Newsgroups: comp.robotics
Path: brunix!sgiblab!darwin.sura.net!europa.eng.gtefsd.com!uunet!elroy.jpl.nasa.gov!decwrl!netcomsv!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: 68HC11 Reset questions...
Message-ID: <nagleCDCuCw.HCz@netcom.com>
Organization: NETCOM On-line Communication Services (408 241-9760 guest)
References: <273m51$n29@telerama.pgh.pa.us>
Date: Tue, 14 Sep 1993 17:38:55 GMT
Lines: 33

steve@telerama.pgh.pa.us (Stephen D. Cohen) writes:
>     The Motoroal reference manuals for the 68HC11 series say, in
>several places, that it is absolutely critical to handle the reset
>line in an inteligent maner during power down of the chip to prevent
>run-away execution from writing to EEPROM and wiping out valuable data
>(such as the config register, or user code).

>1) Has anybody actually *experienced* this phenomenon?  Or is it a
>hoax on the part of Motorola to sell their reset control part?
     It's real.  What typically happens is that as power fails, the
chip continues to execute instructions, but does not execute jumps.
So every instruction in memory gets executed several times.

     You can thus get into trouble if the sequence for writing EEPROM
exists anywhere in memory.  However, if you code your EEPROM writing such
that the instructions are constructed on the fly, and only exist in memory
when an EEPROM update is desired, you can program around the problem.
Motorola has an application note on this, which I have dug up before and
posted to the net.

>2) If it is real, how come so may circuits I see ignore it? 
>(Miniboard, Nuts n' Volts robotics column, one other I don't recall.)
     Because they aren't building hardware which will cause product returns
if it fails.

>3) Is there a simple way to handle this without using some esoteric
>Motorola (or other) part?  I'd sure like the average experimenter to
>be able to build my design without having to scrounge *too* much.
     Regulators with a RESET output and power validation logic are
available from National Semiconductor, and, I think, Maxim.  Probably
others, too.
     
					John Nagle
