-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 07/06/11 21:05, RNGolding@aol.com wrote:
> Hi, not sure if this should be PIC or EE, but I have opted for  PIC
> =20
> I have an application where I need to measure the phase difference betwee=
n =20
> two sine waves of same frequency. I'm using two zero crossing detectors t=
o =20
> trigger the CCP1 and CCP2 inputs of a PIC. The CCP counter is  clocked by=
 a=20
> PLL/VCO at 360 x the sine wave frequency, so the  difference between the=
=20
> CCP1 and CCP2 counts provides a direct phase measurement  with a resoluti=
on of=20
> 1 degree.
> =20
> So far so good, however the sine waves have some jitter so I thought I =20
> would take the average of a few consecutive phase measurements. That is  =
where=20
> I come unstuck, how do you average phase readings?
> =20
> To illustrate the problem, if the phase is jittering about zero  degrees=
=20
> giving 4 phase readings of say, 358, 359, 1 and  2 degrees, a simple aver=
age=20
> calculation yields 180 deg, which is  obviously wrong. Does anyone know h=
ow=20
> calculate the true average,  which  should be 0 degrees? I'm sure the ans=
wer=20
> is obvious, but it has defeated me so  far.=20
> =20
>  Richard Golding
>  =20

In DSP your problem is called phase-unwrapping. See
http://www.dspguide.com/ch8/9.htm for an example when applied to a DFT
or FFT.

If you can guarantee that your real-world phase doesn't change by more
than 180 deg between each sample you could test if | x[n-1] - x[n] | >
180 deg. If so you know you've jumped across the boundary and can adjust
accordingly by adding 360 deg.

Regards,
Brendan

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk3uhwYACgkQHEhZ5Ws5poGJKwCeONWQGnVf35ezD1YpOLFs2J1w
na0AoKU9+U1ewso5utV6QfzR+GufT/en
=3DWw8h
-----END PGP SIGNATURE-----
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.