Interesting assessment that this will be the last project. We
stopped using AVR in 2000 (only one transfer project from Berlin)
because of poor EMC performance of AT90S4433 even though we 
put a reset IC for all the AT90S. The new AVRs should not
need a reset IC since I do not see any more reset IC in
their new designs using AtTiny26 and AtMega8/16/48/8535.

For debugging purpose, why not put a power-up delay of
100s of ms and see if things getting better? A fast
scope is also very helpful to identify this power up
problems. This afternoon we just performed a code
walk-through of my 500 words 12F629 program and solved
two power-up related bugs. The zoom function of
the scope I used (Lecroy Waverunner LT264 4-ch scope )
really helps even though the memory per channel is still not 
good enough.

Regards,
Xiaofan

-----Original Message-----
From: Michael Rigby-Jones 
Sent: Thursday, September 08, 2005 4:18 PM
To: Microcontroller discussion list - Public.
Subject: RE: [AVR] Brown out problems

...
In ~6 years of 16C, 16F and 18F PIC based products we have never 
seen a brown out issue from this.  This is our first (and probably last) 
AVR based one....

The EEPROM corruption is a worry, but possibly workaroundable, 
what concerns me is the spurious garbage I get out of the serial 
link until the watchdog timer kicks in.  I was wondering if the 
oscillator has not started up properly by the time the BOD has 
released the reset, giving some odd baud rate, but it does not 
explain why the watchdog reset occurs shortly after. The only 
logical explanation I can think of at the moment is the PC is 
wandering off in the weeds.  Unfortunately this is pretty 
difficult/impossible to catch with a debugger!

Regards

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