LabVIEW
Saving data to file (frequency measurement)
As a first recommendation, generally it's easier to work with the DAQmx drivers than the DAQ Assistant and the Dynamic Data Type (blue wire).
The DDT can hide all sorts of behaviours...
Certainly it's possible to store multiple inputs together in one file - using CSV/TSV will be easier than an Excel-specific file (e.g. xlsx), and can still be opened in Excel.
If you move away from the Express VIs, the functions you're looking for are Create/Open File, Close File, and Write to Text File, probably along with Array to Spreadsheet String or Format into String.
You'd want to open a file before your loop, close it after, and write in each iteration if using the Format into String.
If using Array to Spreadsheet String, you might prefer to do all of the writes after the loop finishes (RAM allowing).
Then you could also look at "Write Delimited Spreadsheet.vi" or something, which handles the opening/closing for you.
What do you want to put in the Frequency columns? You're currently getting a spectrum from the Express VI, and that has two scalar values and an array of amplitudes - "f0", "df", and "magnitude" (the array).
To write a file in the way you describe, you need to get a scalar value for "frequency". How will you calculate this?
You might prefer looking at other methods to find the dominant frequency given a waveform - try the Waveform Measurements subpalette of the Signal Processing palette.
Hi marcin,
@Marcin_zxc wrote:
There are some random numbers in the saved file and I don't know what they mean.
There are no "random" numbers, only your spectrum data.
Btw. you forgot to include a delimiter char between those two "columns" in your text file…
(By using only one FormatIntoString functions instead of two it would be easier to avoid that error!)
@Marcin_zxc wrote:
I also can't get data from 2 simulated sines simultaneusly, do you have an idea how to this?
Right now you only save the data of one waveform to your file. When you want to save the data of both waveforms you need to handle both!
Somewhat cleaned up version of your VI:
(You might transpose the array before converting to spreadsheet string…)
Hi Marcin,
@Marcin_zxc wrote:
I want the data there to be every 1 Hz, not 10 like it's currently being saved. Do you know how to do this?
I cannot open your VI because I'm limited to LV2017 right now, so just guessing:
When you save the data as shown in my snippet then "df" sets the steps of the frequency values. When you need lower df values then you need to change the FFT parameters…
As GerdW already said, the df value is calculated based on the FFT parameters, which come in this case straight from the sampling parameters (windowing can change this).
If you want a lower df, you need a longer time series when you perform the FFT. You could most easily achieve this by increasing the number of samples from your "acquisition", but you might more generally consider accumulating multiple iterations of acquisition and then performing an FFT on the combination of them (to keep the loop responsive).
Attached as a snippet is a somewhat tidied version of your code for possible reference.
I'd still be tempted from here to adjust some behaviours (ordering of the outputs for example, probably some of the data structures, I attempted to remain close to the originals here) but I don't think I changed any behaviours (although the settings for the express VI might be adjusted on the bottom one, I played with that a little and it might be included in the snippet).
