Mathias Kuester wrote:
:
what kind of radiation and how much you have on the cpu ?
I think the cpu simple crash because emi or some other kinds of
interference.
:
I shouldn't have said the word "radiation". Seems a little bit like
"Jeh*v@"...
Radiation is just an example cause for my bit flip concern.
And JTAG activation is one aspect which cannot be detected by internal
logic, because that's what JTAG is for...
The other thing which came to my mind are bit flips in SFRs (and RAM of
course). These could also lead to complete failure of the firmware
application, but at least bit flips in SFRs can be caught by refreshing
the registers. E.g. a bit flip in P2SEL could turn an Rx pin into a
standard input pin but the application would never receive again a byte
- avoided of P2SEL would be refreshed periodically.
So my questions:
- is anybody taken such cases into consideration or am I a litte bit
hysteric about that?
- if I'm not the only hysterical person: how did the others get out
of the situation?
Hardy
