> > to be able to wirelessly reprogram the flash for firmware
upgrades.
> > The low level drivers for wireless could be copied to ram
together
> > with the function to trigger the IAP routines. Then by
sending small
> > packets I think one should be able to reprogram
>
> Ow. This sounds like a recipe for disaster to me. Wireless is
about as
> unreliable a network medium as one can imagine...
>
> * --
> ; -- Lewin A.R.W. Edwards
(
http://www.zws.com/)
>
Learn how to develop high-end embedded systems on a tight
budget!
>
http://www.amazon.com/exec/obidos/ASIN/0750676094/zws-20
Ever done Power Line Carrier Networks - that's the pits I reckon.
(Maybe better nowadays, haven't kept up anymore)
Milos' scheme could work, _proviso_ the old firmware is retained
until the complete RF network session is final, then the vectors are
reprogrammed.
If you however set up things in RAM, experience a link failure
mid-session,
followed by a Brown Out or whatever that trashes RAM, it's
bye-bye system.
Wireless CAN be very reliable, providing proper link budget studies
are done,
and account for the environment. (blocking, jamming, multipath etc)
I once designed a very low power RF (medical) system,
one that used wireless
transfer of critcal real time acquistion data (EEG,ECG,EMG etc) where
full reliability
(meaning - NO missing concurrent data from any of a dozen or so nodes
at all times)
was paramount.
This was at 915 MHz ISM with ~ +6 dBm TX - it is possible - but
it's the hardest
mixed system I ever designed from the ground up.
It probably retarded me for years to come :-)
The least reliable component probably was the Host PC .....
-- Kris