Time channel

[shefalika]

Alright. I have a time channel in my data set. i am trying to claculate HIC with my data. When i run it, it sometimes crashes on me saying "The X-channel needs to be equidistant". To get around the error I have to reload the data again and start all over again. It seems like the time channel somehow corrupts itself. And this happens very randomly. Any input as to why this is happening and how can this be avoided? Thanks

[Walter_Rick]

Hi,

I never heard something like this and I can't believe that a mathematical function witch is hard coded sometimes delvers corrupted results and sometimes not. So I think the reason we must look for is somewhere else. Please can you send me a step by step instruction or better a script (and the values to be analysed) so that I can try to reproduce this behaviour?
You can send this direct to my email address: walter.rick@ni.com.

Greetings

Walter Rick

[Sarm]

What is happening here is that DIAdem is using all of the data in the time channel to judge if the data points are equidistant. That includes all of the decimal places. If you have an explicit time channel being created by your data acquisition system and then being read by DIAdem it is probably 10 or so decimal places. Since your sample rate is probably from 10 to 20 kHz so you only are interested in 5 decimal places. But DIAdem is using all the info it has available and it is out in the decimal places 9 and 10 that your time channel is not equidistant. You are probably not seeing all of the decimal places on your screen depending on your DATA table attributes.

You might try something like this to work around the problem; depending on what units your time
channel is in (sec, msec, etc.).

Call FormulaCalc("Ch(1):=(Round(Ch(1)*100000))/100000")

This would eliminate the extra decimal places from causing the equidistant error.