What is the best practise to use temperature sensor IC's inside an enclosure?

Thread Starter

zazas321

Joined Nov 29, 2015
715
Hello. I am going to test various temperature sensors including:
DTH22
DS18B20
LM35
BM180
TSYS01

With the sensors such as DTH22 and DS18B20 (waterproof version). Usually I build an enclosure and have the sensor sticking out of enclosure.

However, what is the best practise to use sensors such as BM180 and TSYS01 inside an enclosure? If I plan to use evaluation boards of those sensors, do I need to make sure they are sticking out to the outside of the enclosure as well? I am worried that the microcontroller ( I plan to use ESP32) is going to warm up slightly and affect the temperature reading.

I am interested in reading both indoors and outdoors temperature.
 

Delta prime

Joined Nov 15, 2019
561
Hello there
:)
Testing for certification, thermal dynamics,
Stringent and unwavering parameters must be ahered to. My expertise in this science is limited to quantifying thermal parameters by measuring some physical property of a working material that varies with temperature.

They are then calibrated either by comparing their results with that of certified thermometers, or by checking them against known fixed points on the temperature scale.
I cannot work with inside and outside of a box can you elaborate on how and what you are testing, or if you are employing any one of these test methods. Then it would be my pleasure to help you.

Standard Test Method for Inspection and Verification of Thermometers

Standard Test Methods for Radiation Thermometers (Single Waveband Type)

Standard Test Methods for Testing Industrial Resistance Thermometers

Standard Specification for Compacted Mineral-Insulated, Metal-Sheathed Cable Used in Industrial Resistance Thermometers

Standard Test Method for Calibration of Refractory Metal Thermocouples Using a Radiation Thermometer

Standard Specification for Temperature-Electromotive Force (emf) Tables for Standardized Thermocouples.

Standard Test Method for Evaluating Thermal EMF Properties of Base-Metal Thermocouple Connectors.

Standard Test Methods for Sheathed Thermocouples and Sheathed Thermocouple Cable.

Standard Practice for Calibration and Use of Thermocouple Reference Junction Probes in Evaluation of Electronic Reference Junction Compensation Circuits .

And I can Guide you through Thermocouple Verification
 

Phil-S

Joined Dec 4, 2015
158
What do you mean by testing them? Do they work? Are they accurate? Are they precise?.
As Delta Prime has indicated, temperature measurement is a complex business.
Often the sensor itself will change the temperature of whatever you you are trying measure.
Of course if a sensor is on a board with other components, it will be affected by their heat output.
The sensor will measure the temperature of whatever is in closest contact with the active sensing component which might be a speck on an IC die.
A DS18B20 in a metal sheath and a heavy cable will be considerably affected by the sheath and cable.
Some sensors by way of their construction may not be suitable for external use without protection.
An example of what I use.
To measure the temperature of the water in a copper pipe (heating system), I use a TO220 version of the LM35.
It has a large metal tab. I cut a copper solder joint in half lengthwise and that has a large area of contact with the pipe.
I solder a brass screw onto the joint and use brass washers and nuts to ensure the best thermal contact.
Of course it won't be accurate to 0.1 degree C, but it is robust and is adequate.
Alternatively, I could drill a hole in the pipe and put the tip of an NTC thermistor so that it contacts the water and the copper pipe, but that isn't practical
Generally, you have to go with what is acceptable to you.
As ever, go through the data sheets and application notes of the sensors to see if there are any implementation guides
 

Reloadron

Joined Jan 15, 2015
5,667
Hello. I am going to test various temperature sensors including:
DTH22
DS18B20
LM35
BM180
TSYS01

With the sensors such as DTH22 and DS18B20 (waterproof version). Usually I build an enclosure and have the sensor sticking out of enclosure.

However, what is the best practise to use sensors such as BM180 and TSYS01 inside an enclosure? If I plan to use evaluation boards of those sensors, do I need to make sure they are sticking out to the outside of the enclosure as well? I am worried that the microcontroller ( I plan to use ESP32) is going to warm up slightly and affect the temperature reading.

I am interested in reading both indoors and outdoors temperature.
I believe you mean a DHT 22 as I never heard of a DTH 22? The DHT 22 includes humidity measurement.

When you test a sensor you generally compare your sensor (unit under test) to a known standard. Then you note any deviation. You need a known such as a precision standard. You also develop a test setup making sure nothing is having an adverse effect on your measurement plane. For example I might place a standard precision thermometer beside my DHT 22 in a common thermal environment and compare the readings. The standard and your temperature sensor need to share a common thermal environment free of any outside influences. You also want to test your sensor over a range of temperatures. What you need is a defined procedure and method.

Ron
 
A Platinum Resistance Thermometer (PTR) is a good reference sensor.
A PT100 sensor has a defined resistance of 100 Ohms at zero degrees Celsius.
The resistance of the platinum element will vary with temperature and follows published standard tables.
There are varieties from basic two wire types to four wire types that can compensate for the additional resistance of the wiring.
You need to know the difference between precision and accuracy - they are not the same property.
Just for fun, you could look at how standards laboratories like the National Physical Laboratory do their temperature measurements.
 

Reloadron

Joined Jan 15, 2015
5,667
You need to know the difference between precision and accuracy - they are not the same property.
I see accuracy and precision as follows:
Could we say a good analogy for understanding accuracy and precision is to imagine a shooter shooting a target. If the shooter shoots with accuracy, his aim will always take the bullets close to or into the center of the target. If the shooter shoots with precision, his aim will always take the bullets to the same location which may or may not be close to the center of the target. A good shooter will be both accurate and precise by shooting the bullets the same way each time and each time making it in the center of the target. If we look at it this way precision is a high measure of repeatability and accuracy is then unbiased precision, even as the terms apply to shooting a target. Would that hold true?

Accuracy and Precision.png

Ron
 

Attachments

Yes, the analogy is right.
I used to be involved in quality control and know how difficult it can be to get your head around it.
In the UK, we use the dartboard analogy.
A good darts player will always be going for the bullseye.
He/she might always get close around the bullseye, but the throws are distributed in or around the centre.
He/she is reasonably accurate, but their precision is poor.
He/she tries for the bullseye but always gets the triple twenty. They are precise, but not accurate.
There might be something about the way they throw, the darts might be a bad batch or their glasses are wrong.
The champion darts player achieves accuracy and precision and gets a a large cheque and a trophy.
In analytical quality control, you include blank and standard samples.
You aim to get precision and accuracy in the analytical method such that you can report results with specified degrees of confidence.
The same applies to testing temperature sensors..
The ideal sensor would be very accurate and very precise, but in practice, just hitting the dartboard can be an achievement.
 
Top