Many problems with programs that used to work

[BobEdge]

Hi,

I am having troubles recently. Programs that used to work ok are now behaving strangely, like certain lines of code being ignored, or so it seems. I may have only changed a pause time or something simple, but when compiled no longer work as expected. This started (though may be a coincidence) when I moved to windows 7 at work, PBP V2.5 did not seem to work, so I upgraded to PBP3, but still having problems. Here is an example.

goto start
pulse:
while pulseon = 1
high pulses
pauseus 100
low pulses
pauseus 100
wend
return

start:
if coil1 = 1 then
thy1 = 1
thy2 = 0
thy3 = 0
thy4 = 0
endif

if coil2 = 1 then
thy1 = 0
thy2 = 1
thy3 = 0
thy4 = 0
endif

if coil3 = 1 then
thy1 = 0
thy2 = 0
thy3 = 1
thy4 = 0
endif

if coil4 = 1 then
thy1 = 0
thy2 = 0
thy3 = 0
thy4 = 1
endif

if pulseon = 1 then gosub pulse

goto start

Now no matter which coil input is high, and when pulseon is low all the program will do is switch on thy1, and continually run the pulse subroutine.
Here is another:

if lastgun = 3 then goto jmp3   'if it was last to be used then no need to 
                                          'send settings or change relays
high coil3     ;switch on channel 3 relay
low coil1     ;swirch off all others
low coil2
low coil4
'load channel 3 settings ready to be sent to main card
let ivalue = current3
let tvalue = time3
let gvalue = gas3
let scvalue = sctime3
gosub sendmain ' send the values
pause 150
jmp3:   'jump to here if this was last gun used
let lastgun = 3 'set last gun flag

Here the first line is always ignored, all of the rest of this quite large program works fine, and these programs used to work fine.
I have tried changing chips, and PCBs so I know it cant be hardware faults, or blown chips.
Anyone have any ideas?

Regards
Bob.

[Acetronics2]

Hi, Bob

one of PBP3 main changes is the CONFIG section change ...

so, if you had shown us a FULL listing, probably we could have helped you some more ...

BTW, I do not see any initialization section nor ... is "omission" decided by you ???

Alain

[HenrikOlsson]

Whenever I see several HIGH/LOW etc in a row I suspect Read-Modify-Write issue.
Try inserting a Pause 10 between each row and see what happends. Also make sure you have analog functions (ADC and comparator) disabled on all pins since they will otherwise interfere.
If it doesn't seem to help please also let us know which chip you're using.

/Henrik.

[Acetronics2]

Hi, Henrik

I suppose you meant :

if coil1 = 1 then
thy1 = 1
thy2 = 0
thy3 = 0
thy4 = 0
endif

because there are no RMW issue when using HIGH and LOW ( as they include a TRIS command that takes some time ) ... tested "in live" with 3.5 µF loads ...

Alain

[HenrikOlsson]

I meant those too.
I know HIGH/LOW includes the clearing the TRIS-bit and that it's less likely to get a RMW-issue using it. I probably wouldn't go so far to say that RMW is non existent though since it depends on the load and, obviously, on the clock frequency. What frequency did you run at when you tried a 3.5uF capacitve load? RMW-issues also exists (with and without HIGH/LOW) when the pin(s) isn't properly configured for digital I/O.

/Henrik.

[Acetronics2]

. What frequency did you run at when you tried a 3.5uF capacitve load? /Henrik.

not more than 32 Mhz with a 16F628 and 80 Mhz with 18F chips ( NOT using LAT ! ) ... yess ! I made some hard overclocking tests ..

Alain