Anyone else had problems with the temperature calibration on the MKR ENV shield? Version I have is about 6degC too high. Its sitting on a MKR WiFi 1010. I know the board might give off some heat, but 6deg seems a lot to be out by.
David
Put it in front of a desk fan.
If the temperature goes down then you know it's being heated by the surrounding electronics (or its own electronics).
That is like asking what color Martha's dress is. Since it has an on board regulator and a few other goodies that consume power the heat will vary with the voltage supplied which is then modified by the connected loads such as LEDs etc.
mikb55's suggestion with the fan is a good start. My guess it is the board but that test will verify it. If it is real consistent simply subtract it as a pseudo offset or reduce the power consumption.
Good Luck & Have Fun!
Gil
same problem here, with the same config (MKR Env Shield on the top of an MKR1010 Wifi). I got a ~ 5C higher reading than the actual temperature.
It was easier on my project so I'm compensating the wrong telemetry with an external dht22 sensor. Putting a temp sensor on a temp sensor right? 
I didnt try to move the shield away from the MKR1010 though (on a breadboard for example) so I dont know if the Env Shield itself is flawed from this perspective or not.
It's strange that I could'nt find any info/forum/feedback about this problem, because it's so obvious (small circuitries are heating up each other) maybe no one wanted to complain or just dont care if their readings precise or not.
For me 5C is a lot.
hexyking:
same problem here, with the same config (MKR Env Shield on the top of an MKR1010 Wifi). I got a ~ 5C higher reading than the actual temperature.
It was easier on my project so I'm compensating the wrong telemetry with an external dht22 sensor. Putting a temp sensor on a temp sensor right?
I didnt try to move the shield away from the MKR1010 though (on a breadboard for example) so I dont know if the Env Shield itself is flawed from this perspective or not.It's strange that I could'nt find any info/forum/feedback about this problem, because it's so obvious (small circuitries are heating up each other) maybe no one wanted to complain or just dont care if their readings precise or not.
For me 5C is a lot.
That's was exactly my thinking - it seemed odd that there was no mention of people needing to offset the reading. The sensor claims a high level of accuracy.
davidfmackenzie:
The sensor claims a high level of accuracy.
If the temperature is 5 C higher than room temperature at the sensor due to the heat generated by the Arduino board then it is only right that the sensor should read a temperature of 5 C over room temperature. To do anything else would be inaccurate.
pert:
If the temperature is 5 C higher than room temperature at the sensor due to the heat generated by the Arduino board then it is only right that the sensor should read a temperature of 5 C over room temperature. To do anything else would be inaccurate.
Yes, but I would have expected some directions about a calibration process to ensure that the environmental factors created by the hardware the sensor is installed can be allowed for. For example exposing it to two different known temperatures to allow an interpolation curve to be created. No point in the sensor being highly accurate if there are no real world scenarios where it reflects the temperature properly.
I am going to revive this post as I have the same issue and require some advice
I have two MKR1010 running, one with a DHT11 the other with a MKRENV, I am getting the following reading
MKRENV Temp 24
DHT11 Temp 22.3
Room Temp according to other system 21.7
MKRENV Humidity 32%
DHT11 Humidity 48%
Room Humidity according to other system 31%
does anyone have any suggestions?
It's all the same to me. It seems very strange because this is an "official" package, so I expected more attention on this or at least a warning about it.
somebody find a way to solve it?
Temperature and humidity are totally wrong (+5C T and -10% H)
Thank you