"Ed" wrote:
I suppose just a display for microvolts would suffice, though, and
that isn't hard to do at all. or just calibrate and re-paint the S meter
face to match...
________________

Just don't expect that the S-meter so "calibrated" is reading the real value
of the incident field arriving at the rx antenna. It won't be, unless that
calibration includes (exactly) the real-world performance of the receiving
antenna system at each frequency, including line loss, local reflections,
and other factors.

Otherwise the reading still will be given in fairly meaningless, relative
terms -- the same as S-units.

RF

Just don't expect that the S-meter so "calibrated" is reading the real
value of the incident field arriving at the rx antenna. It won't be,
unless that calibration includes (exactly) the real-world performance
of the receiving antenna system at each frequency, including line
loss, local reflections, and other factors.

Otherwise the reading still will be given in fairly meaningless,
relative terms -- the same as S-units.


The bottom line for the Rx, all it cares about since it doesn't know
what kind of antenna is feeding it, is the signal strength at the
input.... so I'd say a calibrated microvolt reading reflecting that
strength is not very meaningless at all. Any changes in the antenna
system will of course change that, but the whole point of any antenna
work is to maximize the signal voltage to that rx input, so I'd think a
calibrated reading would be extremely useful over an S meter alone.


Ed K7AAT

Ed wrote:

The bottom line for the Rx, all it cares about since it doesn't know
what kind of antenna is feeding it, is the signal strength at the
input.... so I'd say a calibrated microvolt reading reflecting that
strength is not very meaningless at all. Any changes in the antenna
system will of course change that, but the whole point of any antenna
work is to maximize the signal voltage to that rx input, so I'd think a
calibrated reading would be extremely useful over an S meter alone.

I'm afraid it might require more than simple calibration. The S-meter
typically just shows the AGC voltage. The AGC response is only
approximately logarithmic, and depends on the gain characteristics of
the various stages being controlled. Gain characteristics are commonly
very temperature sensitive, so any calibration scheme would have to take
that into account, as well as the common deviation from true logarithmic
response of the various stages. Calibration would also be different on
different bands, with and without preamplifier or attenuators, etc.

Of course, you could make a receiver with very nearly true logarithmic
response, by use of one of the excellent, wide dynamic range log amps
which are available these days. But however much you or I might like
one, the vast majority of amateurs couldn't care less about what their S
meter is really indicating, so they wouldn't pay the added cost for it.

On top of that, most amateurs would consider a 6dB-per-S-unit meter to
be "dead", and would rather have it wiggle more.

Roy Lewallen, W7EL

Of course, you could make a receiver with very nearly true logarithmic
response, by use of one of the excellent, wide dynamic range log amps
which are available these days. But however much you or I might like
one, the vast majority of amateurs couldn't care less about what their
S meter is really indicating, so they wouldn't pay the added cost for
it.

On top of that, most amateurs would consider a 6dB-per-S-unit meter to
be "dead", and would rather have it wiggle more.

Roy Lewallen, W7EL


So true, that last part! Well, I certainly understand the non-
linearity of the RX when comparing gain accross the entire HF spectrum.
I tend to limit my operations to top band and 75, so I hadn't considered
changes in Rx gain when moving down the band (or up, as some would have
it). Currently, when I do an Rx calibration, I just take some readings
on my "S" meter when injecting a signal in the bands of my concern. A
"list" of these readings for each band suffices, although I agree with
you that that is far more than most hams would bother with.


Ed , K7AAT

On Mon, 06 Mar 2006 13:47:04 -0800, Roy Lewallen
wrote:

[snip]


Of course, you could make a receiver with very nearly true logarithmic
response, by use of one of the excellent, wide dynamic range log amps
which are available these days. But however much you or I might like
one, the vast majority of amateurs couldn't care less about what their S
meter is really indicating, so they wouldn't pay the added cost for it.

Bill Carver, W7AAZ, designed such a beast. "A High-Performance AGC/IF
Subsystem", QST, May 1996.

I actually have one of his boards about half constructed... one I
started in 1996. Gotta get back to that someday.

On Mon, 06 Mar 2006 13:47:04 -0800, Roy Lewallen
wrote:

Ed wrote:

The bottom line for the Rx, all it cares about since it doesn't know
what kind of antenna is feeding it, is the signal strength at the
input.... so I'd say a calibrated microvolt reading reflecting that
strength is not very meaningless at all. Any changes in the antenna
system will of course change that, but the whole point of any antenna
work is to maximize the signal voltage to that rx input, so I'd think a
calibrated reading would be extremely useful over an S meter alone.

I'm afraid it might require more than simple calibration. The S-meter
typically just shows the AGC voltage. The AGC response is only
approximately logarithmic, and depends on the gain characteristics of
the various stages being controlled. Gain characteristics are commonly
very temperature sensitive, so any calibration scheme would have to take
that into account, as well as the common deviation from true logarithmic
response of the various stages. Calibration would also be different on
different bands, with and without preamplifier or attenuators, etc.

