Real-Time Measurement and Control

cancel
Showing results for 
Search instead for 
Did you mean: 
Reply

NI 5732 Acquisition

Solved!
Go to solution

Hi all,

 

I have a NI 5732 used with NI PXIe 7962R, and they are embedded in an NI PXIe 1073 chassis. I wanted to test the data acquisition so I connected a function generator to the AI0. However, something weird happend. The acquired data seemed to be the derivate of the signal sent from the function generator: the sine was still sine, but the square wave became pulses, and ramp became square wave! Has anyone saw this happening before? Thank you!

 

Best,

 

Tong

0 Kudos
Message 1 of 10
(3,508 Views)
Solution
Accepted by topic author li1237
08-27-2015 04:09 PM

What was the frequency of the signals you were generating, and what options do you have selected for AC/DC coupling and filtering? If you have the input set to AC coupled or one of the filters enabled that sounds like expected behavior if the signal of interest lies within the band of attentuation.

Message 2 of 10
(3,501 Views)

Hi David,

 

Your suggestion helps me a lot. Thank you! It is the problem of AC coupling. When I set it to DC the acquired signal is correct. 

 

However, another problem arises. When there is no input, the output of FPGA is not stable. It outputs a lot of points having values more than 60000. I think it should stay at about 32768 (=2^15). Why is it happening?

 

Best,

 

Tong

0 Kudos
Message 3 of 10
(3,479 Views)

What clock domain do the analog input signals reside in? The IO node that you get the AI data from needs to be in the Sample Clock domain (or IO Module Clock domain if you are using an older version of the CLIP). If they're not in the correct clock domain you are performing an unsafe clock crossing which could result in glitching. The glitching is likely what you are seeing.

0 Kudos
Message 4 of 10
(3,473 Views)
Hi David,

I'm just using the 40MHz on board clock. I think it's the default one. I'm not familiar with the clock domain you mentioned. Could you tell me more about it? Thank you!

Best,

Tong
0 Kudos
Message 5 of 10
(3,464 Views)

Open the example finder and navigate to the getting started example at Hardware Input and Output>>FlexRIO>>IO Modules>>573x>5732.

 

You'll notice that the analog input node is in a single cycle timed loop that uses the sample clock as its clock source. This means that the "clock domain" that the IO node resides in is the "Sample Clock".

 

Capture.PNG

 

When working with a FlexRIO Adapter Module that has an analog input or output  IO node, there is usually a specific clock domain that IO node must reside in. You'll want to read the help entry for the CLIP that your specific FAM uses for more information.

0 Kudos
Message 6 of 10
(3,448 Views)

Hi David,

 

I tried to connect a "IO Module Clock 0" to the timed loop. I'm using Labview 2011 so there is no "IO Module\Sample Clock". However the following error dialog poped up. I already added the "IO Module Clock 0" in the project. What's wrong with it? Thank you!

Capture.PNG

0 Kudos
Message 7 of 10
(3,425 Views)
Solution
Accepted by topic author li1237
08-27-2015 04:09 PM

The user commands need to be in the 40MHz clock domain. 

0 Kudos
Message 8 of 10
(3,417 Views)

*double post

0 Kudos
Message 9 of 10
(3,417 Views)
Yes you are right! Thank you!
0 Kudos
Message 10 of 10
(3,408 Views)