Recent comments on this topic finally started to focus on one of the weak
links in trying to establish meaningful S-meter calibration. That is, the
question of the true input Z of the RX. This is very important since S-meters
essentially measure the voltage present at the RX input.
In order to get ahandle on this I took some test equipment over to K8LX's FB
contest station (where FT1K's always abound) and Steve and I made accurate
measurements, which are summarized below. The figures show the actual measured
Z
in all positions of the input gain/attenuator switch. The figures for each
band report Z for the switch settings (respectively) NOR, 6db, 12db, and 18db.
Values did not change noticeably between IPO and NOR so we didn't bother
reporting them separately.
1.8MHZ: (37+j18) (68+j15) (108+j63) 108-j80)
3.5MHZ (32-j27) (49-j19) (36-j62) (36-j62)
7.o MHZ (60+j35) (58+j33) (37-j25) (22-j27)
14.0MHZ (17+j24) (16+j19) (18+j21) (10+j16)
21 MHZ (65+j11) (52+j30) (43+j52) (16+j78)
28MHZ (38+j32) (76+j0) (94-j58) (94-j58)
We were particularly disturbed by the 20 meter readings, thinking that maybe
that radio was screwed up in the front end. So we measured another FT1K and it
gave essentially the same readings!
These readings illustrate the futility of trying to calibrate the S-meter in
absolute units, with over 10-20 db of variation at the input terminals.
Another interesting note: I used an HP 4815A Vector Impedance meter which
'excites'
the DUT with a constant current of 12uA, which would produce 600uV across a
50 ohm load (about -50dbm). As we tested at various frequencies, when we tuned
the RX precisely to the 4815 for maximum meter reading, the S-meter readings
went from S3 or S4 to well over S9, as would be expected with the input Z
jumping around as reported.
73
Eric von Valtier K8LV
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|