Forums

Adaptive solution +power line harmonics

Started by DJT November 13, 2008
Hello

sorry to bother you again but I am confused.

What I understand from the threads I read is that a comb notch filter is a
good choice to remove power line interference, assuming that the frequency
is stable...

But say that the frequency fluctuates significantly, some say that there
is no use to go adaptive, while some claim that it should work...

I don't see why the suggestion of measuring the power line frequency and
using it to set the spacing of the comb filter would not work...

or for that matter, why using an NLMS filter with the scaled power line
voltage as reference, in a noise cancellation configuration, would not
work...

Could someone shine some light on this thread

Kind regards

JT
On Thu, 13 Nov 2008 04:09:42 -0600, DJT wrote:

> Hello > > sorry to bother you again but I am confused. > > What I understand from the threads I read is that a comb notch filter is > a good choice to remove power line interference, assuming that the > frequency is stable... > > But say that the frequency fluctuates significantly, some say that there > is no use to go adaptive, while some claim that it should work... > > I don't see why the suggestion of measuring the power line frequency and > using it to set the spacing of the comb filter would not work... > > or for that matter, why using an NLMS filter with the scaled power line > voltage as reference, in a noise cancellation configuration, would not > work... > > Could someone shine some light on this thread > > Kind regards > > JT
Light: You have a bunch of experts who are giving you conflicting advice, and you've tried to resolve the problem by asking them to stop arguing and all please tell you the same thing so you can follow the flock. It's time to turn into a goat, step out on your own, use the collected advice to figure out what's best for _your_ product, and move on. So start prototyping algorithms & see if they'll work for you. Or survey your market to see what you need. Or do just about _anything_ other than trying to get a bunch of engineers to all agree on some technical detail when they've already showed a propensity to (quite gleefully) argue this particular point until the sun goes dark and there are no more AC power lines on the Earth anyway. And if you run into anything else, ask here, and be prepared to sift through the answers for the nuggets that are of value for you. -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
On Nov 13, 5:09&#2013266080;am, "DJT" <JanTorgrims...@hotmail.com> wrote:
> Hello > > sorry to bother you again but I am confused. > > What I understand from the threads I read is that a comb notch filter is a > good choice to remove power line interference, assuming that the frequency > is stable... > > But say that the frequency fluctuates significantly, some say that there > is no use to go adaptive, while some claim that it should work... > > I don't see why the suggestion of measuring the power line frequency and > using it to set the spacing of the comb filter would not work... > > or for that matter, why using an NLMS filter with the scaled power line > voltage as reference, in a noise cancellation configuration, would not > work... > > Could someone shine some light on this thread > > Kind regards > > JT
JT, You use the word "works". What "works" means depends a lot on the application. The same method on one type of signal may "work" to the user's satisfaction, but on another signal it may not be adequate. Some info of interest (definitely not all inclusive): what is application area? what part of your system design has already been constrained, and how? is input signal recorded (analog)? was input signal ever compressed? what Bandwidth/sample rate are you interested in? do you have a powerline noise reference signal available that can be digitized at the same time as your signal? is reverberation tolerable (you would have to define what tolerable means)? do you want to get rid of every harmonic in your reference signal? think about why you answer this the way you do. The more specific you get, the more specific the answers/discussion can be. You might want to try asking again. Dirk
>On Thu, 13 Nov 2008 04:09:42 -0600, DJT wrote: > >> Hello >> >> sorry to bother you again but I am confused. >> >> What I understand from the threads I read is that a comb notch filter
is
>> a good choice to remove power line interference, assuming that the >> frequency is stable... >> >> But say that the frequency fluctuates significantly, some say that
there
>> is no use to go adaptive, while some claim that it should work... >> >> I don't see why the suggestion of measuring the power line frequency
and
>> using it to set the spacing of the comb filter would not work... >> >> or for that matter, why using an NLMS filter with the scaled power
line
>> voltage as reference, in a noise cancellation configuration, would not >> work... >> >> Could someone shine some light on this thread >> >> Kind regards >> >> JT > >Light: > >You have a bunch of experts who are giving you conflicting advice, and >you've tried to resolve the problem by asking them to stop arguing and >all please tell you the same thing so you can follow the flock. > >It's time to turn into a goat, step out on your own, use the collected >advice to figure out what's best for _your_ product, and move on. > >So start prototyping algorithms & see if they'll work for you. Or survey
>your market to see what you need. Or do just about _anything_ other than
>trying to get a bunch of engineers to all agree on some technical detail
>when they've already showed a propensity to (quite gleefully) argue this
>particular point until the sun goes dark and there are no more AC power >lines on the Earth anyway. > >And if you run into anything else, ask here, and be prepared to sift >through the answers for the nuggets that are of value for you. > >-- >Tim Wescott >Control systems and communications consulting >http://www.wescottdesign.com > >Need to learn how to apply control theory in your embedded system? >"Applied Control Theory for Embedded Systems" by Tim Wescott >Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
Can't really argue with your response. I will start prototyping algorithms to see what fits the problem... thanks for the tip-offs... JT
>