With these two sensors I hope to determine a quite exact concentration of CO and H2 gazes (by substracting H2 concentration determined by Q8 from Q7 readings)
Since the impedance is a little high for the ADC, I consider adding an op amp. But when RL is on the low side this means that the voltage for high concentration will be very near VCC. Hence the signal might be wrong at the ouput of the op amp. I would prefer to have it work at mid point.
So I was wondering if I could use a RL on the high side or even better, spillted on both sides of the sensor, like that:
Those sensors are not accurate at all. The resulting value depends highly on the temperature and the preheat or aging time.
Connect them to a circuit and keep the MQ-7 and MQ-8 on for 48 hours. You might have to repeat that sometimes when something changes. Only then you can determine the best value for the resistor RL.
The MQ-7 does not just measure CO and H2, it is sensitive for many gasses, for example also for CH4.
The MQ-8 is sensitive for Alcohol and CO and CH4 and many more.
The start value for the resistor RL is 4k7. That means the right value for you might be somewhere between 470Ω and 47k.
The impedance of a analog input of an Arduino is very high. A OpAmp is not needed and makes it even worse.
Connect the 'A' side to 5V and RL to the 'B' side and GND. Use the Arduino analog input to measure the voltage.
Can your circuit provide enough current to power both heaters ? I hope you are not using the Arduino 5V pin for that.
In the end it will not be accurate at all. To get accurate results, you have to pay a lot more money.
Yes, I suppose these sensors are not really accurate but It seems the alternative are sensors that cost hundreds of euros.
My question was more about the possibility to place RL on both sides of the sensor.
So, you tell me that the MICS5524 is much more accurate? Since it does also measure other gazes, I can't really tell the advantages of this one sensor regarding the accuracy. But it does consume much less.
And yes, I plan to use an STM32F103 for the 10 bits ADC that would be sufficient considering the overalll precision.
I don't know if it is more accurate, probably not. But it does not have a big heater. It uses about 32mA for the heater compared to about 150mA for the others.
I know what your question was, but that is not a solution. You have to put the analog output into a good voltage range with RL. You can only determine the right value for RL after it was on for 48 hours. Adding extra hardware to make things worse is no solution.
I don't think Rl should be changed. I haven't seen any chart of Rs/R0 for another value than Rl = 10 KOhm even though they say RL should be lowered for high concentrations and increased for low concentrations.
I just start to study this. My understanding is that R0 has to be determined in clean air (~0 ppm of CO) then considering the Rs/Ro curve for Rl = 10 KOhm we get the CO concentration. There are ready made libraries for that.
I am studying a board that would measure CO2, CO and in option O2 with the best accuracy possible but the cost should be reasonable. The CO2 IR sensor is around 20 euros. The O2 sensor is also around 20-30 euros. Plus other circuitry. The CO sensor cost could be 20 euros but not let say 50 euros.
