EMRFD Message Archive 205
Message Date From Subject 205 2006-11-20 17:56:26 Rick a few thoughts on minimalist radio If you really want to do minimalist radio, you may want to step way
back and take a look at some very early history. The Pixie circuit
has many more components than an early CW station from the era
immediately after spark.
Rather than starting with the Pixie and trying to figure out what to
eliminate, maybe a better approach is to start from zero and decide
what you need. Combining transmit and receive functions is the last
thing to think about.
Starting with the receiver.... The first thing you need is wire up
in the air. The more, the better. If you have the real estate for a
full sized dipole on 80 meters, you can collect enough signal energy
to hear on a crystal set when conditions are good. I've copied CW
signals on 40 meters with just a dipole, transmatch, a 1N34 diode, a
good pair of headphones, and a one transistor Pierce oscillator
running on the bench. The leakage from the crystal oscillator picked
up by the antenna beats against the incoming signals. I didn't power
the oscillator with lemon juice, but I could have (see Bob Culter and
Wes Hayward, "Lemonized QSO" in March 1992 QST.)
Then for the transmitter, just heat-sink the Pierce oscillator and
key the connection to the load. The shift in load impedance will
offset the crystal oscillator frequency.
A dual pi-net transmatch configuration would take care of the
harmonics and allow maximum energy transfer between the antenna and
diode--but I'd analyze it to make sure the harmonic suppression is
more than legal.
So far I count 5 components for the dual Pi-Net transmatch, a 1N34
diode, 6 components for the one-transistor Pierce oscillator. A
dozen parts, plus headphones, a key, and battery--or some electrodes
to push into a lemon.
That would make contacts, but Wes and I have discussed a basic rule
for radios, which is that a station should be able to work an
identical station over a distance of a few miles. It could probably
be done with the above station, but a single transistor audio
amplifier running at maximum gain between the 1N34 and headphones
would make it possible to extract many more signals from the 80 meter
dipole. That's another 5 or 6 parts. So now I'm up to about 20.
For a more serious station, I'd probably add two more transistors and
a diode, so I could have a separate PA, a balanced mixer, and two
audio stages. The receiver would end up looking a bit like EMRFD
figure 8.7 with a PA tacked on. That would have about 35 parts, but
it would be able to work DX off the ionosphere...about the same
complexity and performance as many other variations on the theme. A
previous version of the Pixie from the 1970s was called "The
Unlike Muntz--instead of starting with someone else's circuit and
trying to eliminate parts until I had something that just barely
works, I'd start from scratch, study EMRFD (and other references too--
but in EMRFD all the circuits have been designed and tested) for
circuit ideas, and then start experimenting on the bench, one stage
and one component at a time. Since one of the joys of minimalist
radios is that they can be understood all the way down to the device
physics, I avoid ICs. (I particularly avoid cell-phone ICs, which I
designed for a number of years. It's like working in a sausage
factory--you are much happier if you don't know what's inside.)
Minimalist radio is one of the more interesting design games that we
play using the methods of EMRFD. It's cheap, it's interesting...and
as we dig in, we discover that the details can be every bit as
challenging for a radio project with 30 parts as one with 30,000.