Sherwood (who sells receiver modifications) has a web page with many
such IP3, dynamic range and phase noise tests.
Many of the early (50s and 60s) radios didn't perform well. Often
limited by the small dynamic range of multigrid converters (humongous
noise generators with little top end) with the selectivity only after a
couple frequency conversions. Some used paper bypass capacitors for the
RF and mixer stage that were well above resonance on the higher HF ham
bands. I've suggested more than once (and been told it works) that
replacing the paper capacitors with disc ceramics to lower the
inductance make a great improvement in performance.
Early synthesizers (primarily phase locked loops with programmable
dividers) tend to be outrageously noisy. One trick that Tentec uses in
the Omni V and VI is to synthesize at VHF and divide down. Good
frequency dividers reduce phase noise significantly. Radios of the
TS-850 and 870 era use one synthesizer with large (probably 10 KHz)
steps so the loop amplifier can work on phase noise out a few KHz from
the carrier, then do fine tuning with a direct digital synthesis
synthesizer that can have very good phase noise, but tends to have large
spurs in its output spectrum.
Collins (and Drake) with a single range PTO and a variable first IF were
a step forward in stability, though that wide first IF means they are
fairly easily crunched. The 75S3B in my collection when behind a 2m
converter found one ham about 20 miles away 20 or 30 times in 200 KHz
from intermod products. I replaced the 6AU8 noise generator in the
second mixer with an Amperex tube with 12 times the Gm, which lowered
the receiver noise so much I could cut the signal level into that stage
by a significant factor. With the gain of the 2m converter followed by
an attenuator, I was able to optimize the system performance so that the
strong station was only heard 3 times, once at full scale, the other at
the noise level in that 200 KHz interval, yet the system noise figure
(and thus MDS) was as good as it was with excess gain in front of the
filter.
The RF, mixer, and first IF stages of the Corsairs have the capability
of far greater dynamic range than any tubes. The Corsairs have only
those three stages before the first IF filter at 9 MHz, no roofing
filter wide enough for FM or the fine tuning synthesizer. The Corsairs
premix the PTO (except on 80 and 20 meters) with crystals to maintain
the PTO's pretty good phase noise.
I'm working on (quite slowly) a modified receiver measuring scheme where
I'll need only one high level low phase noise signal generator (I have
an HP 8640B which is what the ARRL lab uses in pairs for intermod
measurements and several crystal sources including a very low phase
noise HP 5100 synthesizer). I figure on using an old fashioned vacuum
diode noise generator to measure MDS effectively independently of
receiver bandwidth and frequency, then combine that with the signal
generator to determine the effects of strong signals on that noise
figure or MDS and I can easily do that as a function of spacing from the
desired frequency to the strong signal. It may take some correlation to
compare to ARRL measurements, but a single 8640B is much more reasonably
priced than a pair of them.
I am also working (slightly faster but not fast enough) on a 2m receiver
that promises to work very well. It uses a voltage controlled crystal
oscillator for the LO, (which is very hard to beat for phase noise),
really sturdy double gate MOSFET RF stage, diode ring mixer, and double
gate MOSFETS through the IF stages for keeping up the S/N. Just before
the product detector it has another roofing filter to narrow the pass
band of IF noise that the detector sees. I plan on building my own
filters for each end of the IF to make that tail end filter pass only
one side band to improve the S/N seen at the product detector output by
3 dB. The basic radio is available as a kit form a German QRP group,
called the Hohentweil. I can't recommend the kit yet, the manual is all
in street German (not quite the German I learned in school once), and a
large percentage of the small capacitors are not marked so it takes
having a capacitor bridge good to 1 pf with a range of up to 300 pf. I
have a Boonton Q-meter that does that well, but its not a common ham
shack item.
Ulrich Rohde has written three books on receiver design and many ham
magazine and technical magazine articles. The Corsairs use his favorite
amplifier and mixer stages. Other receivers tend to not do so well. Some
do distinctly poorer neglecting the proper termination of the first
mixer, like the Yaesu FT-736 and that kills the receiver sensitivity and
makes it easily intermoded. The earlier Yaesu FT-767 does it right and
performs a great deal better.
Then there's the ringing qualities of various filter designs. Tentecs
are fairly good, Collins and Kenwood are fairly poor.
I see I've been eclipsed. Try: http://www.sherweng.com/table.html
73, Jerry, K0CQ
--
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.
|