Of course, you could make a receiver with very nearly true logarithmic
response, by use of one of the excellent, wide dynamic range log amps
which are available these days. But however much you or I might like
one, the vast majority of amateurs couldn't care less about what their S
meter is really indicating, so they wouldn't pay the added cost for it.

On top of that, most amateurs would consider a 6dB-per-S-unit meter to
be "dead", and would rather have it wiggle more.

Roy Lewallen, W7EL


Back many years ago, and probably still today, many hams would turn
away from a receiver that had what they called a "scotch" S meter. To
them a receiver that read S 6 while another receiver only read S 4 on
the same signal "had to be much better". Manufacturers started making
receivers with more lively S meters.
Looking at some of the older receivers such as the Collins had much
more realistic S meters than most today.

The calibration points that Mike did on his receiver should be valid
for any band for his antenna comparisons. An actual signal strength
measurement is not required nor would it be valid between bands. All
that is really needed is the difference measurements between the two
antennas so his calibration between points on the meter scale will be
valid on any band.

A really nice instrument that would be good for signal strength
measurements is an old HP 3586C selective level meter. It covers from
around 100 hz to 32 Mhz and has a digital readout to 2 decimal places
in dbm signal strength. Hard to use with other than a steady signal
though.

73
Gary K4FMX

Gary Schafer wrote:
. . .
The calibration points that Mike did on his receiver should be valid
for any band for his antenna comparisons. An actual signal strength
measurement is not required nor would it be valid between bands. All
that is really needed is the difference measurements between the two
antennas so his calibration between points on the meter scale will be
valid on any band.

I'm not sure I fully understand this. The difference from one S meter
division to another *is* likely to be different on different bands,
since it depends on the gain-vs-voltage characteristics of the
controlled stages which can vary with frequency. But I do agree that he
can make good comparative antenna measurements without good S meter
calibration, because he has a step attenuator. By simply setting the
attenuator so he gets the same S-meter reading on both antennas, S-meter
calibration is completely irrelevant -- the antenna gain difference is
the attenuator setting.

I find it useful, however, to be able to see the difference with
reasonable accuracy just by looking at my S meter. But that does require
calibration for the band in use.

. . .

Roy Lewallen, W7EL

On Mon, 06 Mar 2006 18:43:22 -0800, Roy Lewallen
wrote:

Gary Schafer wrote:
. . .
The calibration points that Mike did on his receiver should be valid
for any band for his antenna comparisons. An actual signal strength
measurement is not required nor would it be valid between bands. All
that is really needed is the difference measurements between the two
antennas so his calibration between points on the meter scale will be
valid on any band.

I'm not sure I fully understand this. The difference from one S meter
division to another *is* likely to be different on different bands,
since it depends on the gain-vs-voltage characteristics of the
controlled stages which can vary with frequency. But I do agree that he
can make good comparative antenna measurements without good S meter
calibration, because he has a step attenuator. By simply setting the
attenuator so he gets the same S-meter reading on both antennas, S-meter
calibration is completely irrelevant -- the antenna gain difference is
the attenuator setting.

I find it useful, however, to be able to see the difference with
reasonable accuracy just by looking at my S meter. But that does require
calibration for the band in use.

. . .

Roy Lewallen, W7EL

Hi Roy,

I have been away for awhile so haven't had a chance to reply.

My point of the S meter being the same relative difference between S
units on all bands comes from the assumption that like most modern
radios, there is no AGC controlling anything in the front end of the
radio. All the gain control is done in the IF so it is impartial to
frequency.

So even if there is some difference in the gain of the first mixer, or
preamp if there is one, the gain controlled IF sees the same relative
signal level regardless of band.

I did a quick check on my old Kenwood TS430 using a Wavetek 3001
signal generator that has a step attenuator in it and got the
following results:

+60= -10dbm
+50= -20dbm
+40= -30dbm
+30= -40dbm
+20= -50dbm
+10= -60dbm
S9= -70dbm
S7= -80dbm
S4= -90dbm
S1.5= -100dbm

Test was done on 80, 20 and 10 meters with the readings the same on
all bands as close as I could determine the meter reading. I would
suspect that the signal generator leveling accuracy may be no better
than what error can be read on the S meter.

73
Gary K4FMX

"Ed" wrote:
The bottom line for the Rx, all it cares about since it doesn't know
what kind of antenna is feeding it, is the signal strength at the
input.... so I'd say a calibrated microvolt reading reflecting that
strength is not very meaningless at all.
___________

Received signal strength, whether in terms of relative S-units or the
measure of real incident fields, is at least as much the result of local
conditions as of the ERP of the transmission source on the path toward the
receiver, the frequency, and propagation conditions.

So the received signal strength indication, whether relative or "real,"
isn't a hugely significant indicator of any of these parameters.

RF