On Thursday, May 2, 2013 10:23:58 AM UTC+1, The Natural Philosopher wrote:
On 02/05/13 09:59, Andy Champ wrote:
On 01/05/2013 13:54, whisky-dave wrote:

You ought to measure the time for 50 (or something) cycles instead. If
you time it at the zero crossing that doesn't move about much.

Even so - yes, it's hard, especially with a noisy waveform.

don't be silly. an active low pass filter or even a 50hz high Q bandpass
filter with around a 30dB/octave fall off will remove all the trash first.

Yup. A capacitor, repeated counts and a few mistimes are cheaper.

You aren't even bothered about phase issues - just the fundamental
frequency. And in the limit you can use something like an FM detector if
you want.

countings cheaper

There are a million ways to skin this cat, and none are hard.

quite


NT