Accuracy Calculation

In case of no verifiable method Stacking should be applied to accuracy, means it should be 0.15+0.05= 0.20

and range should be -40 to 300⁰C

But best way is contact to manufacturer,

Any need to add the resolution also?

That means have to add 0.15+0.05+0.01.

Indicator resolution 0.01 is already factored in its accuracy of 0.05, So I Think its not needed,

Consider I done the a measurement of 300°C.

The probe having an error 0.15%. Error due to probe is 0.45°C.

Error due to the indicator is 0.15°C.

But there will be another error which have to be taken in account. Which is the error by indicator of the value due to the error of probe.

So 0.05% of 0.45°C have to add in this case, which is 0.000225.

So the total error is 0.45° + 0.15°C + 0.000225°C = 0.600225°C

Assume the error of probe is 'x' and the error of indicator is 'y',

Then we can generalize the error as x + y + xy

What you think about this ?

First of all, as many, too many, you make an error when you write about "accuracy". It is a value which cannot be measured it is only obtained by deduction. In fact what you can measure is the "not-accuracy" i.e. the "uncertainty" of your measurements. Using "accuracy" as term is an error since the ONLY aspect you can quantify is the made ERROR. Now the uncertainty of your measurement depends on what was understood by the manufacturer when he indicated the possible errors of probe and display Those errors are variable so that their distribution can be normal or not. The "probable error " will be the "probable sum" of the 2 effects. If the distribution is normal then the probable sum is square root of the sum of squared errors. An other aspect is the resolution of measurements. It is no reason to measure with a resolution of 0.01° and have an error of 0.158°. It will be better you either mask the second digit after comma or at least record only with a digit after comma.

>Probe accuracy is 0.15°C and indicator accuracy is 0.05

Ummm, there are no units specified for the indicator uncertainty.

Is the indicator uncertainty 0.05°C, or 0.05% full scale?

0.05% of a 1572° span is 0.8°C, an order of magnitude greater than 0.05°C

A claim of 0.05°C uncertainty in a thermocouple indicator that by design must incorporate cold junction compensation is incredibly optimistic; it sounds like it belongs in a laboratory. I'm skeptical.