Re: [avr-chat] avrdude : chip not responding

[Vincent Trouilliez]

On Wed, 2005-08-31 at 11:30 +0200, Vincent Trouilliez wrote:
> Hi again,
> 
> Just got my ATmega32 chip, fired up avrdude to try and see if the chip
> is alive, but avrdude says it's dead ("not responding", he says :-/ ).


Seeing as it said "not responding" but could somehow retrieve the chip's
ID, and seeing as I saw activity on the MISO line, I asked avrdude to
read the entire content of the flash program memory (into hex file
format).
To my surprise, it did read it, I could see a progress bar, and at the
same time I saw intense activity on the MISO line during the entire read
process. So if the chip wasn't "responding", how comes he sent all that
data ? 
however when I look at the content of the resulting 32KB hex file, I was
expecting to see "FF" everywhere (no ?), but instead I can see this
"pattern" repeated all over the hex file :

20000000FFFF01010303030307070707070707070F0F0F0F0F0F0F0F0F0F0F0F0F0F0F0FAC
:200020001F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1F1FE0
:200040003F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3FC0
:200060003F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3F3FA0
:200080007F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F80
:2000A0007F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F60
:2000C0007F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F40
:2000E0007F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F7F20
:20010000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
:20012000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFDF
:20014000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFBF
:20016000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF9F
:20018000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF7F
:2001A000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF5F
:2001C000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF3F
:2001E000FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF1F
:20020000FFFF0101FFFF0303FFFF0707FFFF0707FFFF0F0FFFFF0F0FFFFF0F0FFFFF0F0F52
:20022000FFFF1F1FFFFF1F1FFFFF1F1FFFFF1F1FFFFF1F1FFFFF1F1FFFFF1F1FFFFF1F1FDE
:20024000FFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FBE
:20026000FFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3FFFFF3F3F9E
:20028000FFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7F7E
:2002A000FFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7F5E
:2002C000FFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7F3E
:2002E000FFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7FFFFF7F7F1E


Which has me worried a bit.
Seeing as the chip was obviously "alive", I compiled the pwm demo
project detailled in the avr-libc documentation. Avrdude sent the hex
file to the AVR, but when it read it back, it found a mistmatch/error on
the first byte ! I then asked avrdude to read the entire flash memory
again, and horror, none of the my program's bytes had been written : the
content was exactly as before I programmed it (as above then), like if
nothing had been written to it !! :-O

I hope that the AVR chip I got somehow sn't faulty and got shipped with
some kind of write protection ! Or my chip is lost !! :-O :o(

Of course it could still be my cable that's a tad too long, but my prior
tests (with function generator and scope) showed it should be okay at
50KHz, and also, a long cable would introduce random errors I guess,
whereas the content of the memory clearly has a very repetitive pattern
to it.

So I don't know what to do/think... heeeelp....



Regards,


--
Vince