DAQmx delay in reading?!!

[T-REX$]

Hi Amir,

Does this example work for you? How exactly are you measuring the delay in front panel response that you are seeing?

[ballbouncing]

Hi TJ,

what's the pupose of the vi that you just posted?

Actually I am moving my object, which is attached to the linear control, and then I watch the output signal on the monitor. So when I move my physical object the curve on the monitor follows that but with a 50-60 msec delay. I usually record a video from both objects on the same scene. Then I analyze the .mpg file frame by frame.

Amir

[T-REX$]

Hello Amir,

The example I posted sets up a CI read task that is clocked by the ai Sample clock. In the example that you posted before, I noticed that you were not configuring a clock in your code, so I wanted you to try my code. Did it work for you, or are the results the same? The posted VI is a snippet, so all you have to do is save the image to your disk, and then drag the file to a blank block diagram.

[ballbouncing]

Dear TJ,

I have tested the card delay and it was less than 100 usec. So I think the problem is not with the DAQ card or even the code.

I have imported the periodic pulse data from function generator and via AI-DAQmx blocks to Labview and I sent the same data out on AO-DAQmx. I plotted both analog waves on the same window ossiloscope.  

So now I am suspicious to the graphical presentation of data in labview and my monitor. I have already checked my Nvidia VGA card and it has a very large frame/second data visualization for 2D data and also 3D one. I am using 60 Hz monitor which can add up to 16.66 msec delay. but I have 50 mse delay. 

do you have any idea where the 33.4444 msec delay is coming from?

Thanks

Amir

[T-REX$]

Hi Amir,

Well, as you're running Windows, a non-deterministic OS, it could be anything. The Windows scheduler is going to decide when the data is pulled out of the PC side FIFO. If Windows is busy doing other things, it may take it a while to get the data to your screen. If this was the case, I wouldn't expect a consistent 50ms delay. What kind of resolution does your measurement system provide?

[ballbouncing]

Dear TJ,

Sorry for the delay as I had to work on another project for a while, but now I am back to fix this monster delay with your help of course. As far as it goes with windows I think you're right. But I can not measure windows delay, though it shouldn't be that much large. I should also note that I give the realtime priority to labview in Windows Task Manager. Also what do you mean by resolution; If you mean the monitor resolution it is 1280*1024 16 color bit. 

One more thing that I am not sure is the connection between labview Front Panel and Block Diagram. How long does it take for a data to be updated in the front panel? How synchronized are the two panels?

Amir

[T-REX$]

Hello again Amir,

You said you were using a video camera to take these time measurements. Do you know how accurate this is when generating the time stamp (ie. what and how consistent is the frame rate)? Also, Could we get some more basic information about your application? Why is 50ms, a relatively small amount of time, so important to your application?

[ballbouncing]

Dear TJ,

Actually I was in contact with some of your on call engineers at NI. One of them even came to our office and looked at the experiment. His idea was to replace a DAQmx USB card, that we own, with the PCI card. So it went. I have boroowed the PCI 6035E from NI and I made a lot of tets comparing both DAQ cards. 

To my measurement the difference of these two is in the orther of 1-2 msec, which is not an issue for me. So I am intrested in edlays lower than 20-30 msec.

We are studying human motor control. and my experiment is a ball bouncing task, so the subjects move a physical racket in his hand and the virtual one moves correspondingly on the screen. I used LabVIEW to program that. Now there is a significant delay of 60-70 msec between the physical and virtual racket. 

I have been working on that for a long time. no succes. 

I am using an accurate camera to take a video of both physical and virtual rackets on the same frame. So after that I can estimate the time difference between two objects. 

I am using a pretty high performance dell precision computer.

E5620 @ 2.40 GHZ (2 processors)

8.00 GB RAM

Windows 7 64-bit 

NVIDIA Quadro FX 580

120 Hz monitor

and

Labview 8.6

[leean]

Hi Amir,

Were you able to reduce your delay? We too have encountered similar problem in our lab.

Angela

[YOGESH(YOGI)]

Try this