Notch/Bandstop filter not giving proper result

[Bgmnt]

Hi

I have a signal with with FFT peak at 66Hz as shown ( figure at the top).

I wish to filter out this frequency.

Here's what i wish to ask:

1. Since the LabVIEW versio i am using does not have an inbuilt notch filter, i decided to use a Bandstop filter.

I feel that a precise bandstop ( very liitle stop band) can act as a notch filter or at least help in my case.

Is this true? Or notch filter cant be realized from bandstop? Am I wrong in thinking so?

2. I made a bandstop filter. gave a passband of around 3-4Hz centered at 66Hz BUT no matter what order i choose, The response no doubt filters out the the required band but adds other frequencies! (as shown- figure at the bottom)

What am I doing wrong?

applying a filter is pretty straight-forward isnt it?

                     

Kindly help

[Bgmnt]

please help!

[johnsold]

It looks like the energy at 66 Hz has been spread out into two broader peaks above and below that. 

It is almost impossible to tell what is actually happening because (1) we cannot see what is below -85 dB in the data, (2) you did not post your VI with default data included, and (3) filters are meant to be applied to the time domain data not the spectral data, although it is not clear what you have done.

Please post your VI with data so we can take a look.

Lynn

[Bgmnt]

Hi.

Here's my vi.

To test what was wrong, I took up a signal from DAQ ( named as ' voluntary') added a sinusoid at frequency 100(controllable) and appplied a band stop filter with cut off (100-2)Hz and (100+2)Hz. 

Here's what i am getting.

the signal of my DAQ is a very low frequency signal (0-5 Hz)

.

[Bgmnt]

Also,

case 1 : When i add 2 sinusoids (one low frequency -voluntary) and another high frequency and then apply the same Vi, it works perfectly fine.

case 2 : when i add daq signal(low frequency-voluntary) and a sine with high frequency and apply the same vi, it works for a certain limit of amplitude of high frequency sine. I.e below a certain amplitude, it gives the weird response as shown in figure posted earlier.

NO such abnormality occurs at any amplitude for the first case i.e two sinusoids.

[Bgmnt]

Please help. i am in a fix.

[billko]

Is there a reason you are not actually using the physical filter?  Unless this is some kind of simulation, I would try for physical conditioning of a signal over software.