🚀 go-pugleaf

In article <9102182138.AA22294@en.ecn.purdue.edu> ditz@ECN.PURDUE.EDU (Michael R Ditz) writes:
>
>I am working on an X-10 based home control system for a senior
>project.  The part that I am having trouble with is the
>zero-cross detector.  What we need is the AC line as input, and 
>some "nice" voltage output pulse at each zero-cross.  One of
>the problems is that a transformer CANNOT be used, because it
>will shift the phase.

	Actually, a transformer would do nicely.  They don't shift
the phase unless you're saturating something.
	A useful approach is to use two phase-shifting networks,
one leading, the other lagging, like:

                                         |\
                                         | \
 AC hot------VVVVV-----+----||-----+-----|  >-----output
              R1       |   C2      |     | /
                       |        R2 >     |/
                   C1  =           >
                       |           >
 AC neutral------------+          GND


Use R1*C1= R2*C2, make R1 much smaller than R2 (factor of a hundred?)
so that the two RC pairs are both lightly loaded by the next stage,
and try for time constants large compared to 1/60 second (because 
you NEED some attenuation here).  The amplifier shown could be a
CMOS Schmitt trigger, or comparator; 'GND' can be any convenient DC
bias point (if you use +5 regulated power, +2.5 V would be a good 
choice).  I assume that you are using NONisolated power for your
logic circuits; else that amplifier up there will have to include 
some isolation.

	John Whitmore