[DXBase] Radio interface problems Win2k ( & ICOM?)

[Brendan Minish]

At 15:15 11/10/2002 -0500, Ron Stordahl wrote:
>Brendan

HI Ron I am also using 19200 Bd


>You will find a variable called PACING in the radios.ini definitions for
>these radios.  For the 756 series the value set is 5.  You can try
>increasing this, a first try might be setting it at 10.  This will slow down
>set up quite a bit, so if 10 is reliable, you could then try, say 7 or 8.


I had a play with the pacing value a few months ago and it didn't help, 
just slowed things down :-(


>The additional parm which is associated with timing are the 'eoc(nnn)''s you
>will find scattered in the code.  I think they are all eoc(100) except those
>that are eoc(0).  The ones which are 100, you may want to try 150.  Again
>this will slow set up, but might help you.


I am having a go at modifying the EOC settings (currently running at 120 
instead of 100 ) and will see if that helps.

Are the pacing and EOC values related to the com port speed, I.e will 
higher values be needed at say 1200 Bd ?

the problem with the fault is that it only happens occasionally (but more 
than enough to be annoying!) so it is hard to say if changing a setting has 
made it go away or not until I have tried it for a while

I am tempted to stick a scope on the ICOM interface and see what the 
timings look like as I change settings, Eg the space between the outgoing 
CI-V command and response from the radio

Surely DXBASE  should be smart enough to try again a couple of times before 
giving up and reporting an error (maybe it is already..)

Is DXB sending commands blind or does it wait for the response from the rig 
before sending the next command?
Blind sending would be much more prone to collisions


>It doesn't seem possible that timings would vary on one Pro II to another
>Pro II but.....

NO, I had the same problems with the original 756pro.
If there is a change in timings it is much more likely to be down to WIN2k 
vs Win ME however my problem does not appear to be CPU related

Shutting down driver X CERTAINLY dramatically reduced the frequency of the 
error

73
Brendan EI6IZ