transformer and do the 1st IF and S-meter circuits
-- the S-meter is driven by a bridge circuit from the cathode (voltage
drops on signals) and screen grid (voltage increases ...) of the 1st
IF stage.

That S-meter circuit is another item stolen from the Tempo ONE.  It
avoids the need for a separate S-meter amplifier tube as used by W6TC.
However I think the circuit I'm using pretty much requires a 200 or
100 uA meter and W6TC's will work with a (then cheaper) 1 mA movement.

Walt
KJ4KV