I2CWRITE not writing anything on PIC18F45K80

[tumbleweed]

just saying... "Problem is definitely in arrays... but as soon their total length exceeds 24 bytes, issues start to appear. " doesn't seem to pan out.

The following reads three 16-byte arrays (48 bytes total) 20K times and verifies the arrayreads. No errors.
Same with the original two 16-byte arrays.

' device = 18F45K80

topline var byte [16]  'top part of the screen array
botline var byte [16] 'bottom part of screen
line3   var byte [16] 'more array data

ix var byte
ct var byte
cb var byte
c3 var byte

i var word
fail var word

fail = 0

for i = 0 to 20000
    ' clear arrays
    for ix = 0 to 15
        topline[ix] = 0
        botline[ix] = 0
        line3[ix]   = 0
    next ix

    ' read array data
    arraywrite topline,["ABCDEFGHIJKLMNOP"]
    arraywrite botline,["QRSTUVWXYZ[\]^_`"]
    arraywrite line3,  ["abcdefghijklmnop"]

    'read array char by char and check
    for ix = 0 to 15
        ct = topline[ix]
        if (ct-"A" <> ix) then fail = fail + 1
        cb = botline[ix]
        if (cb-"Q" <> ix) then fail = fail + 1
        c3 = line3[ix]
        if (c3-"a" <> ix) then fail = fail + 1
    next ix
next i

' check for failure
if (fail > 0) then
    i = fail
endif

Tried both PBP 3.16 and PBPL. No difference.

I just wrote another code, which simply reads values of array and writes data on 1602LCD.
Issue still exists

Post the complete code that doesn't work.

[CuriousOne]

Thanks, sure
this weekend

[CuriousOne]

The "complete" code is about 700 lines (and it is not complete yet).
Found another bug.
I call some subroutine, which uses own set of variables (verified)
after calling it, value of totally different variable (not used in that part of code) gets random values.

Very disturbing.
Can this be compiler issues or maybe faulty chip itself?

[richard]

Found another bug.
I call some subroutine, which uses own set of variables (verified)
after calling it, value of totally different variable (not used in that part of code) gets random values.


Very disturbing.
Can this be compiler issues or maybe faulty chip itself?

if you can produce a minimal, complete and verifiable example [MCVE] that demonstrates the problem it would be the first
ever case. do continue

[tumbleweed]

The "complete" code is about 700 lines (and it is not complete yet)

You just said...

I just wrote another code, which simply reads values of array and writes data on 1602LCD.
Issue still exists

That's the code that's 700 lines?

If you're having issues then continuing to add more code isn't going to help.
You should stop, and start trimming things down bit by bit until it starts working.
Then you'll know what to look at.

While compiler and chip issues are always a possibility, there's a much better chance you're just doing something wrong.

[CuriousOne]

No that different code
Now I'm back to "big" code.

I'm trying now to isolate the problem and writing all from the scratch, part by part, verifying.

[CuriousOne]

What I 100% figured out, by writing the small code is the 18F issue with ST7920 LCD.
When I turn on the graphic mode, by sending LCDOUT $FE,$2E , there are high chances that random character in text mode will be displayed in top left corner.
I tried adjusting timings and OSC settings - no change. Same code, same PCB, same LCD - no issues with 16F1939.