LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Data acquisition using Labview FPGA

Solved!
Go to solution

What do you have for the number of samples to read from the DMA?  Hopefully that is not 0 (though, your error description tells me it is set to 0).


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 11 of 23
(690 Views)

Hi I have number of samples as 1000. Please find the images of my VI and waveform chart. I expect a 4Vpp sine wave of 1Khz frequency

Download All
0 Kudos
Message 12 of 23
(689 Views)

At least we have the right general shape now.  Your FPGA can sample at up to 1MS/s.  So with a 1000 samples and one full cycle captured, you are getting 1kHz.  And the amplitude looks correct.

 

You might want to add a wait inside of your FPGA to have a more consistent and known sample rate.  I am thinking a 10us wait should be good, giving you a 100kHz sample rate.

 

What are you using to generate the waveform?  What sample rate is it using?  I am eye balling a sample rate of ~10kHz.


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 13 of 23
(686 Views)

Yes we have the proper amplitude and wave. I have added the wait in my FPGA, but that didn't seem to improve my output. How do I get a smoother sine wave? I also wanted to increase the range of x axis to get a clearer waveform. How do we do that?

I am using a sine wave generator AFG 3021B with sample rate 250MS/s.

 

0 Kudos
Message 14 of 23
(681 Views)

@pratikpade wrote:

I also wanted to increase the range of x axis to get a clearer waveform. How do we do that?


Increase the history of your chart.  Right-click on the chart and choose "Chart History Length".  The default is 1024 samples.


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 15 of 23
(677 Views)

@pratikpade wrote:

How do I get a smoother sine wave?


Please attach your entire project. I am talking the project file, target VI, and host VI.  I cannot debug anymore with pictures.  I need actual code and environment.

 

My latest observation is that the read waveform only has integer values.  So I think you have a really bad FXP configuration for your DMA FIFO, which is causing the decimal places to be truncated/rounded off.  More evidence of this shows up in the Target VI since there is a coercion dot going into the DMA Write node.


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 16 of 23
(673 Views)

I tried changing the FIFO data type to u32 which gave me a square wave. Also I tried using the "Get number of elements" method but I was not able to place it correctly.

Please find the attached project as well as the VIs.

 

Please refer to Vis:

Host: test_6_22_part2(HOST).vi

Target:test_6_22.vi

 

Thank You

 

P.S: I would like to apologize in advance for the excess VIs in the project and bad designing of VIs.

Download All
0 Kudos
Message 17 of 23
(668 Views)

@pratikpade wrote:

I tried changing the FIFO data type to u32 which gave me a square wave.


Yeah, that is definitely not going to work.  The data type of the DMA should match exactly the FXP configuration of the AI node or a SGL.

 

I am away from the office, but I can look at your code in the morning.


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 18 of 23
(661 Views)

Thank You for your efforts. By that time I will keep trying and let you know if I can get something.

0 Kudos
Message 19 of 23
(658 Views)

Yep, I just confirmed what I thought was happening.  The FXP settings for your DMA currently has a resolution of 1.  So any decimal places are going to get truncated or rounded.  My suggestion is to just make the DMA data type be a SGL, which is a floating type that is now supported by LabVIEW FPGA.  This is also a data type that can be wired straight into your chart.


GCentral
There are only two ways to tell somebody thanks: Kudos and Marked Solutions
Unofficial Forum Rules and Guidelines
"Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
0 Kudos
Message 20 of 23
(646 Views)