My 2 cents worth - if you decide you need an anti-brownout
circuit.
I presume that you are interested in volume production - if
nit then the answer is YES - always use an antibrownout
circuit.

I would recommend one almost anywhere unless you are CERTAIN
that the probability of unexpected power supply failure
justifies the savings per design compared to the cost of the
inevitable failures.

I have found the single transistor, three resistor circuit
in Microchip's AN522 (figure 7 in 1997 embedded control
handbook) to be very successful. {{Anyone had trouble with
this cct???}}.The operation point is a little "soft" as
reset occurs when Vcc x (R1/R1+R2) falls below the Vbe of
the transistor. Vbe is not the most precise reference
voltage! As long as the power supply voltage which causes
reset is well above the point at which the circuit will
really fail you should not need the slightly dearer circuit
which also uses a zener diode (figure 6). In your case with
an automotive power supply this should not be a concern. .

You will also want to be *** CERTAIN *** that your regulator
is "automotive environment rated" . You say that it is "very
clean" but does this mean that you have tested it in all
automotive conditions and/or designed it for them or that,
when tried in a few apparently typical cases, it worked OK.

If you are not familiar with what can happen on a vehicular
battery supply then Harris Semiconductors Application Note
AN9312 (available on Harris web-site or I could email it to
you (68KB, PDF)) should give you a good scare. Nat Semi make
some regulators specifically designed for vehicle
environments (and I'm sure many other companies do too.
These are in the LM293x family. The LM2931 is available in
TO220, TO92 and smd (8 pin) and is available with fixed 5v
or variable 3v to 24v output.



Disclaimer:

I don't have any financial interest in any of the companies
mentioned above but I may use their products.





-----Original Message-----
From: G.Daniel Invent Design
<g.daniel.invent.design@xtra.co.nz>
To: PICLIST@MITVMA.MIT.EDU <PICLIST@MITVMA.MIT.EDU>
Date: Thursday, 30 April 1998 21:50
Subject: Re: Power-on Reset IC's


>I've seen two answers to dennis's question so far and
generally I would
>agree. Your first answer mentioned 16f84 self reprogramming
problems
>which I will remember. Here is my opinion for what it is
worth:
>
>If you realy need high volume AND are trying hard to reduce
cost AND you
>want to use a  NON brownout protected PIC (they do come
with brownout
>protection now) AND your application is not a super
critical one then:
>
>1) Do try to have the reset pin perform a usefull function
for your
>program; most ram will still carry your operating
parameters (except
>prog cntr) after an MCLR.
>
>2) Do periodically REinitialise TMRO, baud rate, I/O
direction etc as
>much as possible like a genuine POR reset.
>
>3) Do enable the watchdog timer.
>
>4) Do check your code to be assured that no sub loops may
become
>un-exit-able if ram contents become corrupted.
>
>5) Do use some of your extra memory to insert "goto *"
(here    goto    here)
>loops in between code modules; this will trigger the WDT if
the PIC PCL
>incs blindly past the real code jumps.
>
>6) Don't use CLRWDT instructions too freely.
>
>Hope this helps,
>
>Graham Daniel:end
>
>Dennis Plunkett wrote:
>>
>> At 07:09 AM 29/04/98 -0400, you wrote:
>> >Piclisters!
>> >I am currently using DS1233 econo-resets that hold the
MCLR line low for
>> >approx. 15ms, just after power-up.  My PIC application
is automotive.  I
>> >have a very clean 5v regulator supply.  Slow power-up is
not a problem.
>> >
>> >My question is:  Is this uP supervisor necessary w/ the
PIC (16c76, 16c63,
>> >and 16F84) ????
>> >
>> >Cya-
>> >                         6500
>> >                              7000
>> >Dave Celsnak                    ___7500
>> >                               / /     8000     (Shift!)
>> >                             /-/
>> >                           / /              RPM
>> >                         / /
>> >                       / /
>> >
>> >
>>
>> Slow power up may not be your problem in automotive, but
brown out is. The
>> GM spec puts the 12V supply down to 4Volts under cranking
condtions. If the
>> support cicruits connected to the PIC have the same
supply rating, then
>> there is not a problem. I do think that a reset chip
should be used, as
>> there are other considerations such as the location of
the target device in
>> the auto and what other things are active (On the same
loom)
>>
>> Dennis
>