> for further information read AN521 (DS00521C)

This is an OK app note as far as it goes BUT it is dangerously misleading
because it fails to mention the very very major problems its methods  may
cause.

BECAUSE it is an official Microchip app note and BECAUSE it mentions techo
sounding specs (eg max protection diode current) and BECAUSE it discusses
reliability and failure modes, you may be mislead into thinking that this
method will not cause problems.


The absolute truth is

- When used like this the PIC is being operated outside its data sheet specs
and if it causes problems you have only yourself (and the app note writer)
to blame.

- While the current level used here with the processor type mentioned here
MAY never cause problems

    - this is not guaranteed by Microchip (or anyone else).

    - it is certainly not the case with some other PICs.

- Protection diodes should NEVER be allowed to conduct during operation in
any design that you want to be certain will operate correctly while this is
happening OR at any time after this has happened until power has been turned
off for several minutes (or more) after it happened *.
(And even then it mightn't :-) ).
Allowing protection diodes to conduct during operation may be OK if you have
exhaustively characterised the processors operation in this condition (100s
to 1000's of hours of work) OR if you don't care if your system behaves in
strange and entirely unpredictable ways.

- If anyone tells you that the prior point is incorrect then you can be
certain that they do not know what they are talking about and that their
other technical advice may also be suspect. (Flame shields up).




        Russell McMahon

* internal nodes that usually NEVER conduct during normal operation can
place charge in locations where it is not expected and does not easily
discharge from, with the result that this "charged capacitor" causes
misoperation until it discharges. if there is no easy discharge path leakage
current is the only means of things coming right. I have seen ICs (not PICs
** ) that take several minutes to resume normal operation from this sort of
condition. This could NOT be fixed by grounding all IC pins - only waiting
for the internal discharge to finish would allow proper operation to resume.
I have seen PC I/O cards that "went out" for over 30+ minutes after power
off after certain events - this MAY have been caused by on board capacitors.

** ISD2590 and others.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads