Well, it is fixed.  See below.

Jan-Erik Soderholm wrote:

>Hi !
>I have not copied the full code, it's there in yur post for everyone to 
>see...
>
>Now, let me see if I understand here.
>You have TMR1 interrupting each 0.5 sec.
>You count those interrupts ("rollovers").
>  
>
Yes.

>At each rollover you take a *new* A/D sample and
>compare to the current "number of rollovers".
>
>Right so far ?
>  
>
No.  I actually was using the last A/D result for the comparison.  A new 
conversion was not being not started until after the comparison was 
complete.

>Now, I think one problem is that, even if the pot is left
>alone, your A/D value will "flicker" a few bits between
>each A/D reading. And since you are making a new
>A/D reading each time, you might "miss" the "hit".
>I'd only make a new A/D reading right after the
>current reading have had a match with the rollover
>counter.
>
>  
>

Almost.  My rollover counter was incrementing before I got my first A/D 
result causing my 'equal to' comparison to fail.  I just changed it to 
get a conversion first..  Actually, the real fix should be to change the 
comparison to >= instead of just ==.  If I'd do what you suggest pot 
changes won't take immediate effect, but rather just after a match is made.

>Have you made a quick test-case to test that you actualy
>can read the A/D line at all ?
>
>  
>
Well, that was my quick test case...

>A few other things...
>
>  
>
>>;       GP1 - Not used.  (Seems to be blown on this particular chip.)
>>    
>>
>
>Hm, are you sure that GP0 works ???
>  
>
At the time I wrote that, no.
Now, I do know it works.  :)

>  
>
>>;       GP5 - Drives an LED on a clock tick.  Visual 
>>indication that program is working.
>>    
>>
>
>Well, at least that the program is *running* and the *LED*
>is working... :-
>  
>
Well when the code is debuged, that comment will be true...

>One often hears that one should put as little as possible in the ISR
>and do the rest in main. But in *this* case, since your main routine
>doesn't do anything else then wait for the ISR to set the flag,
>and there is no other interrupt sources, you could just as well
>move more of the code into the ISR. That would simplify the
>structure a little. I don't know if it solves your problem...
>
>  
>
Yes, might simplify a little now, but other things are likely to be 
added later.  :)

>I'd make a test-case to verify if the A/D module is working at all.
>
>Oh, I nearly forgot...
>Where do you wait for the A/D conversion to finish ?
>Or are you just hoping that it will be finished in the next
>0.5 sec (it probably will, but I think one should check the
>GO flag anyway...).
>
>  
>
It should be finished, but I'll add the wait.

>And I'll also make sure that I just a chip without any "blown" pins !
>It's hard to know what else might been broke inside...
>
>  
>
I only have this one '675.  Need to get another one on the way...
Actually GP1 is working now.  CMCOM was not initialized.  (See Jinx's post.)

>Jan-Erik.
>
>
>
>  
>
Aaron

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist