PLL for Zero Delay LPF
in [Design]
|
From: John Larkin on 24 Jun 2010 09:42 On Thu, 24 Jun 2010 08:18:28 +0300, Paul Keinanen <keinanen(a)sci.fi> wrote: >On Wed, 23 Jun 2010 19:51:32 -0700, John Larkin ><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > >>On Thu, 24 Jun 2010 02:34:38 GMT, glennkenroy(a)protech.com (Glenn >>Kenroy) wrote: > >>>You mean perhaps to use two identical sensing coils; one tuned to ELF >>>and the second to 60Hz? The output from the second could then be >>>adjusted in amplitude and inverted to null the 60Hz component in the >>>output of the second. >> >>Actually, I was thinking of using the local power line. But a >>divided-down crystal oscillator would work as well. It only needs to >>be pretty close to 60 Hz. > >If the device is mains powered, why not use it directly as a reference >for cancellation. Yes. > >You may need the ability to shift this reference phase, since the pick >up coil may also pick up magnetic fields from other phases, creating a >resultant, which is not in phase with your equipment power. > The i/q cancellation loop automatically corrects for phase and frequency variations. John
From: JosephKK on 25 Jun 2010 10:25 On Thu, 24 Jun 2010 02:48:17 GMT, glennkenroy(a)protech.com (Glenn Kenroy) wrote: >On Wed, 23 Jun 2010 16:58:39 -0700, dplatt(a)radagast.org (Dave Platt) >wrote: > > >>>>Does your filter really need to be realtime? > >Yes please, for this application. > >>You can split the signal in half, and feed >>part of it to a very narrow-bandwidth "band-pass" filter which lets >>through only the 60 Hz component, > >Or, as I replied to John's post, how about deriving this signal >instead from a second sensing coil tuned to 60Hz? Then adjust >amplitude, invert and mix with to the output of sensor #1 to null the >60Hz component there. > >Glenn Kenroy It took a long time for this to percolate back up from old memory. About 30 or so years ago Popular Electronics did an article on a old school analog distortion analyzer. It used a tunable twin T active filter to remove the fundamental, and since it was servo'd in both frequency and phase you could get up toward 100 dB rejection. All this with little delay and well understood and simple phase response. I suppose this technique could be used here, perhaps with a phase canceling pair of servo'd twin T active filters.
From: Phil Hobbs on 25 Jun 2010 10:34 On 6/25/2010 10:25 AM, JosephKK wrote: > On Thu, 24 Jun 2010 02:48:17 GMT, glennkenroy(a)protech.com (Glenn Kenroy) > wrote: > >> On Wed, 23 Jun 2010 16:58:39 -0700, dplatt(a)radagast.org (Dave Platt) >> wrote: >> >> >>>>> Does your filter really need to be realtime? >> >> Yes please, for this application. >> >>> You can split the signal in half, and feed >>> part of it to a very narrow-bandwidth "band-pass" filter which lets >>> through only the 60 Hz component, >> >> Or, as I replied to John's post, how about deriving this signal >> instead from a second sensing coil tuned to 60Hz? Then adjust >> amplitude, invert and mix with to the output of sensor #1 to null the >> 60Hz component there. >> >> Glenn Kenroy > > It took a long time for this to percolate back up from old memory. > About 30 or so years ago Popular Electronics did an article on a old > school analog distortion analyzer. It used a tunable twin T active > filter to remove the fundamental, and since it was servo'd in both > frequency and phase you could get up toward 100 dB rejection. All this > with little delay and well understood and simple phase response. I > suppose this technique could be used here, perhaps with a phase canceling > pair of servo'd twin T active filters. I just bought an HP 339A Distortion Meter that does this. It's a great tool. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal ElectroOptical Innovations 55 Orchard Rd Briarcliff Manor NY 10510 845-480-2058 email: hobbs at electrooptical dot net http://electrooptical.net
From: John Larkin on 25 Jun 2010 12:31
On Fri, 25 Jun 2010 07:25:48 -0700, "JosephKK"<quiettechblue(a)yahoo.com> wrote: >On Thu, 24 Jun 2010 02:48:17 GMT, glennkenroy(a)protech.com (Glenn Kenroy) >wrote: > >>On Wed, 23 Jun 2010 16:58:39 -0700, dplatt(a)radagast.org (Dave Platt) >>wrote: >> >> >>>>>Does your filter really need to be realtime? >> >>Yes please, for this application. >> >>>You can split the signal in half, and feed >>>part of it to a very narrow-bandwidth "band-pass" filter which lets >>>through only the 60 Hz component, >> >>Or, as I replied to John's post, how about deriving this signal >>instead from a second sensing coil tuned to 60Hz? Then adjust >>amplitude, invert and mix with to the output of sensor #1 to null the >>60Hz component there. >> >>Glenn Kenroy > >It took a long time for this to percolate back up from old memory. >About 30 or so years ago Popular Electronics did an article on a old >school analog distortion analyzer. It used a tunable twin T active >filter to remove the fundamental, and since it was servo'd in both >frequency and phase you could get up toward 100 dB rejection. All this >with little delay and well understood and simple phase response. I >suppose this technique could be used here, perhaps with a phase canceling >pair of servo'd twin T active filters. Unless they're using their own generator, the 60 Hz line is incredibly stable, so a fixed-frequency high-Q notch will work fine. Its width needs to be compatible with sidebands caused by any amplitude variation in the amount of 60 Hz pickup. John |