Looping through parallel model starts at index 1 instead of 0 on second execution

[raphael.wittwer]

Hello forum

I have an issue with looping through the UUT loop. Even though I configured the loop to start at 0 it starts at 1 from the second time I start the test without stopping and restarting the sequence. I attached the modification to the process model as a screenshot.

Thanks for any help!

Raphael

 

[CyGa]

What is you purpose exactly ?

Test only 3 UUT on a socket, and then stop the socket execution ?

[raphael.wittwer]

yes, loop three times then stop and wait for the next start command from the GUI.

[CyGa · Accepted solution]

Why do you keep the counter in a SG ?

This loop is for a specific socket to run as much as needed. If you keep a counter as normal, but use a variable (local) to set the max number of loops I guess you won't have any problems.

From what I see, it should work, but keeping the counter in a SG sounds weird to me.

[raphael.wittwer]

The idea behind the station global was that after a crash the sequence would start with the right batch, but as it is now I don't need this functionality. I will try a normal local variable and see what happens.

Thanks!

[CyGa]

HI Raphael,

 just ran into this few days ago : in case of a crash (meaning complete app crash or PC being unplugged for example) there is a chance that your SG won't be saved !

Indeed SG are saved whne the Engine shuts down or if you call a specific Commit method.

See here : https://www.ni.com/docs/en-US/bundle/teststand/page/station-global-variables.html

[raphael.wittwer]

I tried locals for this counter and propagated it through the sequences, unfortunately this has the side effect that the sequence starts running without user input. I undid the change since this seemingly breaks other things.

I will need to find a point to reset this counter since the loop does not work as I expected, probably right after the GUI is called.

[raphael.wittwer]

I tried today reinitialising the for loop after the GUI gets called, unfortunately this does not change anything. The "first" iteration still is 1 instead of 0 on the second pass. On top of that I noticed that the loop counter goes up to 4, which should be impossible with the parameters from above.

 

[raphael.wittwer]

the solution was to use Locals.xy and propagate them into and from the sequences. What was missing the first time I tried was the parallel model was still using the state from the station global instead of the local which led to unexpected behaviour.

Thanks for your help!

[CyGa]

I didn't do much !

Glad you found a solution ! 🙂