03-09-2015 12:02 PM
I just started working for a company as a test developer and they seem to be using CVI along with TestStand. They also have there own test software that uses TestStand and CVI together along with some SAP integration. I am using the software to test some computer that is connected to a test station via a serial port. The serial port monitos the activities of the uut and I am able to read the timestamps from the logs
The problem that I am facing at this moment is that the timestamps from my computer are different compared to the timestamps from other peoples computers. I installed TestStand and CVI runtime engine on my own and I have found that different people have different versions of the runtime engine. I tried installing various versions of the runtime engine but none of them seem to give me the right time stamps. Basically the amount of time taken for my sequence to generate timestamps is more than twice the time taken by other people.
If there is someone out there that can help me out, I won't be able to thank you enough in words.
03-09-2015 04:37 PM
What kind of time discrepancies are you dealing with ? ms or more than that ? Also you may want to read about the SLEEP_MORE (and similar) option of CVI.
03-11-2015 12:07 PM - edited 03-11-2015 12:08 PM
If the discreptancies are small, there could be a few things throwing you off:
1. CVI's Delay() function will always delay for at least the amount of time passed to it, but can (and will often) delay slightly longer than the passed delay time.
2. Timers on the user interface are not gaurenteed to fire at the rate they are set up for. They run in the same thread as the UI and can be blocked by function calls, other timers, and other events on the UI such as having a menu open.
3. All Asyncronous timers using the asyncimr.fp instrument run in the same thread. If you have more than one async timer they can interfere with each other.
03-11-2015 02:29 PM
The discrepancies are "LARGE" in seconds. Everyone else can run their test in about 7 seconds and it is considered a PASS. While my machine takes 15-18 seconds to run the test and it is considered FAIL. I have narrowed the problem down to software by switch the uut as well as the test equipment. Further I have copied every file from an existing install folder from a "FAST" computer and replaced the files on my "SLOW" computer. Yet the slow computer remains slow. Upon swapping the "FAST" files with the "SLOW" files the fast computer still remained fast. I copied all the dll files from the system32 folder as well. There is something that I seem to be missing over here and since I am new to this I have no clue what to do.