SI7021 - to high temperature values

[parbar]

Dear all,

I have a wemos D1 mini with an I2C Si7021 and SSD1306 oled display connected. The SI7021 shows to high temperatures. About 4 Celsius above actual room temperature. I have tested 2 different SI7021 but the same issue. Anyone experienced the same? Could it be self heating ?

Just installed latest R145, but same result.

oled.jpg

scr1.jpg

Thanks for any help.
Regards,
Pär

[Drum]

Try putting a fan on it, I have seen heating from the Esp. 2 to 3 degrees at 3 cm, and also 2 up to degrees over a couple of feet in height at my desk at work, at home I need to have a ceiling fan on pretty fast to equalize the temperature from floor to top of a bookshelf.
What are you using to measure room temperature? I prefer to use something I can calibrate by boiling, and place it next to the sensor, at least 15 cm from the ESP, and board.

[parbar]

Thanks for the reply!
I have compared the temperature with standard indoor/outdoor termometers. I could understand and accept an devistion of +/- 0.5 degrees or even 1 degree. Now I have 2-4 degrees.
As I have the sensor on short wires I just tested to insulate the ESP unit from the sensor but same result.
I think it is selfheating. I made test and disconnected power to let it cool down. The first minutes the temperature seems okey then it increase about 2-3 degrees.
https://thingspeak.com/channels/21739

The module is also named GY-21. It seems to have a 3.3v regulator on it. Could the voltade regulator couse self heating?

I will investigate further...

[parbar]

Removed the voltage regulator and connected direct to 3.3 v but same issue, about 2-3 degrees to high temperature.
There are other IC and components on the module but I don't think I can remove them.

The wemos it self reach about 40 C and can of course have impact by heating. Next steep will be to insulate the sensor better from the wemos board and see what happens.

[Drum]

If you want to be sure, get a good quality glass thermometer and put the bulb next to the sensor. It needs to be rright next to the sensor. And Check the datasheet for the sensor. If you want that kind of accuracy you cannot assume your indoor/outdoor termometer is that accurate unless it is tracable to a standard. Try something like this:
https://www.amazon.com/Enviro-Safe-B605 ... meter+nist

From what you describe it sounds like you need to make the wires longer....

[parbar]

I extended the wires for the sensor to about 300 mm. Placing the sensor at that distance from the wemos d1 mini solved the problem. Heating from the Wemos increased the temperature readings. Issue solved.

I also have a similar setup with DHT22 and it is of course the same problem.
This make it difficult to do small compact units for temperature measurments. Any ideas how to solve this, it would be great to make compact units. Thinking of using deepslepp between readings but then it feels like the full potential of espeasy is lost...

Thnaks for the help!

Regards
Pär (Sweden)

[Drum]

I consider the biggest potential of ESPEasy to be flexibility and of course it s REALLY EASY.

You could try to use a small compact unit and use a formula to bring the temp down(less accurate), or use deep sleep to keep it from getting hot enough to affect the sensor. In any case, having a housing with a thermal barrier and good ventilation would be helpful.

Do an experiment, measure the temperature at several different heights in a room. I used a DS18b20 on a 1 meter cable. What is the "room temperature"?
It takes a ceiling fan set pretty high to mix enough to get the same value at all heights.

Generally you do not need to take readings continuously, so not a problem.