Clock drift on a cRIO ... getting way more points than expected?

[josborne]

I discovered an odd issue today.  I have a really simple LV RT application that is running on a cRIO 9040.  It captures AI voltages at 20kHz. 

I noticed that I was getting a lot more samples than I expected.  It seems like either the clock on my CRIO or my system clock is really inaccurate.  Or I am just doing something wrong.  

I suspect it is the cRIO because when I sample some OPC tags, they match the count on my Windows clock.  

Below is a screenshot of my code.  And I attached my code, too.

Can someone explain this to me?  I should be getting 20,000 points a second, but it looks like I am getting a bit more than that?

[GerdW]

Hi j,

Which DAQ module do you use?

[josborne]

It’s a NI 9232. Going to use it for some remote vibration measurements. 

[mcduff · Accepted solution]

The 9232 is a DSA device. It can only sample at specific divisors of its internal master time base, 13.1072 MHz.

I am guessing that it is sampling at 20.48kSa/s. The formula is

[josborne]

Oh wow that is totally it!  Thank you!  

[GerdW]

Hi j,

---

@josborne wrote:

Oh wow that is totally it!

---

Yes, reading the specs/datasheet often helps... 🙂