To gage or not to gage...that is the question!

[Brad Gover]

Hi all! I have a question on picking an instrument for a study. On of the requirements of the study is to measure airflow given a supply pressure of 3 inH2O (inch columns of water). The instrument I have (Dwyer Series 476A) specifies its Accuracy to be +/- 1.5% full scale. The scale reads from -20 to 20" w.c. (meaning inches of water column). The question I have is validating the required pressure of 3. Using this gage if I get a reading of 3, my uncertainty is between 2.4 and 3.6. There is no tolerance associated with the specified value of 3. In this case would the "rule of Tens" apply? If so, then this gage would not be accurate enough. Are there other ways of assessing the validity of this gage?

Really appriciate the feedback!
Brad

 

[dgriffith]

Hi all! I have a question on picking an instrument for a study. On of the requirements of the study is to measure airflow given a supply pressure of 3 inH2O (inch columns of water). The instrument I have (Dwyer Series 476A) specifies its Accuracy to be +/- 1.5% full scale. The scale reads from -20 to 20" w.c. (meaning inches of water column). The question I have is validating the required pressure of 3. Using this gage if I get a reading of 3, my uncertainty is between 2.4 and 3.6. There is no tolerance associated with the specified value of 3. In this case would the "rule of Tens" apply? If so, then this gage would not be accurate enough. Are there other ways of assessing the validity of this gage?

Really appriciate the feedback!
Brad

Check with the manufacturer of the gauge, or check the spec sheet. Many instruments are single-sided spec'd. Many are not. So you think the tolerance is 1.5% of 40, when it may be 20. Honeywell PPT transducers are single-sided--a differential spec of ?20" wc is toleranced at the derated 20" wc only, for example.
Taking a look at the somewhat Spartan specification sheet, I would go with 20 FS, even though the span/range is 40. Otherwise, call or email them.

There is a tolerance associated with the specified value of 3, unfortunately it is 1.5% of FS.
Accounting for the basic contributors (specification and 2 digits of precision/resolution at 3"), the uncertainty at 3" w.c., with a 95% confidence level, is 0.294" w.c. @60?F.