LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Graph precision

Hello everybody.

Recently we tried to bring back to life an old vapor compression cycle with a NI SCXI-1000 connected to LabView 10.0.1. In the front panel, there are 2 graphs, one with temperature and another with pressure. The one with pressure works okay, but the temperature one only shows variations of 1 ºC. I have tried to find when is set to display like that but I have not been able to find it. Below I attach a capture of the 2 cases, pressure as it should display and temperature with the problem. Also, I attach the LabVIEW file.

Thank you in advance. 

Pressure.PNG

Temperature.PNG

  

0 Kudos
Message 1 of 4
(714 Views)

In your lower graph, the data seems to be quantized to whole numbers. This could be because you use the wrong decimal delimiter (comma or  period) when reading the data. Can you attach a datafile?

0 Kudos
Message 2 of 4
(692 Views)

OK, on a computer with LabVIEW now, looking at your code:

 

Oh boy! I think most her are not willing to untangle a gigantic diagram with mostly dynamic data and peppered with glaring Rube Goldberg constructs..

 

  • Adding zero to dynamic data does nothing interesting.
  • setting the tab dimension can be done exactly once, not with every iteration of the loop.
  • You are pining the dynamic data unchanged across each FOR loop, so the loop makes no difference.
  • The chart is not designed to take a cluster of dynamic data (notice the red coercion dot!). If I create a constant from the chart terminal, I get a cluster of integers for the T chart and a cluster of DBLs for the G. chart)
  • I cannot look at your express VIs, so I have absolutely no idea what's in the dynamic data wires. Maybe it matters?
  • Never hide the label of terminals. Makes the code unreadable. There is no need to maximize the front panel to the screen.
  • You typically should not wire N when autoindexing on a FOR loop.

 

altenbach_0-1676064296944.png

 

0 Kudos
Message 3 of 4
(683 Views)

Maybe you could just wire a cluster of DBLs to define the type better. No idea if that would work. (Better programmers stay away from dynamic data!)

 

altenbach_0-1676064755814.png

 

0 Kudos
Message 4 of 4
(677 Views)