Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

Synchronize NI9234, NI 9203, and 9237 at 200 Hz

For a research project, I am preparing an acquisition system based on a NI CompactDAQ (NI cDAQ 9184) with NI 9234 (4 accelerometers), NI 9203 (4 current sensors), and NI 9237 (4 strain gages). The hardware will be controlled through a LabVIEW program, saving all data in a common file.

In my project is fundamental that all sensors (in 9234, 9203 and 9237 in same 9184 cDAQ) are perfectly synchronized and it is mandatory that the final acquisition rate is 500 Hz for all sensors, too.

 

Is it possible?

 

After a general search I am still with doubts how can achieve the synchronization of all sensors taking into account that they will be in three different modules:

 

- Is it enough to configure all devices into the same “DAQ assistant” in LaBVIEW to ensuring the perfect synchronization?

 

- In that case, which should be the native acquisition rate since 9234 and 9237 have specific supported rates and they are not coincident (see http://digital.ni.com/public.nsf/allkb/593CC07F76B1405A862570DE005F6836)

 

- Are there a more recommended (proper) ways to synchronize the modules for this purpose?

 

- I found this explanation http://www.ni.com/white-paper/11549/en/ but I am not sure that this may be effective in my case (which figure should I follow?) and if it updated taking into account the last “DAQ assistant” function.

 

- In LabVIEW, how should I perform the decimation to keep the synchronization of data?

 

Excuse my questions but it is not simple to test and to assess the validity of the synchronization.

Thank you in advance for your support.

Regards,

CR

0 Kudos
Message 1 of 6
(4,739 Views)

Hi CRodrigues,

 

I think DAQmx Timing property node is very helpful for testing. Please look at the attached snippets how to use it. In nosychro snippet channels may use different sampling rates which you can see in "SampClk.Rate" indicators. In synchro snippet channels are synchronized and you can see used sampling rate in "SampClk.Rate" indicator.

 

I hope it helps.

 

Best regards,

CaravagGIO

Download All
0 Kudos
Message 2 of 6
(4,692 Views)

Thanks, CaravagGIO, for your support.

I implemented the last solution (the second image) and it seems is working fine and synchronized. Thanks.

 

A last question about the final rate for the saved data:

With the synchronized procedure, I guess that the acquisition rate of all sensors is compatible with the 9134 supported rates, only. To have the final 200 Hz (in an external data file), I need to compress or decimate the acquired signal.

What procedure do you recommend for that propose?

 

Basically, I am with doubts about which of the following two may be the best option:

  1-  Define the lowest rate multiple of 200 Hz, in case 3200 Hz, and then apply a decimate function.

In this case, which function should I use to decimate the multiple channels in the common waveform format adjusting the time field, too? Has Labview an anti-aliasing filter (lowpass filtering) intrinsic to the decimation?

 

  2-  Define a lower acquisition rate (2560 Hz, per example) and apply the expressVI "align and resample" (deselecting the align option).

 

My concern is now to avoid errors concerning the synchronization of all sensors with the final decimation.

 

Thanks again for your support.

CR

 

0 Kudos
Message 3 of 6
(4,611 Views)

The first option sounds better for me because I am not so familiar with this expressVI. I attached an example how to select samples with 200Hz frequency if your sampling rate is 3200 Hz. I hope this will help you to understand the way of "decimation" waveform. 

 

Comment to snippet: The while loop is not synchronized with DAQ but it does not matter because your perfectly synchronized data are stored in buffer. In while loop you read 32 samples (16 samples per channel) and use only the first one which will effectively produce 200Hz sampling rate.

 

Regards,

CaravagGIO

0 Kudos
Message 4 of 6
(4,589 Views)

Thanks CaravagGIO,
I think the attachment is missing. Can you reattach, please?
Regards, CR

0 Kudos
Message 5 of 6
(4,574 Views)

I am sorry. Now it should be attached.

0 Kudos
Message 6 of 6
(4,571 Views)