Andy Kunz wrote:
>
> At 09:03 AM 1/29/98 +1300, you wrote:
> >Does anyone know whether the discrete brownout circuit that MChip
> >recommend i.e the transistor and zener circuit, is significantly less
> >reliable than the Dallas chip, which I suppose will be much more
> >precise. How much do they cost?
>
> Low voltage zeners are not especially accurate, and the cost is probably
> less of a problem.  I have a quote here for 858 pieces DS1233-5 (best
> accuracy, TO-92) at 1.01 each, and that wasn't pressing the issue at all.
>
> For the reliability it was a bargain.
>
> Andy
>
> ==================================================================
> Andy Kunz - Statistical Research, Inc. - Westfield, New Jersey USA
> ==================================================================

I'm not sure if a brownout circuit will cover all aspects of using a
watch-dog. Only errors caused by too low rise time is covered. I guess
that it is much better to add an external watch-dog timer, by using e.g
LM555 to pull off and on the power to the chips. The internal watch-dog
may be used, and the external only if the internal watchdog not
'answer'. If the current is limited, then latch-up is protected too.

Regards,
Jens Dyekjfr Madsen