Sounds like I should put a couple boxes on the Config page for linear fit
parameters.
Then an explanation to users on how to compute them.

I think I'll give that a go.  Also, could put the observed watts into the
log box too.
I was also thinking it might be nice to have the ability to put in "10W" as
your output power goal and WSJT-X could adjust the input level automagically
to give you that all the time.


73
Mike W9MDB


-----Original Message-----
From: John Nelson [mailto:j...@rmnjmn.demon.co.uk] 
Sent: Wednesday, August 05, 2015 3:31 PM
To: WSJT software development
Subject: Re: [wsjt-devel] Power/S-Level

Mike,

STRENGTH vs External power meter (Daiwa) (output to 50ohm load)

Smeter readings are not accurate since these are discrete dots not
continuous:

                For S1 through S9 the linear (r=0.999) relationship is:
Strength = 5.4 n - 48.2   where n is S meter.  From this, difference between
S1 and S2, for instance, is 5.4 not 6 as one might expect for Smeter reading
power.

               Strength between S9 and S9+20  =  18
                                          S9+20 and S9+40 = 18    and not
20.

In both cases  delta-Sreading is low.  

It appears that S meter is calibrated to 6dB per unit as you thought, which
is what I expect for Smeter power.

Since Strength is evidently on a log scale (dB), plot Strength vs log power:
linear (r=0.96) relationship:    Strength = 52.7 log (power) - 76.5

Now, if I change HamlibTransceiver.cpp as:

//      strength.i = pow(10,(strength.i-30-107)/10.0);
        strength.i = pow(10,(strength.i + 76.5)/52.7);

then I get a sensible answer.   There are obvious inaccuracies here, not
least the measurement of power from the Daiwa meter - all I have at the
moment.  But basically this works.  Power on your meter also changes when I
alter RFPOWER with  L RFPOWER  0.25, for instance.   How this fits in with
the service manual on the Smeter and the meaning of "Strength" returned from
hamlib is something you will no doubt have a look at.

--- John G4KLA





----------------------------------------------------------------------------
--
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


------------------------------------------------------------------------------
_______________________________________________
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Reply via email to