Hi,

there are several sensors that require to set up amplifier gain & load resistor in order to get correct output values.

Those gain & resistors values should be calculated after the sensors are calibrated. Calibrating is not simple, apart from expensive, which makes a bit difficult to do it while internal developing is done.

Do somebody know some empirical method to estimate gain & resistor without calibrating the sensors? In particular, I would need to do it for CO2 sensor (only gain is needed) and CO & NO2 sensors (both gain & resistor are needed).

According to a different post, gain of 10 is more or less valid for CO2 sensor.
Similar numbers for NO & CO sensors?

Thanks

Anybody working with sensor boards could help me on this calibrating issue?

Thanks

Hi edpi,

Calibration process is recommended to determine the output voltage range in the concentration range that you need to measure. In order to enable the possibility of adjusting it to the Waspmote input a variable load resistance and a variable gain stage have been introduced. But to configure them properly you will need to know the voltage range output by the sensor before.

If you can't calibrate, then I'd recommend you to set the load resistance to the minimum specified in the sensor data sheet and the gain of the stage at 1, and modify it after the first tests, if needed.

The case of the CO2 sensor is different because it outputs a continuous voltage that decreases when the gas concentration rises. You can measure the output voltage of the sensor at the environmental concentration (about 350ppm) and, if it is going to be your reference, fix the gain so that the output voltage is as near as possible to the 3.3V that the Waspmote admits at its input (this gain will be between 7 and 10 approximately).

Regards.

Thanks for the answers.

We know calibration is required, but we would like to start working with the sensors before performing such calibration.

Couple of questions:

- Do you have an estimation of minimum load resistor for MICS-2710 NO2 sensor? we haven't found it in datasheet.

- Do you know any good provider in Spain for "pattern gas" (not sure if this is the correct term in English for "gas patrón") and other equipment required for calibration?

Regards

Hi edpi,

you are right about the NO2 sensor, there is no information about the minimum load resistance in the data sheet. We have tested it using a load resistance of 3kOhm and got good results. Take into account, when selecting this resistance, that this is a different case, because the resistance of the sensor increases instead of decreasing with the gas concentration.

Regarding the pattern gas, we do not have this information, we calibrated our test sensors in a laboratory at the University of Zaragoza, and the gases that we used where provided by them.

Regards.

Hi sayantanib,

1) The reason to adjust the output to 3.3V is to get the biggest resolution possible. If you want to measure concentrations below those 350ppm you should fix the output at that value to a lower voltage, or you will miss that part of the range.

2) If the measurements have been taken at normal air their output is out of range, below the voltage it should be at. Have you tried it under different concentrations?

Regards.

What I suggest you do is to take your sensor to an open area (outside). and take a few results there, and observe the output. Once you get a stable output (Within the sensitivity levels) then you can make that your CO2 at 350 ppm. Then as libelium-dev said, amplify your output until you get 3.29999 Volts, and from there on, you know that if you expose your sensor to high concentrations of CO2 then your output will decrease.

Hi sayantanib,

answering your questions:

1) you won't damage the sensor, it is just an amplification stage prior to the analog to digital conversion. You can adjust the gain by the procedure AmroQuandour indicates before.

2) the output is slightly out of the typical range, but we won't know if its just a small deviation from the normal values or a malfunction of the sensor until you perform some test.

Regards.

Hi sayantanib,

it is not necessary such a complex process as in calibration, it will be enough with putting the board with the sensor in a small closed space (we used a bowl upside down) and burn something inside (a couple of candles, for example) so CO2 will be generated inside the "chamber" and you will be able to see if the output voltage decreases with the gas concentration.

Regards.