Zero Crossing detection fluctuations

[temtronic]

When the dinosaurs roamed, I remember using 1/4 of an MC1489 as a zero cross detector. Had builtin hysteresis and optional 'response control' pin.
hmm, probably make good 'logic level' converters, say +12 to +5......

[Ttelmah]

I think ZCD's, have actually 'got harder'. TRIAC dimmers produce so many
nasty spikes on supplies. Combine these with the huge number of switching
devices present in the home, and things like the old school demonstration,
of what mains looks like with a scope, now are likely to show things closer
to mountain ranges in many places....

[Humberto]

A zero crossing detection interface, obviously in an AC environment, will be subject to strong electromagnetic actions,
producing undesirable electrical effects as we all know.
I have done projects with triacs where I had to deal with these events, seen on an oscilloscope in the lab, undesirable
pulses appear all the time for different reasons, as has been said in this thread.
In the real world, we use this technique to control light intensity, temperature, the speed of a universal motor, etc.
and similar elements, which due to their characteristics, none of them will be affected due to their own inertia,
therefore these effects will be negligible in these type of applications.

Suppose that in an hemicycle a trigger occurs at the wrong time due to an undesirable glitch, we must know that this
is possible only in the zone of the hemicycle prior to the triggering of control device (triac). For all the remaining time,
that is, once it enters in full avalanche conduction, the device is insensitive to any type of glitches or undesirable event,
until the end of the hemicycle when the main voltage becomes = 0V
Let's suppose that these random events appear and as a consequence of this, they generate an error (trigger occurs
before its time), well, the whole process from when it appears until it is extinguished can never be greater than 10 ms.
In the next 10 ms (@50 Hz) the error disappears without real affectation, unless the error pulse is repetitive or synchronous.
I assume that the design of the interface hardware is correct and taking into account the element to be controlled
- in this case we don't know - it will be irrelevant if the trigger occurs +/- 1000 us due to fluctuations.
_________________
Humber