waveform graphs

[Rutz]

In one application I need to display & measure peak to peak amplitude of a sinusoidal waveform. Due to some unwanted spikes present in one or odd cycles the measured values are wrong as it is measuring the peak of unwanted spikes. I dont know the source of these unwanted spikes. I could know about it only from the waveform displayed.

How do i ignore these spikes?

The waveform is actually measured by NI Express scope VI. Do I need to configure it? how?

[pallen]

Sounds like you might need to filter your data.  LV has some built in filters, as well as add-on toolkits that might be useful. 

You could also use a sort of "debounce" on your data as well and discount peaks that aren't long enough in duration to be part of your actual signal. 

[Rutz]

Thanks for replying!

Are there any worked examples in LV Help for this? Actually, I am very new to LV and know very little about it.

[pallen]

Hi Rutz,

I personally have not done a lot of work with filters in LabVIEW.  But I know has flexible and powerful filtering capabilities. 

Have a look in the Example Finder for "Filter" or "Filtering".  (Click on Help -->Find Examples to open the Example Finder.  Make sure to click on the "Include ni.com Examples")

There's one there Called Express Filter.vi that should help at least a little.

Also have a look in the NI Search for more information on Filters in general.

http://zone.ni.com/devzone/cda/epd/p/id/2467

Good Luck.

[Paul_C.]

Hi Rutz,

Although it might be important to determine what is actually causing these "spikes" in the first place.  You could get away with filtering out the spikes as suggested.  This would still require you to know a little bit about the "spikes" to determine which is the best filter to remove them.  If you can determine this information, I would recommend taking a look at the example finder as suggested.  I would also take a look at the signal processing » filter palette.  You can get to this palette in LabVIEW by right clicking on the block diagram and going to signal processing » filters.  You can also use the filter express VI which can be found by right-clicking on the block diagram and going to Express » Signal Analysis » Filter.

I hope this helps,
Paul C.

[Rutz]

Hi Paul!

In my application I am using a LVDT demodulator followed by a filter. to test the frequency response of the filter i need to feed amplitude modulated input with varying frequency of the modulating signal. the filter is a 1st order low pass filter with a cut off freq of 19.4hz. so i need to check the filter's response over a range of 1.94 Hz to 194 Hz. I am measuring the demodulated ac output of the board using NI scope VI available in LV 8.0. This is the signal where I was getting unwanted spikes. I tried capturing this signal in another CRO and observed that spikes' freq is around 50 Hz to 200 Hz (over a range of 1.94Hz to 194 Hz).

Usage of a filter, as suggested, in LV to filter out these spikes may suppress my filter's actual output itself. 

Best would have been to kill this noise at the hardware level itself but unfortunately the hardware is frozen and I cant touch the hardware. Thats why I was looking for some solution in the " techniques to extract"  the true signal. Any other suggestion?

I tried using tone measurement instead of amplitude/level measurement. It helped to a good extent but is there any better alternative?

Thanks

Rutz