Re : Re: Re : Re: Re : Acquisition measurement index

[Zinnk]

I was testing the Labview example Acq and Graph Voltage-Int Clk.vi and trying to acquire some analog signals. I applied the same signal in to 4 channels and then measure the max and min values and their respective index. The result was that all the channels have the same value for max (as expected) but with different index. How can this be possible when it’s about the same signal?? Anyone has an idea about this??

Thanks

[Steve__M.]

Hi,

It depends on the way that you do your acquisition. When you said that you applied the same signal into 4 channels, is it at the same time ? Or did you do it channel per channel and execute the VI each time ?

Regards,

[Zinnk]

Hi Steve M.

The aquisition and measurment of 4 channels was at the same time. 

[Steve__M.]

Hi,

What's your DAQ device and your LV version ? In order to reproduce what you see, could you give us your VI ?

Regards,

Message Edité par Steve M. le 29-12-2009 03:53 PM

[Zinnk]

The code is basically a LV example with just some extra measurement vi. I'm using DAQ PCI-6024E and LV version 7.0

Thanks for the work!

[Steve__M.]

Hi,

I reproduced this behaviour but according to your screenshot you should set the Sample Rate at 50k maximum for the 4 channels, because the 6024E has a max sampling rate of 200 kS/s for 1 channel. What you see is expected because it might have some noise in your measurement which can affect the place where it detects the max and the min for each channel, therefore the index of them. And to finish justa little tip you can just write Dev1/ai0:3 instead of Dev1/ai0: Dev1/ai3.

Regards,

Message Edité par Steve M. le 04-01-2010 12:08 PM

[Zinnk]

I’ve tried with the Function Generator from BNC-2120 to provide the signal and the result was affected by the same problem. So where the noise came from? And how can I remove it? I changed the sampling rate as you said.

Once again tank you Steve M.

[Neil_S.]

Zinnk,

The 6024E is a multiplexed board which means that channels are never actually acquired at the exact same instance but are acquired one following the other.  At max rate for 4 channels the delay is ~20us between channels.  In addition you are looking for the max in a sine wave.  Since the max of each cycle is almost identical each channel might find the max on a different peak of the sine wave.  Any noise at all will cause different channels to find different peaks.  If the shared signal were more of a single peak I would expect the indexes to be much closer.  By the way, what exactly are you trying to do?  What will be you test signal in the end?  Would it be possible to test your application with a test signal that is closer to what you will actually be testing?

Regards,

Neil S.

[Zinnk]

Hi Neil S,

I’m trying to acquire 4 signals and make some measurements, but first I want to make sure that the acquisition is being correctly performed. That´s why I applied the same signal in  4 different channels. This way I can test the synchronism, graph it and detect peaks.

[Neil_S.]

Zinnk,

Thanks for the additional explaination.  Using a test signal to test your application before deploying it makes sense.  However, it is important to consider what a valid test signal will be.  If you are interested in the index of the maximum value (which apparantly you are) than using a sine wave doesn't make much sense.  If you have a sine wave with multiple periods what is the 1 index of the maximum?  You can't really reliably choose a maximum since each period has a maximum.  Even in the ideal case this 1 maximum is undefined since there is more than one.  The maximum would actually be any index that is pi/2 into the sinewave.  If you only have 1 period of the sine wave than the maximum becomes more deterministic.

Why don't you use a single peak to determine if your application is working correctly?  Will your final application be measuring a sine wave?

Regards,

Neil S.