LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Slow stripchart update with LabView 8.0.

On one of our LabView panels, we have several x-y stripcharts.  When we run the program, these plot update very slowly, (5-10 seconds per update).  However, we have two "Viewer" computers, running the same compiled code, that do not have this issue.  The Viewer computers receive data from our main computer over ethernet, then display all of the same panels/data as the main computer.  When the same code is compiled on LabView 7.1, there is no stripchart update problem.  Any suggestions would be appreciated.
0 Kudos
Message 1 of 4
(2,894 Views)
Hello,
 
Is this something you can post, or can you reduce your code to something which you can post so I can take a look at it?  Just to clarify, you see fast/normal behavior in LabVIEW 7.1 and slow behavior in 8.0 with the same program on the same computer?
 
This is not the first time such an issue has come up, but most of the time it has a reasonable explanation, such as trying to plot and retain huge amounts of data.  In general, you can also turn off autoscaling or even defer panel updates at times since those require screen updates which can definitely affect execution time.
 
I look forward to your repost, and hope we can find an underlying cause and corresponding solution in your case!
 
Best Regards,
 
JLS
Best,
JLS
Sixclear
0 Kudos
Message 2 of 4
(2,877 Views)
We can't post our program at this time.  Normal behavior was observed in LabVIEW 7.1 and slow behavior in 8.0, (and the 8.0.1 patch) with the same program on the same computer.  Computers were switched out with no improvement.  We store data at 1Hz intervals for about 300 channels, which is a large amount of data. The program is in two main parts, one for insturment control, the other for data observation.  We require fast updates on the stripcharts as this information is used for control, so defering panel updates is not an option for us.  Autoscaling is not used for our application.
 
If we split the program into two sections and start LabView twice on the same computer, the problem goes away.  So it does not seem to be a problem with our computer hardware.  It appears as though there is a processing loop somewhere between our control panel and stripcharts that is slowing everything down.
 
Thanks for your help.
 
0 Kudos
Message 3 of 4
(2,873 Views)
Hello,
 
The fact that launching two instances of LabVIEW helps is interesting, and perhaps points to a threading issue; can you or have you tried breaking the code into two VIs and playing with the execution properties (file -> vi properties -> exectution) with only one instance of LabVIEW running?
 
Also, how is it that your control is dependent upon specifically updating the strip charts - is there an operator constantly monitoring those?  Basically I am wondering if perhaps it will be possible to defer panel updates for certain periods of time, or update the chart less frequently by writing more than one point at a time (maybe try 5-10 points at a time).
 
In any event, to that affect I wanted to see the strip chart performance so I put 20 strip charts on a front panel being updated with random uniform(0,1) values (using the LabVIEW random number function) and observed the average execution time over a hundred or more executions of the VI in both versions 7.1 and 8.0 using the profiling tool (See Tools -> Profile -> Performance and Memory... in version 8.0, and Tools -> Advanced -> Profile VIs in verison 7.1).  The performance was very similar in both 7.1 and 8.0.  Thus, perhaps this is less due to differences across versions as far as panel updates are concerned, and more to do with how the underlying buffers are being handled.  What buffer size are your charts using, and how many charts are you updating?  For reference, I used 1024 point buffers.  Perhaps you can isolate at least the performance difference (doesn't have to be on the same scale) so that you can post a simple example (I understand you can't post your code at this time).  If I can reproduce the problem, I will explore it and be in touch with R&D accordingly.
 
I look forward to hearing from you!
 
Best Regards,
 
JLS
Best,
JLS
Sixclear
0 Kudos
Message 4 of 4
(2,857 Views)