Data Socket: More data loss in LabWindows/CVI than in LabVIEW?

[YSKIMO57]

Hello.

My code for data socket transmission in LabWindows/CVI is having problem of data mismatching. I'm testing with sine/cosine curves and set-up with 'WriteAutoUpdate' and 'ReadAutoUpdate'.

It's ok for low rate (10Hz with 20 double values/patch). However, for higher rate, say 50Hz, the curve distorted with jagged curves. It actually reads the same value for several loops. Looks the loop itself is fast enough, but accessing is overlapping or mismatching.

I acknowledge it can happen in data socket application, especially for higher update rate, but, I think, 50Hz is not supposed to be the expected data socket performance. Also, the thing is - when I read it with LabVIEW (not LabWindows/CVI), it works ok. The transmission rate can be as much higher as 500Hz!

Wondering what makes LabVIEW faster than LabWindows/CVI.
Or would there be anything wrong with my code?

Expecting a guru's advice...

Thanks.

[Scott Y]

Hello,

Have you tried including a timer in your loops so that you can get a rough idea of how often the writer and the reader are actually running?

If they appear to be running at similar rates, then I would like to ask you to post your code, and a note about which LabVIEW examples you are using for the LabVIEW side of the testing.

Scott Y
NI