On Sun, 19 Jul 2009 12:20:40 -0700, Roy Lewallen
wrote:
Thanks for the references.
One more (in German):
http://www.s-meter.de
This one is rather interesting in that you can use the site to make an
S-meter scale.
Software to convert a perfectly good digital receiver data output,
into marginally useful S-Units:
http://www.seed-solutions.com/gregordy/Software/SMeterLite.htm
Where does the fiction come from that an "S-unit", presumably the marks
on our S-meters, is or for some reason should be, 6 dB -- that this is a
"correct" or "ideal" value?
Using S-Units is its own punishment. I couldn't find anything on the
history of the S-Meter that would substantiate my guess that long ago
some major radio company incorporated a signal meter into their
receiver. Soon after, everyone had to have one. Whether they did
anything useful is probably debatable. I guess it's not considered a
proper ham radio unless it has an S-meter. For example, the new
Elecraft K3 now has a software S-Meter for those that just can't live
without one:
http://www.sight.net/K3Meter/
My guess(tm) is that the 6dB per S-Unit was probably fairly close for
the early tube type receivers. Below S-9 was probably below the AGC
threshold. Above S-9 was more compressed above the knee in the AGC
curve. AGC thresholds in old tube radios was probably a fairly strong
signal, mostly because the RF/IF chain didn't have much gain. 6dB per
S-Unit was probably fairly close and fairly linear below the AGC
threshold.
Roll forward 90 years and we have sensitive receivers that have an AGC
knee at a few dB above the noise floor, with lots of overload handling
ability to maintain a wide AGC range, and sensitivities much better
than tube radios. If I maintain my assertion that S9 is the AGC knee,
there would be perhaps one or two S-Units below S9, and the rest of
the scale would be highly compressed "dB over S9" units. That would
be a rather strange looking S-Meter, but might also be more useful
than the current dual range meter.
I'll give cellular designers minimal credit for replacing the
inaccurate S-Meter with a less accurate 5 bar graph. Not satisfied
with the bar graph abomination, they also invented that rather
non-linear RSSI indication, and then cut it in half again by using
0-100 instead of 0-255.
I suspect that in the future, when we're all carrying FTL sub-space
communicators, Tricorders, and direct thought transmission radios, the
associated projection or heads up display will probably include an
S-Meter.
To me it's the same as "defining" pi to be
3.2, as the Indiana House of Representatives once did. "Defining" an
S-unit to be some value has no effect on our S-meters, any more than the
proposed Indiana law changed the ratio of the circumference to diameter
of a circle.
So it is written, so it must be.
Actually, I didn't think of it much until you mentioned it, but
6dB/S-Unit is rather improbable. The base line noise level for an
antique AM radio is probably about 10uv which I presume is S1. S9 is
loosely defined as 50uv. 6dB per S-unit implies a:
9*6 = 54dB
range, which is considerably larger than the:
20 log (50/10) = 14dB
range.
If the 14dB range from S1 to S9 were considered the basis for
calibration, then:
14dB / 9 S-Units = 1.5dB per S-Unit
nowhere close to 6dB per S-Unit
Or, if 6dB per S-unit, and S9=50uv are considered sacred, then 54dB
below 50uv (-72dBm) =
-72dBm -45dBm = -117dBm = 0.3uv, which is rather improbable.
However, that will work for a modern receiver, with fairly good front
end sensitivity and gain.
Roy Lewallen, W7EL
Jeff Liebermann wrote:
On Sun, 19 Jul 2009 03:00:51 -0500, "Rollie"
wrote:
www.qsl.net/k5lxp/projects/SMeter/SMeter.html
Check the chart to see actual input readings. I've always used the (6db per
S unit) as a general rule-of-thumb.
More receiver S-meter testing:
http://www.seed-solutions.com/gregordy/Amateur%20Radio/Experimentation/SMeterBlues.htm#
http://www.smeter.net/slc/signal/strengths.php
--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558