> Hello Cristiano,
>=20
....
> In my case, Texas doesn't guarantee if the flash lock bits would prevent =
flash
> corruption.
> The recommended solution is to add a voltage supervisory IC to protect yo=
ur
> microcontroller from these external events.
>=20
> In order to acknowledge if you are having this problem and not some USB
> related issue, I suggest you to pick a good board and read the entire fla=
sh
> using your programmer, after that, pick a bad board and repeat the proces=
s.
> Compare the results and you will have the exact location where the
> corruption occurred.

The corruption may not occur in the same place every time, it could be rela=
ted to just which (programming) page of code is being accessed at the time,=
 or it could be totally random.

One of the problems is that Cristiano says: -
" The power pins of the pic are directly connected to the power pin of the =
usb connector. (there is a bypass capacitor)"
Which to me is a no-no. Even with a number of ceramic capacitors for high f=
requency bypassing and tantalum (or large value ceramic) this asking for al=
l sorts of problems, as the micro will be operating at some unknown voltage=
 which could be below the minimum operating voltage for the chip. USB power=
 is notoriously NOT the nominal 5V and I wouldn't connect a micro like this=
 without having it on a voltage regulator. I suspect this is the base probl=
em here, and the Flash memory is operating in a marginal state that sometim=
es lets it partially erase over time until it becomes unreliable and it nee=
ds reprogramming.

Which reminds me, is this chip a 5V one or 3.3V one? If it is a 3.3V one th=
en you cannot connect it straight to the USB supply as that voltage will be=
 too high.

--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.