02-06-2022 07:58 PM
Dear Sir:
I purchased a TC01 and I recently came across this NI article:
USB-600x Accurate Thermocouple or Low Voltage Measurements - NI
Can an NI engineer clearly explain the point the author was trying to make when he wrote:
" These resolutions would mean that these devices can only detect changes of approximately 23-29 °C. "
Please explain what he had meant by the above statement in terms of the accuracy of a measurement taken within and outside of the temperature ranges set forth above. If a specific thermocouple is required for the explanation, assume that I am using a K type thermocouple.
Thank you for your consideration.
Solved! Go to Solution.
02-06-2022 11:39 PM
Since you purchased TC01, this article is not applicable. This article explains why you need to get a TC01 instead of using the low cost 600x series DAQ for Thermocouple measurements.
In short, the ADC on 600x are lower-resolution 14/16-bit and fixed higher input range. Typically thermocouple signals are in the order of mV. Absolute accuracy of the 600x DAQ is already in the range of mV and hence, most times the Thermocouple signal changes are not picked up properly and typically it would require 23-29C change on the thermocouple for the DAQ to pick up the voltage change.
02-07-2022 01:12 PM
With all due respect, Santo, I can see that you took some time to re-state what was already asserted in the article.
Unfortunately, that type of response does not add to clarity.
"These resolutions would mean that these devices can only detect changes of approximately 23-29 °C"
Verification - clarification - of the above statement is needed.
The NI 9211 and the USB TC01 appear to be 2 different devices and may, possibly, have 2 different circuits - leading one to believe that they may 2 different performance specifications. The author of the webpage bundles these 2 devices into one category and references a temperature range.
"These resolutions would mean that these devices can only detect changes of approximately 23-29 °C"
USB-TC01 Specifications - Temperature Input Device Manual - National Instruments (ni.com)
NI 9211 (24 bits of resolution).
The person who wrote this article should have let the reader know how he came up with the above statement, instead, of having the reader look at data sheets and try to figure out the meaning of his statement. Technology articles written by any author should be intended to clearly convey their points, not leave it up for the reader to research what point(s) they were trying to express. "Vagueness" defeats the whole point of writing the article.
Let's say I theorized what the author wrote and I could create equations which appear to be consistent with what he had written. At best, I am "theorizing" what the author had meant. This is not verification with any "certainty" of what the author had meant and never will be. However, I am willing to settle for a mathematically supported theory at this point.
I thought there would be simple answer, I guess I am going to have to go to the data sheet and obtain my answer.
Thank you for your effort anyway. I thought I could come into "contact" with a specialist at NI and immediately obtain an "instant" answer. This does not appear to be the case. To the data sheet(s) I go.
02-07-2022 02:53 PM
Let me take your example, voltage response of K type: 41 uV/°C
This means, for every deg C temperature the voltage generated by the thermocouple is 41uV.
Now, coming to the resolution part, the resolution of analog input is the smallest incremental change it can detect. Now, you know that the K-type TC will increase the output voltage by 41uV if the temperature increases by 1 degC. If the analog input has to detect this as a change, its resolution has to be at least 41uV or smaller.
For example, a 14-bit analog digital conversion of a signal ranging from +/- 10 V provides a resolution of around 1.22 mV (voltage range/2^bits = 20/2^14).
Now, this 14-bit ADC can only detect input changes at every 1.22mV increment, anything smaller than this 1.22mV change, it cannot detect.
Now, to know how much degC is this 1.22mV resolution is, 1.22e-3/41e-6 = 29.75 deg C. Thus the approximation of 23-29 degC is calculated.
02-07-2022 04:23 PM
Thank you, Santhosh (santo_13), for your most excellent reply!
Kudos for you, (hopefully, that "Kudos" button will be available to me after I post this message)!
08-05-2022 04:07 AM
Can I suggest that you use Type N rather than type K. Better in almost every respect, especialy resistance to long-term issues, but not much more expensive.
Obviously, if you want long-term high temperature stability (>3 months at >1000°C) you need the more expensive type R.