Subject: Sherwood laboratory measurements.
Date: Thu, 28 Sep 2006 09:08:46 -0600
From: Rob & Terri Sherwood
To:

Two years ago I wrote Robert Sherwood of Sherwoood Engineering
to clarify some points. Will had asked me some questions and I
decided to verify my thoughts on receiver measurements. While I
was comfortable with my facts, I wanted to be certain of everything
before trying to explain technical terms to a non technical person.



Noise Floor(dBm): The signal level required to produce a S+N/N of
?10?dB

REPLY: Sensitivity is 10 dB S+N/N. Noise floor is 3 dB S+N/N.



AGC Threshold: (uV) and XdBm

REPLY:" Yes. I put in a -33 dBm (5 mV) signal, and set the audio
output to reference. Then drop the signal level until the audio
drops off 3 dB. This is the AGC threshold. Some radios have a
less stiff AGC, and the drop off may be more than 3 dB before
the AGC runs out of steam. The table lists the dB point, be it 3
dB or some larger number. Usually the S meter has just started
moving at the AGC threshold." RS



100KHz Blocking /Sensitivity (uV)Higher dB at lower uV is better?)

REPLY: These are not tied together at all. Blocking number should be
120 dB or greater. Having adequate sensitivity is has not been an
issue
since the 50s, but a reasonable nominal value would be about 1/4 uV
in
SSB mode. With any modern quality radio, galactic and man-made noise
is usually higher than the sensitivity rating of the radio.

This is easy to test. Tune to a dead place on the band on 14 MHz or
higher. Connect and disconnect the antenna.

Does the noise go up when the antenna is connected?

If so, sensitivity is more than adequate.



LO Noise Spacing (dBc) @ KHz LO noise at "X" KHZ in dBc
(Lower dB at closer KHz is better)

REPLY: 120 dB @ 10 kHz or larger.

Non-synthesized radios will be unmeasurable at 10 kHz, thus there may
be a number at a few kHz.

Phase noise is a non issue in classic radios from the 1970s and
earlier.



Front End Selectivity: Descriptor of tuning type (Tracking
Preselelctor
is best with an Octave filter being the worst for serious
communication
receivers)

REPLY: Correct.



Ultimate Filter (dB):attenuation of filter, signals stronger then
this
will leak around the filter. (Higher dB is better)

REPLY: Yes.


DNR WS (dB) @ KHz
Dynamic Range in dB at 20KHz
(Higher dB is better)
DNR NS (dB) @ KHz

REPLY: You would like 80 dB or greater.



Dynamic Range in dB at "X"KHz: X usually 2KHZ, in more modest
receivers may be 5KHz.(Higher dB at narrower spacing is better)

REPLY: At close 2 kHz spacing, on CW you would like 80 dB or higher,
while on SSB 70 dB or higher is adequate due to transmitted splatter
the adjacent signal. Passport tests at 5 kHz since shortwave signals
are spaced at 5 kHz intervals. CW pileups / contests are the worst
case scenario. Transmitted phase noise and transmitted key clicks are
problems, and even if one had a "perfect" non-synthesized radio, most
transmitters today are not clean like they were in the 1970 and
before,
so you have to contend with imperfect signals close to the weak
signal
you are trying to copy.



I just rediscovered this while trying to organize the 20G of technical
data I
have on our new home server. I thought it might help others unravel
the
techno jargon that floats around.

Robert Sherwood is one of the few beacons in the modern world of
radio.
He offers lot's of nice accessories for radios and has no beef as he
doesn't
sell radios. He also has and does some very good modifications that
will
boast the performance of any high end receiver.


Terry