[Elecraft] Elecraft technology

[Ron D'Eau Claire]

Bill makes an excellent point. If you have access to QST archives, see
"What's Wrong With Our Present Receivers?" by Byron Goodman, W1DX. The
subtitle is "A 1957 Approach to Receiver Design". It appeared in the =
January
1957 issue of QST (Page 11).=20

One of the things in radio that I find most interesting after more than =
half
a century of pounding brass is to see if I can spot anything =
fundamentally
"new" in the art. Any basic new concepts that weren't understood before.
Since I first built a receiver in the 1940's, that hasn't happened. What =
has
happened is that we have been able to make better PARTS to build the =
designs
that go back to the first days of radio. The Superhet itself is a =
compromise
receiver. The detector connected to an antenna (coherer, crystal or
regenerative detector) is still the most "spur free" receiver we can =
make.
Armstrong evolved it into an incredibly sensitive receiver with the
"Regenerative" designs. Unfortunately, we don't have the parts to make a
really good tunable filter so Armstrong's next great breakthrough, the
superhet, is still the best. But with it comes a host of tradeoffs with
spurious responses, etc. The way to deal with some of those was with
multiple-conversion, as Bill explains. There was a time - in the 50's =
and
little beyond - when the "quality" of a receiver was in the number of
frequency conversions it used. "Triple conversion" was the ultra, until
Byron pointed out to hams the loss of other important performance one =
got to
get the "ultimate selectivity". In the years since we've been able to go
back to "single conversion" superhets because of better filters and =
better
oscillators, as Bill said.

Single Sideband was understood before World War I, when almost everyone =
on
the air was still using spark, and no "real" radio operator had even
considered 'phone. Short skip antennas, what are now called NVIS, go =
back
almost to Marconi's experiments. FAX, RTTY, phase modulation (like PSK) =
and
virtually everything else I can find today is described, at least on =
paper,
in the early days of radio. What has changed is that we have parts =
needed to
make them work today and the tools to better predict and measure their
behavior.

One thing to keep in mind is that radio design goes through "fashions" =
like
car designs and architecture. There's a couple of reasons for that: it's
what manufacturers are investing money in building parts for and what
designers know how to build. Regenerative receivers were so very popular =
in
the 20's and early 30's because they got the essential gain with few =
tubes.
Tubes were very expensive. Everyone understood the concept and the =
circuits
and so they were used for years after a Superhet design could be built a
cheaply. Indeed, I have a regenerative receiver today.=20

Today, the designs we have - including the K2 from Elecraft - is a
compromise between cost and performance based on what parts are =
available.
The real "engineering genius" in a modern rig is someone who can make =
the
best trade-offs between cost and performance. Maybe that's what =
"engineering
genius" has always been about; what value is there in designing =
something no
one can afford to buy or build?=20

What's different about radio today and radio 100 years ago when Colonel
Armstrong was dreaming up the Regenerative detector and Superheterodyne
receiver is that we have better parts to implement their designs with, =
not
fundamentally new designs.=20

Ron AC7AC


-----Original Message-----
...Back in the 70s or so, there were many tranceivers (or receivers) =
that=20
used a dual-conversion architecture. Typically this would involve=20
conversion to a broad-band IF from a fixed heterodyne oscillator, then=20
conversion through a 5-5.5 MHz LMO to a crystal filter for the IF.=20
Sometimes there was a third IF frequency, but this dual conversion =
seemed=20
typical.

The dynamic range on this architecture wasn't fabulous, but it was =
stable=20
and reliable. Since the LMO only covered 0.5 MHz, a fixed heterodyne=20
crystal was needed for each band (or 4 for the full 10m band). On some=20
rigs, the LMO was even more narrow, so more expensive crystals were=20
required.
...Bill Coleman, AA4LR, PP-ASEL        Mail: aa4lr@arrl.net