FieldPoint Family

cancel
Showing results for 
Search instead for 
Did you mean: 

Confidence Interval in cfp-tc

Solved!
Go to solution

Hi there,

 

We measure temperature with thermocouples which are connected to a cFP-TC-125. I'm interested in the measurement accuracy of this unit. Of course I know the diagrams in the manual, which contain different measurement error curves for different measurement situations. For a specific ambient temperature and thermocouple, there are two curves. One does show a so called "Max" and the other one a "Typ" measurement error. I understood the difference, however we want to use the information to calculate the total measurement error (which additionally contains the error of the thermocouple sensor and the error from averaging different measurement points). Well, I'm getting to the point: If you totalize different measurement errors you need to use the same confidence intervall, because this is scaling the measurement error.

 

The question is, if someone could tell me to which confidence intervall does "Max" and "Typ" belong. Is that 1sigma, 2sigma, 3sigma? In a first approach i would think that "Typ" is 1sigma and "Max" is 3sigma. But, assuming a normal or a uniform distribution, this doesn't seem to fit to the given curves.

 

Thanks for reading, regards

Gerald

0 Kudos
Message 1 of 4
(6,606 Views)
Solution
Accepted by topic author GBesch

Hi,

 

I have been doing some research and I can answer your question:

The figures related to the errors for different thermocouples uses the values for Error/range from page 15 of the same manual to calculate typical (15-35 °C) and maximum (40, 70 °C) error.

And I confirmed that these max errors stated in the manual are taken at the 99% confidence interval, the very worst case scenario readings.

I hope this answer your question.

Have a nice day  

Message 2 of 4
(6,577 Views)

Hi Carmen,

 

Thank you for your reply!

 

That was the information I was looking for. I just wondered how to treat the error from this instrument in complex measurement, because you have to bring all errors to the same level...so your answer clearly helps me to treat the error correctly.

 

Have a nice day!

Gerald

0 Kudos
Message 3 of 4
(6,573 Views)

I am glad to help!

Have a nice day too

0 Kudos
Message 4 of 4
(6,569 Views)