MH-Z14A CO2 sensor

[alexsh1]

@korttoma values are completely off. Co2 cannot go down below 300 level. This is a value of co2 outside (300-400).

You have to calibrate it regularly. Zero calibration (this is what you are talking about) is different.

8.1 The sensor should be calibrated regularly. The cycle time is better to be no more than 6 months. 8.2 Do not use the sensor in the high dusty environment for long time.
8.3 Please use the sensor with correct power supply.
8.4 Forbidden to cut the sensor pin.

Try to put it outside for 10mins and then short pins 8 and 12 (see the photo below). This worked for me when I had some weird readings

[alexsh1]

@FotoFieber Netatmo is using index also cannot be compared with mh-z14a.

I think the idea putting the sensor outside is to stabilise it - the reading is in the range of 300-400

[korttoma]

@alexsh1 said:

Try to put it outside for 10mins and then short pins 8 and 12 (see the photo below). This worked for me

Where did you get this kind of information from? The MH-Z14.pdg I found online does not mention anything about this.

[alexsh1]

@korttoma It was recommended by an Aliexpress seller who sold me the sensor.

Let's agree that Chinese documentation is not always comprehensive :)

[korttoma]

I finally found some time to really look in to this 562 issue. I looked at the raw data in DataMine2 and I noticed that the 562 value appeared every 10 minutes on the minute.

Then I knew that this value cannot possibly come from the sensor. I search around in my Vera device for a bit and I found some lua code that copied the "Variable1" value from the sensor to the "CurrentLevel" every 10 minutes. Then I remembered that in the previous version of this sensor the value was reported via V_VAR1 and I had to copy it to another variable in Vera to be able to draw any graph. The current version of the sensor reports the value directly to the "CurrentLevel" variable and the value 562 that was still in the old variable was copied over the "real" value every 10 minutes.

I also previously had some really low values from time to time from the sensor mostly when responseHigh = 1 and responseLow = negative value

so I forced the responseLow to always be a positive value like this:

int responseLow = abs((int) response[3]);

I will keep an eye on the behavior of the sensor now that I have finally removed the confusing 562 value.