Problem with PIC18F2550 ICSP

[dmitri]

If we eliminate design error, which we can do as proven by others successfully building it, I can see these sources of the problem you have.

1. Loose/bad soldering causing intermittent shortscuts/clearances on the PCB
2. ICSP cable irregularities or it is too long
3. PC software conflict, e.g. other programs trying to control a COM port used for programming.
4. COM port failure.

So, here what I can suggest:

1. Thoroughly check you PCB and soldering.
2. Thoroughly check your cable
3. Check if no other program tries to control the COM port
4. Test the COM port with some other equipment, e.g. a modem.

[jonga]

The cable is aproximately 40cm long. Is it too long?

[dmitri]

The cable is aproximately 40cm long. Is it too long?

Mine is longer, about a meter.

[jonga]

So that's not it...

I'm checking the soldering but I'm pretty sure it's all fine. I was carefull enough to check each one of them after soldering. Also the com port is fine. Could it be an IC failure?

[dmitri]

It could.

Also check this thread viewtopic.php?f=3&t=67&start=20&st=0&sk=t&sd=a and especially page 3 and 4.

Can you confirm you get same results as I did when testing COM port?

My results:

Vpp checked -> pin 26 = +5
Vpp unchecked -> pin 26 = -0.5V

Clock checked -> pin 27 = +5
Clock unchecked -> pin 27 = -0.5v

Data (to PC) checked -> pin 28 = +3.6V
Data (to PC) unchecked -> pin 28 = -0.5V

You should get similar readings.

[jonga]

I read this post earlier. That's when I saw that this problem was not rare. Why do you think this is? Could it be a problem with the software? Because programming the Atmega was soooo easier.

Anyway, my readings are pretty close to yours (just a little higher):

Vpp checked -> pin 26 = +5.85
Vpp unchecked -> pin 26 = -0.58V

Clock checked -> pin 27 = +5.85
Clock unchecked -> pin 27 = -0.61v

Data (to PC) checked -> pin 28 = +3.82V
Data (to PC) unchecked -> pin 28 = -0.60V

Do you think it's worth to get the cable a little shorter?

[dmitri]

Try to write this file into your PIC and straight after that, regardless of if WinPic reported errors or not when finished writing, read the flash memory back and post the file here.

[jonga]

Ok. When I tryed to programm it, the message thrown was:

No EOF-record
Aborting LoadAndProgram because of error(s)

Then I read the PIC:

[dmitri]

Try with this file.

[jonga]

The messages I got now was:

When writing:

Programming...
Erasing ("bulk" or "chip") ...
Programming CODE, 0x000000..0x00001E
Verifying CODE, 0x000000..0x00001E
Verifying 0x000000..0x00001E
Verify Error: 000010: read 00FFFF, wanted 005156, mask 00FFFF ==> diff=00AEA9
Verify Error: 000012: read 00FFFF, wanted 006F58, mask 00FFFF ==> diff=0090A7
Verify Error: 000014: read 00FFFF, wanted 006B59, mask 00FFFF ==> diff=0094A6
Verify Error: 000016: read 00FFFF, wanted 0090D8, mask 00FFFF ==> diff=006F27
More Verify Errors, unable to list all (total=8)
Programming CONFIG, 0x300000..0x30000C
Verifying CONFIG, 0x300000..0x30000C
Verifying 0x300000..0x30000C
Verify Error: 300000: read 00FFDA, wanted 000E32, mask 00FFFF ==> diff=00F1E8
Verify Error: 300002: read 00FFFF, wanted 001E3F, mask 00FFFF ==> diff=00E1C0
Verify Error: 300004: read 00FFFF, wanted 0080FF, mask 00FFFF ==> diff=007F00
Verify Error: 300006: read 00FFFF, wanted 00FF85, mask 00FFFF ==> diff=00007A
More Verify Errors, unable to list all (total=7)
ERROR: Programming FAILED !

When reading:

Reading 0x3FFFFE - 0x3FFFFF ...
Reading 0x300000 - 0x30000D ...
Reading 0xF00000 - 0xF000FF ...
Reading 0x000000 - 0x007FFF ...
Unknown algorithm name : "16F716 ; ex: ERASE_16F87xA" !
Reading PIC ok.