The set consists of a one-MPF 102 JFET beat frequency oscillator
(BFO) loosely coupled to a modified MRL 39 single-tuned crystal set.
The BFO is at left. You can see the MPF 102 JFET centered just
above the back edge of the mahogany base. The UTC LS-12 matching
transformer and RCA 'Big Cans' sound powered phones are not shown.
Also absent are parallel tuned wave traps, mainly because they
weren't used. The BFO is coupled to the MRL 39 by wrapping a
few turns of a clip lead from its output around the antenna
lead-in. The basic principle of operation is exalted carrier
reception. More about that later.

Thanks to Steve McDonald, VE7SL, for reminding me of possibilities
of this idea for the One-Active Device (1-AD) Contest. I used it
for low frequency (LF) listening a year ago -- and proceeded
immediately to forget about it. The resurrected idea has also proved
eminently successful for LF listening this season.

The original BFO a year ago was a small EICO 330 solid state signal
generator. Because of the 1-AD Contest limit of one active device,
I planned to try a simple, straightforward, one-JFET Hartley
oscillator as "proof-of-concept" and then to build from there.
As it turns out, there is little need to go beyond this most basic set.
The coil and tuning capacitor should be high Q to keep the oscillations
sharp as possible. At one time I'd thought about more "sophisticated"
FETs like the J310 and U310 but decided against these devices because
of the harmonics issue described below. Their high frequency
response might be detrimental.

Exalted Carrier Reception. Signal from a local oscillator (the BFO)
is tuned to the carrier of the desired signal. This raises the voltage
of the signal to the region of much more efficient detection. Remember
that crystal sets are notoriously inefficient on weak signals. Exalted
carrier --where the wanted signal is amplified -- is the active
counterpart to passive wave trapping -- where the un-wanted signal is
attenuated. It is amplification of sorts, but it is quite different
from regeneration or reflexing -- the set is continuously in
oscillation. In the case of an isolated weak DX signal, exalted
carrier raises the signal to the level of intelligibility (readability).
When the weak DX signal is 'buried' by a strong QRM signal, successful
extraction of the weak DX signal becomes a game of changing
the relative amplitudes of the DX and QRM signals.

There must be more high-powered explanations around,
but that's the gist of it.

Exalted carrier reception is not new. Early on, hams discovered
that weak AM signals could be greatly enhanced by tuning them in
using the single sideband (SSB) mode. Here again a local oscillator
in the receiver is tuned to zero beat with the AM carrier. These
rigs had the further advantage that you could select the sideband
(lower, upper) you want to listen to. Many of the older top-of-the-line
communications receivers, Hammarlund, National, etc., include it in
their bag of features. Exalted carrier reception differs from the
more sophisticated 'synchronous detection.' In synchronous detection
the carrier is replaced by an oscillator in the receiver, locked to
the carrier frequency and phase. While synchronous detection
reduces fading, its main benefit is increased fidelity, making it
particularly useful for listening to music programming. I'm not sure
synchronous detection could be had with only one active device.

Performance:

An indicator of the sensitivity of this set is the ring in the
phones when either BFO or MRL 39 tuning capacitor is lightly
rapped, causing the plates to vibrate. This is classic frequency
modulation about the zero beat. Equally significant is greatly
enhanced selectivity -- so much so I didn't use a wave trap.
Sure, there is splatter on the baddest actors,
but that's the station, not the set.

The strongest (but non-splattering) local stations were noted to
'pervade' the spectrum several tens of kilohertz to either side
of their center frequencies. This is because the MRL 39 set,
good as it is, has only so much selectivity. You can tell when
an off-channel local is present because its audio quality is
not affected by re-adjusting the BFO. As described below,
on-channel audio quality is greatly affected by tweaking the
BFO exactly to its frequency. Next year, I would consider
using a wave trap to attenuate the QRM from off-channel locals.

The BFO is very frequency-stable in the short term in spite
of there being no supply voltage regulation provisions.
(Voltage regulation was planned for Phase II, but I never
got that far.) 'Warm-up time' is 10 to 15 seconds. I can tune
in a station, turn off the BFO for an hour or so, then turn
the set back on to hear a high-pitched whistle descend exactly
to zero beat in a matter of seconds. One thing working for
frequency stability is the low current draw, about 0.3 mA.

Now, you don't get all this without paying a price.
Here are some of the downers:

The simple BFO does have harmonic content. The BFO is going
to exalt any strong signals at 2x, 3x, 4x, etc. the frequency
you're tuned to. For me, this means local stations at 1080,
1500 and 1540 make listening to 540, 750 and 770 kHz difficult.
2.5 MHz WWVH ties up 1250 kHz -- I also hear them on 625 kHz.
A far greater threat is all the short wave (SW) going on,
especially in the evening. Fortunately, my location is in the
SW boonies. Other localities may find the SW situation
impossible for this type set.

The spotter radio (a Sangean pocket digital) is near-useless
-- dead as a doornail. Front-end overload by the BFO signal
causes AGC clamping in the Sangean. When I turn the BFO off,
the Sangean comes back to life. But I can't listen to both
receivers simultaneously to verify identical program content.
It's not that difficult without the spotter. One can tune to
a known station and 'count channels' to either side by noting
the ascending/descending pitch squeal to the next zero beat.

The BFO, although verniered, requires some deft handling to get
exact zero beat. At the top end, this is virtually impossible
-- so I tune as close as I can and then zero-beat by judiciously
positioning my left hand near the BFO -- a hand capacitance
vernier. Because BFO tuning is so touchy, I went with the MRL 39
single-dial set instead of the two-dial Lyonodyne-17. Even when
the carrier is zero-beat the audio is degraded because the
modulation sidebands are off-frequency. Speech sounds like SSB
'duck talk,' and music can get trashed beyond recognition.
But hey, we're talking DX here, not hi-fi, easy music
listening from the Lazy Boy.

I discovered that frequency stability does suffer over the long
term as the battery supply decays. On the last night of the
1-AD Contest, I switched out of desperation to a new battery.
(The old one had been in continuous evening use for almost four
months.) With the new battery, BFO calibration points were shifted
by a small but noticeable half-a-percent or so.

Future Improvements:

Fair Radio 500 pF capacitor to replace the 180 pF var. cap.
plus 100 pF padder. May need to reduce the inductance of
the coil a little to spread the band out.

Improved vernier dial -- something larger,
perhaps even a double vernier drive.

Variable level injection -- through an actual var.
cap. instead of the gimmick.

Harmonic suppression -- low-pass filter, or tuned parallel tank
as a notch filter. At the same time I don't want to shoot myself
in the foot by attenuating high-end BCB BFO output.

In spite of all this -- or perhaps because of all this
-- it sure has been a load of fun!

-Mike Tuggle- 03/28/05