Some years ago, I developed a protocol to implement an improved characterization of the SPICE diode. In time I intend to publish this protocol but to do so I need to be able to make lab measurements of a diode’s forward voltage drop as a function of its current at three named ambient temperatures. In order to do this laboratory measurement work, I will need a capability to accurately measure the ambient temperature of a device under test. To do this I will need a thermocouple thermometer such as that pictured above and available from Amazon. The above pictured model was so incredibly cheap that I thought it was worth taking a chance on.
Once a device like this arrives, it is necessary to establish its operating parameters. You need to figure out what it is trying to tell you. The way to do this is by performing a calibration procedure.
The means of calibration is to use the two easy to obtain standards that are readily available. The first is the boiling point of water with is approximately 100oC at one atmosphere of pressure (101.4 kPa) at sea level. But the site of the test was to be at 275 meters (900 feet) above sea level so a little correction may be in order. Please refer to the chart provided by EngineeringToolBox.com which gives a rough indication of atmospheric pressure as a function of meters above sea level.
Here it can be seen that at 1,000 meters, an atmospheric pressure is likely to be 100 kPa. The difference in accounting for height above sea level is adjudicated to be insignificant so we will use 100.0oC as the boiling point of water for this exercise.
The other standard readily available to us is the freezing point of water which was taken to be 0.0oC for this exercise.
The premise of this exercise is to bathe our thermocouples in these two standards and allow for any thermal equalization to take place. We will then arrive at indicated values to correct for. For example, if a thermocouple reads 99.3oC in a boiling water bath, we will then have established that its reading of 99.3oC is in fact an indication of 100oC. We will then do likewise using an ice bath to establish a calibration for the other thermal extreme of any one thermocouple.
With two calibration points we will then be positioned to solve for an equation that will return the factual temperature for any indicated temperature. For this we must assume a linear relationship for thermal variations which is justified.
The table to the right documents the above described procedure.
We took three readings for the boiling point. The reason for three points separated by five to ten minutes was to establish the absence of any significant thermal capacitance and that the water had in fact reached a boiling point by means other than visual.
Although not visible in the table, thermocouple number 2 was giving irregular readings for the ice bath and so it was necessary to fully disqualify that thermocouple for all use.
To evaluate these results, our primary objective is to establish an offset. At a boiling temperature, any one thermocouple should read 100oC. There will be an offset for each thermocouple. On average, the offset for each thermocouple was about 3.5oC which is excessive. The factory specifications for the instrument say that error should be less than ±2oC. Just the specification itself is quite loose. A reasonable thermocouple instrument should do much better than ±1oC. But this device was considerably outside of even its factory specifications. But still, with the calibration that is being accomplished here, a correction equation will be developed to allow it to provide reasonable results. The table also records temperatures taken at an unknown temperature. The purpose of this measurement is to establish a relationship between the entire group of thermocouples. We may observe that there was roughly 1oC offset across the board for the entire group. While the temperature was unknown, there was nevertheless a temperature identical for each thermocouple.
The table on the right lists the results showing linear interpolation equations explicit for each thermocouple. These equations can be used to correct for any readings obtained from them. For example, suppose that thermocouple #1 returned a measurement of 38.2oC. The actual temperature after correction will be 37.5C.
Note, however, that #2 has been stricken. This thermocouple was found to be reporting irregular results so the best choice is to not use it at all and that for all time.
CONCLUSION:
Don’t buy this thermocouple instrument unless you are only interested in extremely rudimentary results. I’m going to buy something else and, time permitting, I will compare these results.
To say the least, I was very much disappointed that I had to fully disqualify thermocouple #2. I must place it permanently out of service.
This instrument has no capability for the user to set any offset. The result is that you simply have to take what you can get which is not a good thing.