Noisy Data Acquisition (Multiple Scans, Multiple Channels)

[C. Wennmacher]

I had to perform 3 tasks in 1:
1. Noisy Data Acquisition
2. Multiple Scans
3. Multiple Channels

My problem is to solve effectively all 3 task in 1 application. The Measurement Manual proposes for noisy data acquisition the mean function. But this is too course and not effective for multiple channel measurements.
Any suggestions?

C. Wennmacher

Platform:
LabVIEW 6i
OS: w2k
DAQ 6.8
DAQPad-6020E

[Nikos Lipsanopoulos]

I had to do the same thing for some other testing. I could not find a built
in function to eliminate such noisy signals, so what I did to solve this was
develop a CIN vi, using C programming, that performs a "moving average"
similar to the one that is embedded in the MS Excel Spreadsheet. That CIN
along with some magic from our electrical engineers on the electronics I
use, completely smoothed out the output signal.

Knowing how to use and develop CIN's (using Code Interface Node) has helped
me a lot, because many of the functions included in my version of
BridgeView, are not what I need. The mean function you are thinking of using
is just the average of an array of 10 numbers for example, where as the
moving average of an array of 10 numbers will take a num
ber entered by the
user (e.g. 2) and it will take the average in increments of 2 down the array
list, smoothing out your data.


"C. Wennmacher" wrote in message
news:50650000000800000010160000-983677784000@quiq.com...
> I had to perform 3 tasks in 1:
> 1. Noisy Data Acquisition
> 2. Multiple Scans
> 3. Multiple Channels
>
> My problem is to solve effectively all 3 task in 1 application. The
> Measurement Manual proposes for noisy data acquisition the mean
> function. But this is too course and not effective for multiple
> channel measurements.
> Any suggestions?
>
> C. Wennmacher
>
> Platform:
> LabVIEW 6i
> OS: w2k
> DAQ 6.8
> DAQPad-6020E

[Greg Caesar]

You might want to consider re-posting this question to the data-acquisition newsgroup. LabVIEW Datalogging and Supervisory Control (DSC) is an add-on package for LabVIEW for developing SCADA/HMI applications for industrial automation.

If the over-sampling and averaging the data does not meet your noise-rejection specs, then consider over-sampling in hardware combined with software filtering. You could use the "Acquire N Scans.vi" example as a starting point. The example is located at Search Examples >> I/O Interfaces >> Data Acquisition >> Analog Input >> Multiple Point (Buffered) Analog Input >> Acquire N Scans. You would then enter your channel list and a single scan rate for all channels. Pick a scan rate that is several multiples of your highest noise frequency
. In software, you can then use LabVIEW's filter vi's to individually filter each channel.

Another possible solution is to use SCXI hardware, which would allow you to add hardware filters individually to each channel.

Regards,
Greg