Olin, the PIC input parasitic capacitance (somewhere between 5 and 10pF)
with a 4.7M will already gave a time constant of  4.7exp6 x 10exp-12 =
50uS. So the high frequencies are rejected from the very begining. A
simple low pas filter will be then enough. (1..3mS time constant)

However I presume there is another problem there, like an inefficient PIC
supply filtering or a bad MCLR via configuration word ...


top 10 wishes,
Vasile
http://surducan.netfirms.com


On Mon, 15 Sep 2003, Olin Lathrop wrote:

> > The Neutral terminal of the Mains is connected to a PIC input through a
> > 4.7M resistor for detecting zero crossings. (Some may consider this
> > dangerous, but since the power supply itself makes the whole circuit
> > "live", this doesn't make any big difference.)
>
> You should put some low pass filtering in there.  You circuit will detect
> the zero crossings of the power line just fine, but it is very susceptible
> to noise near the zero crossing.  A 5V spike near 0 will cause additional
> edges, and 5V is a very small noise spike as power line standards go.  I
> would use two successive R/C filters to reduce the high frequencies, then
> measure the zero crossings.
>
> Of course now there will be some phase shift.  I don't know if this is
> acceptable in your application.  If the filters aggresively attenuate the
> line voltage (you've got plenty to start with), then the phase shift will
> be very predictable and quite independent of part value variations.
>
>
> *****************************************************************
> Embed Inc, embedded system specialists in Littleton Massachusetts
> (978) 742-9014, http://www.embedinc.com
>
> --
> http://www.piclist.com hint: The list server can filter out subtopics
> (like ads or off topics) for you. See http://www.piclist.com/#topics
>

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics