Koellner, Guenter schrieb:
>Tobias,
>
>why do you always take the old value into your calculation. It would be much,
>much simpler when we only look for absolute values, but I see, my thoughts are
>not accepted here, and so I will implement it as working, even when I think it
>is an not optimum solution.
>
>73, Günter
>
>PS: Keep in mind: English is not my mother tounge.
>
>
Hallo Günter !
The algorithm can as well be used without knowledge of the last
frequency, provided that the current LO is exactly on a step-frequency
(as it usually is -or should be- after initialisation is done). At large
differences e.g. using the MHz -wheel you just can proceed without
knowledge of the last LO. The more interesting part is with low
differences (from an exact step). At a 25000 Hz stepwidth the user
cannot change the frequency for more than 10000 Hz by just one click at
the LO-Numbers in either direction which is less than half the
"distance" so you can assume what the last frequency could have been.
Knowing this you can assume, that at any frequencies less than one
stepwidth apart the user has tried to move the LO into that direction
and move the LO for one stepwidth in that direction.
This version works on the same principles, but more assumptions have to
be made. (you have to calculate the "nearest" step) This version will
not work when the step-width is exactly two times of any wheel-turn. (at
a stepwidth of 20000 there is an ambiguitiy when a 10000 Hz change has
happened).
> PS: Keep in mind: English is not my mother tongue.
Die ganze Sache funktioniert ähnlich der vorgeschlagenen, nur muß man
die Ausgangs-Frequenz als genau auf dem Frequenzraster annehmen (nach
Init !). Dann können Änderungen als kleine Abweichungen vom Raster
angesehen werden und es wird wieder entsprechend dem Vorschlag
gerechnet, also immer mindestens einen ganzen (Frequenz-) Schritt zur
Seite gehen. WinRad übergibt nach jedem einzelnen Click auf den LO den
neuen Wert sofort an die DLL, die kann das entsprechend auswerten. Ist
diese Version wirklich einfacher ?
73 de dg3lv Tobias
_______________________________________________
Winrad mailing list
[email protected]
http://winrad.org/mailman/listinfo/winrad_winrad.org
