[600MRG] Uncalibrated S-Meter

[Edward R Cole]

Frank,

At 04:07 AM 7/18/2016, Frank Lotito wrote:
>In discussing the measurement of noise some of us have used the 
>phrase "uncalibrated S-Meter."  What is an "uncalibrated S-Meter?" 
>What makes us think that without challenge, S-Meters are unreliable, 
>while signal strength meters reading out in microvolts or dBm are 
>any more accurate?

Most ham radio s-meters are not an accurate measure of signal 
power.  But do not take my word for that - measure it.  Inject a 
signal of known level (say from a signal generator of recent 
calibration) and read what the meter indicates.  An awful lot will 
not show S9 with 50 uv carrier signal.  Then lower the signal in 6-dB 
increments noting what the meter indicates.  My FT-847 was not too 
bad but S2 to S3 was 7-dB and S7 to S8 was 5-dB so the meter was not 
even linear with power.  That's OK if handing out signal reports, I 
suppose, but hardly a reliable indicator if you are making technical 
measurements.  My K3 appears much better in both regards (I'd say its 
within 1/2 dBm and quite linear).

The difference for the Elecraft and other SDR's is that signal level 
is measured in the digital domain and displayed digitally (LCD screen 
has no magnetic movement that requires precise alignment).  Or if the 
signal is displayed on a computer screen it represents the digital 
information obtained from DSP and not a simple micro-ammeter with 
physical movement.  Note I'm not saying all receivers but most 
receivers.  Laboratory instrument meters can be calibrated and 
measure accurate ac levels (e.g. a good audio voltmeter).

>  If for example, if S9 means 50 uV at the antenna terminal of a 
> receiver having a nominal 50 ohm input impedance, wouldn't the 
> microvolt readout option indicate 50 uV?  The dBM readout option 
> indicate -73 dBm? (For those receivers where we can select the 
> signal strength unit, S-unit, dBm, or microvolt.)

If you have not tested it with a calibrated signal then the s-meter 
is uncalibrated.  If it passes both accuracy and linearity tests then 
you are good.


>Granted, the glow-in-the-dark, heavy-iron boat anchor's S-Meter in 
>all likely hood  has a signal strength metering circuit that lacks 
>not only accuracy, but also "repeatability." Maybe some of the newer 
>design solid state receives suffer the same ills?   Ditto for RF 
>spectrum analyzers, RF level meters, RF signal generators with 
>built-in output meters / step attenuators.

Depends on mfr quality and how recent the instrument was 
calibrated.  That's what the calibration sticker is all about.  Most 
hams have surplus stuff which is mostly out of date for 
calibration.  But the instrument may not be that far off and probably 
provides original linearity.  But we do not know for certain unless 
they are recertified.  Example is I have a HP432A (circa 1970's) 
which I depend on for mw level power measurements.  It could be off a 
dB or so, but I have no access to calibrate it so I just decide its 
my standard against which all my other equipment is measured.  Same 
for frequency; I have a surplus Rubidium source of 10-MHz which is 
supposed to be within +/- 5 E-11 = +/- 0.000000000005 Hz.  It might 
not be dead on but good enough for my use.


>At our individual stations, repeatability is more important than 
>accuracy. I submit even linearity is more important, e.g. each 
>S-Unit really is a step of "X" dBm? When we compare the measurement 
>results between stations, accuracy and repeatibility are both important.

See my comments , above.


>There are ways to determine a signal measuring device's "accuracy 
>and repeatibility."  Can you suggest one or more methods  of doing 
>that other than "using a calibrated signal generator that has not 
>seen the inside of a bona-file RF calibration lab" in the past 10 to 
>20 years?  I don't believe we can for accuracy.  However, within the 
>confines of our own shack, our "A" vs "B" measurements, especially 
>when we can do the comparison by flipping a switch in a fraction of 
>a second, should not be clouded by accuracy concerns.

Yes and no.  Granted our test equipment is probably not lab-accurate, 
but some items are much better than others.  The typical s-meter is 
very low in accuracy in that universe.  I trust my SDR displays and 
my surplus HP spectrum analyzer over s-meters (unless I have 
rigorously checked the s-meter against a more trustworthy instrument).

An example is measuring noise figure.  I don't own a NF meter but I 
can verify my system performance by measuring sun noise.  Of course 
that varies but one can get an accurate daily solar flux reading from 
one of the solar observatories.  That probably gets you within +/- 1 
sfu which I believe is measured in Janksies.  I don't have to do the 
math to determine my solar reception as I have sw to do that.  The 
performance figure of merit is a comparison of cold sky noise to 
solar noise called a Y-factor which is expressed in dB (and 
convertible to dBm).

But I agree that for hams its all relative; some hams are more 
exacting than others.  If I really need to get an accurate 
measurement or calibrate an instrument I can ask one of them to do 
that.  May not be cal-lab level but I'll bet its much closer than I 
can get by myself.  I take my preamps to VHF conferences which offer 
NF measuring with equipment borrowed from commercial labs that are 
certified.  Helps me know how my stuff is doing.

>
>
>73 Frank Lotito  K3DZ / WH2XHA
>
>_______________________________________________
>600MRG mailing list
>600MRG at w7ekb.com
>http://w7ekb.com/mailman/listinfo/600mrg_w7ekb.com

73, Ed - KL7UW
http://www.kl7uw.com
     "Kits made by KL7UW"
Dubus Mag business:
     dubususa at gmail.com