Getting accurate temperature reading from sensors

[I14R10]

Hi

I've been playing with MCP9808 temperature sensor. In the datasheet it says its accuracy is under +/- 0.25 degrees C, and can be around +/-0.1, but It doesn't seem that way to me. I made two identical boards, each using one 9808 sensor, and compared it to my weather station reading. Now, my weather station is also not that accurate, it's WS-2300 model which costs about 70 euros for temperature and humidity sensing unit. But the problem is that one 9808 reads on average 0.5-0.7 degrees higher than my weather station, and the other reads about the same, or slightly lower. From the datasheet I expected them to be a lot closer. What I got from these two 9808 is that they can be around 0.7-0.8 degrees apart on average. And it shouldn't be that large difference between different chip readings.

I'm also looking at TMP117 sensor, which too claims that it has accuracy around 0.1 degree C, and it is quite more expensive than MCP9808, which sounds good.

So my question is, can I trust datasheets at all? Is it better to buy one relatively expensive sensor, or buy 5-10 cheaper ones, like 9808 and average their results? Do you have any sensor that you would like to recommend?

 

[I14R10]

Testo is a pretty well respected name for measuring instruments (www.testo.com).

What do they say, what's the accuracy of their sensor?


Also, you should use better resolution on 9808. 0.25 degrees is very large.

 

[Andy_Sweden]

> What do they say, what's the accuracy of their sensor?

+/- 0.5 degrees Celsius but they don't write if that's "typical" or "max".

> Also, you should use better resolution on 9808. 0.25 degrees is very large.

I did that on purpose. I tried 0.125 and even 0.0625 which worked but then I went back to the default of 0.25 degrees. As long as the MCP9808 was 1.0 degrees away from the other sensors I didn't see a reason to show the glaring problem with more decimal points.

 

[KlausST]

Hi,

Just to verify that this is not an ambient problem:
Interchange the location of the most extreme sensors ... and see how temperatures behave then.

I didn't read the datasheets ... but there are sensors that have built in heaters to verify their operation. Be sure the heaters are switched OFF.

Klaus

 

[I14R10]

In my opinion, this seems like a systematic error. How is your pcb layout? Did you put 9808 on an "island" far away from other components that may heat it?

 

[stenzer]

Hi,

one possible canditade would be the ADT7422 [1]. It even states an accuracy of ±0.1°C from 25°C to 50°C at 3.0 V. The datasheet also shows statistical results (histogram) and serveral measurement series within one plot.

[1] https://www.analog.com/media/en/technical-documentation/data-sheets/ADT7422.pdf

BR

 

[Andy_Sweden]

> I didn't read the datasheets ... but there are sensors that have
> built in heaters to verify their operation.
> Be sure the heaters are switched OFF.

No, there is no heater.

> In my opinion, this seems like a systematic error.
> How is your pcb layout? Did you put 9808 on an "island" far away
> from other components that may heat it?

The data does not support your opinion.
The sensor is on a separate breakout board from Adafruit with nothing else on it. I didn't design my own pcb and I didn't solder it. It's a commercial product.
And the "heat issue" is not the issue. The sensor is showing temperatures LOWER than all the other sensors.

 

[I14R10]

And the "heat issue" is not the issue. The sensor is showing temperatures LOWER than all the other sensors.

Ah yes, sorry, my brain wasn't working properly.

 

[MDR]

Hello,
I bought 20 MCP9808 modules and the accuracy is always near the 0,5 'C. If the accuracy is not your first priority, the module do the job for the price.
Also, the factory calibration can be easily lost if you make a programming error on configuration register.
The best way to have accuracy, is to add a personal calibration "value" by software.

I bought also 3 SHT31 sensors, the 3 mode give exactly the same temperature (±0.1 'C) and the same humidity (±1 %).

I think that the methodology of factory calibration of these modules must be updated.

 

[simon mugo]

Hi

I've been playing with MCP9808 temperature sensor. In the datasheet it says its accuracy is under +/- 0.25 degrees C, and can be around +/-0.1, but It doesn't seem that way to me. I made two identical boards, each using one 9808 sensor, and compared it to my weather station reading. Now, my weather station is also not that accurate, it's WS-2300 model which costs about 70 euros for temperature and humidity sensing unit. But the problem is that one 9808 reads on average 0.5-0.7 degrees higher than my weather station, and the other reads about the same, or slightly lower. From the datasheet I expected them to be a lot closer. What I got from these two 9808 is that they can be around 0.7-0.8 degrees apart on average. And it shouldn't be that large difference between different chip readings.

I'm also looking at TMP117 sensor, which too claims that it has accuracy around 0.1 degree C, and it is quite more expensive than MCP9808, which sounds good.

So my question is, can I trust datasheets at all? Is it better to buy one relatively expensive sensor, or buy 5-10 cheaper ones, like 9808 and average their results? Do you have any sensor that you would like to recommend?

Datasheet specifications are always accurate in most cases and you need to trust them. The temperature sensor above description are usable. I have been reading extra information about temperature sensor here https://www.pcbway.com/project/share/Temperature_sensor_6.html