Signal analysis of non periodic Signal

[db2nc]

Hi,

I successfully established a communication between my sensor and Labview. First Labview sends out the "send-message" and then I read them back. It works fine. In the next step I have to analyse the non periodic signal. As I've problems loading up pictures in the forum, I put the image in the VI. I

The signal I have is non period and I have to measure the exact time (µs) of the signal in yellow and red part (see VI). How can I do that? The normal VI don't work as it is non periodic.

[GerdW]

Hi db,

search for edges in your signal (rising/falling). Get the distance between those edges...

P.S.: Es gibt auch ein deutschsprachiges LabVIEW-Forum. Vielleicht bist du dort auch gut aufgehoben...

[Norbert_B]

Most important question:

The waveform ALWAYS looks similar to this?

If so, you have to develop an algorithm identifying the keypoints for your signal you want to measure (start point of the yellow/red line as index, both end points again as index).

The time "for each line" is (index end point - index start point) * delta_t. delta_t is your 1/samplerate.

hope this helps,

Norbert

PS: I cannot insert images here when using Microsoft IE. Using Firefox works fine.

[db2nc]

Well, my first problem is how to cope with the output signal - I like to convert it into an array of double and then go throw the measured values and compare them. But how do I convert this signal-data into something else? My first results did not lead to something like a 1D Array where the first element is the first measured value and the last element is the last measured value.

[Norbert_B]

There are several ways to get the numeric array from the waveform datatype which you are currently dealing with:

a) Reconfigure the DAQmx Read function to return a 1/2D Array (depending on number of channels!) instead of waveform/1D array of waveform.

b) Extract the data array from the waveform by using the "Get Waveform Components" function.

The delta_t is included in the waveform, so you might go for option b) and also read this. But (at least as first approximation) the samplerate you configure as inverse is also your delta_t.

Norbert