First determine what is taking the time using profiling, or by just unwiring
the graph. You can measure your loop rate and watch your backlog. In another
VI, you can hook the graph to arrays of data that are constants or simulated
and don't involve the DAQ calls.
With these separated out, you know whether you should focus on the DAQ
or the
graph. If it is the graph, then as stated, it could have to do with the
array building on the diagram, but I doubt it. The update rate of a graph
has lots to do with the size of the graph, the depth of the monitor, the
number of points being plotted, and the style of the plots.
Using the simultated data, vary these parameters and see how the speed changes.
You can tune it to where a graph updates hunderds of times a second or
less than
once. One of the most surprising speed hogs is the video depth. By
running the
computer in 16 bit or perhaps even 8 bit mode, you can often double or quadruple
the update rate compared to true color depth.
Greg McKaskle
>
> XY Graphs require four 1-D arrays for two plots. A chart uses a 2-D aray
> that is about half the size. Arrays require construction and construction
> requires CPU and memory. There are "good" and "bad" ways of building and
> manipilating arrays. This is most likely the source of your woes.
>
> >>I'm a fairly new user to Labview but have a good basic Knowledge of general>programming
> requirements. My problem is this. We are developing some software>using
> Labview running a DataTranslation DT2811 board and A ComputerBoards>CIO QUAD
> 4 counter board simultaneously. I am taking the data from both>boards and
> writing to file and also producing some on-screen plots in real-time.> My
> problems are that in using an XY graph to behave as an XY chart using>buffers
> the logging rates are significantly reduced to an unacceptable level.> Is
> there an alternative way of plotting the data from the ComputerBoard>against
> the data from the DataTranslation board that will not compromise>the logging
> rate. >>Any help would be appreciated>>Thanks in advance>>Richard