[AMPS] s meter calibration

[Ian White, G3SEK]

Skipp wrote:
>
>Had a quick look at the site...   Their amplification abuse page 
>says 3 dB is a one s-unit change on your receiver meter.  I was 
>under the impression receiver s-meters were calibrated 50uV 
>for S-9  and 6 dB per s-unit.  The occasional "other mfgr" using 
>100uV for s-9
>

50uV/6dB is an IARU recommendation, but it has little effect in real
life.

The idea of a single input level for S9 came long before modern
transceivers, where there is a choice of 2, 3 or even 4 different levels
of RF gain. That means the reported strength of the other guy's signal
depends on which buttons you choose to push on the RX... which if you
think about it is pretty absurd. 

Also, the idea of a constant number of dB per S-unit assumes there is a
linear relationship between dB gain in the IF/RF stages and the applied
AGC voltage (which is what the S-meter actually measures). That was true
in old tube receivers, but not with modern solid-state.

According to the lab reviews in the magazines, most modern receivers
seem to be calibrated so that the difference between S9 and S9+20 is
pretty close to 20dB. Below S9, the scale looks linear but the dB per S-
point is not! It typically takes many more dB to get from S2 to S3 than
it does to get from S8 to S0 - often less than 3dB per S-point at the
top end. 

Right down at the bottom end, it's like your car speedo - there may not
even be an S1 marking.

It doesn't have to be that way - there are engineering solutions that
could easily deliver the full IARU specification - but when everybody on
HF is "five nine" anyway, who cares any more?

73 from Ian G3SEK          Editor, 'The VHF/UHF DX Book'
                          'In Practice' columnist for RadCom (RSGB)
                           http://www.ifwtech.com/g3sek

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com