Hello everybody.
I'm quite a newbie in Labview (2009 on WindowsXP).
Here is what I want to do :
1) acquire Raw datas continously from DAQmx (typically one digital channel, 500 KS/s)
2) "decode" them on the fly, continously too, and store the datas on hard disk.
3) process the decoded data, like display them, compute for example sliding average, PSDs, PDFs, etc...
I want to have all those tasks to be done inside a program one could build, so it need a good GUI : user must just run the main VI and then select what to do (and not changing values, run the program, stop it and re run it with different values like I usually do).
My first attempt was to do nearly everything in the same VI, using a producer/consummer pattern with a queue to acquire (producer) and decode (consummer). The result is working but clearly it is a nightmare to read and add a functionality.
So I try to make a more readable program.
To do so, I use a lot of multiple subVIs, each one of them dedicated to do a generic task (initialize the GUI, Initialize the DAQ card, etc..) and a main subVI with the Producer/consummer loops inside.
There is a lot of parameters to give to subvis and I want the subVIs to refresh indicators from the main VI, I create references for each cluster or indicators, and pass a big cluster containing them to every subvis.
My question is what is the best way to do my tasks ? Is the design pattern I use last the good way ? Is there any better solutions ?
As you could use state machines, global variables, event structures, etc... It is not clear to me what would be the most effective and readable solutions.
For example, refresh a graph of a main VI from a sub VI using reference and writing inside the Value property node of the reference is a good thing ?
I didn't find any example of large program in labview and I have no idea about effiency of the choices I made at this time...
If someone has made something similar or have some clues to share I would greatly appreciate !!
Thank you for reading my poor english and help me !!
