Air Quality Sensor

AnneFR

@epierre hi, could remove some doubts regarding the mq131 sensor?
thanks,
Anne

epierre

(shinyei ppd42ns)

I corrected the following sketch which had bad yg/m3 to ppm conversion thanks to LouiS22 from Domoticz forum remark (type conversion errors leading to 0 values).

https://github.com/empierre/arduino/blob/master/DustSensor_SamYoung_DSM501.ino

I guess I'll have to do the same here (but I lack time...):
DustSensor_Shinyei_PPD42NS.ino

alexsh1

@epierre Thanks - I'll have another go at Shinyei PPD42NS sketch once I have time. Ideally, I'd like to combine it with MH-Z14A sketch.

epierre

@alexsh1 which is a good idea, also a temp/him is quite usefull though I didn't added the impact in the formulas so far.

alexsh1

@epierre I have combined the two sketches - works just fine, but need to find a spare temp sensor.The hardware setup looks messy so adding an extra i2c sensor will complicate it even more. I'll come back with the combined three-in-one sketch shortly.

BTW, this

gw.send(msgPM25.set("ppm"));

still throws the following error. What version MySensors are you using?

 MySensors: Unknown/Invalid sensor type (43)

alexsh1

@epierre I started looking at the formula you have used in ppmv calculation. Why do you need ppmv? All values in EPA or Europe are in μg/m3.

The ppmv equation is:

ppmv = mg/m^3 * (0.08205*T)/M

T = atmospheric temperature in kelvins = 273.15 + °C
M = molecular weight of the air pollutant = 28.97
(http://www.engineeringtoolbox.com/molecular-mass-air-d_679.html - this is a good link)
0.08205 = Universal Gas Law constant in atm·l/(mol·K)

Your code for this equation is:

ppmv=(float)(((concentrationPM25*0.0283168)/100) * ((0.08205*temp)/0.01))/1000;

Now, I think it should be as follows:

temp = °C + 273.15

int temp=273.15 + 22;

22C - is a typical temp inside though the intention is to use a sensor

The amended ppmv equation is going to be:

ppmv=(((concentrationPM25*0.0283168)/100) * ((0.08205*temp)/28.97))/1000;

I have not changed 0.0283168 /100 - not sure that this is. And the whole thing is divided by 1000? why?

IMPORTANT UPDATE:

I have just received the following result:


PM10: 7373

send: 11-11-0-0 s=0,c=1,t=37,pt=7,l=5,sg=0,st=ok:1.744

If I use the following web-site http://www.herramientasingenieria.com/onlinecalc/ppm-mg_m3.php and the result is
mg/m3 = 7373 * 0.283168/100 = 2.08 mg/m3
The molecular weight is 28.97 for dry air

2.08 mg/m3 is equivalent to 1.74ppm for a gas with molecular weight=28.97 Pressure=1013.25, Temperature=22C

Success!
Now I have found a spare BMP280 (temp and pressure sensor) which I can use here.
Additionally, I need to use 24-h and 1 year averages - https://www3.epa.gov/airquality/particlepollution/2012/decfsstandards.pdf

Stay tuned!

epierre

@alexsh1 said:

@epierre I started looking at the formula you have used in ppmv calculation. Why do you need ppmv? All values in EPA or Europe are in μg/m3.

because domoticz only knows ppm... and many AIQ like use only that except for particles.

temp = °C + 273.15
int temp=273.15 + 22;
22C - is a typical temp inside though the intention is to use a sensor

not for me ;-)

The amended ppmv equation is going to be:
ppmv=(((concentrationPM250.0283168)/100) * ((0.08205temp)/28.97))/1000;
I have not changed 0.0283168 /100 - not sure that this is. And the whole thing is divided by 1000? why?

was a volume conversion

The molecular weight is 28.97 for dry air
2.08 mg/m3 is equivalent to 1.74ppm for a gas with molecular weight=28.97 Pressure=1013.25, Temperature=22C 
Success!

success for the mysensors value ?

MySensors: Unknown/Invalid sensor type (43)

Domoticz doesn't recognize this command... I use 1.5.x but I have my own gateway to domoticz

alexsh1

@epierre Yes, all works fine now. I modified the combined sketch and here is the final result.

0_1463152922051_Screenshot (21).jpg

moskovskiy82

Which one to use for detection of fire smoke? Currently have got several MQ-7 but they don't seem so good at it

epierre

@moskovskiy82 said:

Which one to use for detection of fire smoke? Currently have got several MQ-7 but they don't seem so good at it

as discussed just above, a particle sensor could be good for smoke is a particle concentration, coupled with heat this would be a good indicator

if (concentration > 315000) {
     Serial.println("Smokes from cigarettes detected! Or It might be a huge fire! Beware!");

moskovskiy82

What about mq2 or mq135? Any experience? As a gas sensor will be more suitable detecting early fire

epierre

@moskovskiy82 said:

What about mq2 or mq135? Any experience? As a gas sensor will be more suitable detecting early fire

please see what it does, this is not a gas sensor this only detect particle size whatever the gas...

moskovskiy82

@epierre
It still detects concentration. Both state CO detection. So in case of fire won't they detect the increase in concentration much faster that the particle sensor like Sharp’s GP2Y1010AU0F or alternative?

alexsh1

@moskovskiy82 For a fire smoke, you can use pretty much any gas or particle sensor - there are a quite few gases formed during the burning process. MQ2 is highly sensitivity and has a fast response time. I can recommend it for a fire detection usage. However, I have been disappointed in MQ* sensors in general - there are not accurate, require 24h heat-up time, consume a lot of power etc. The only advantage is the price.

To detect fire to can use a flame sensor - http://www.instructables.com/id/Flame-detection-using-Arduino-and-flame-sensor/

rollercontainer

Can someone tell me how to read this line?

float mq135_ro = 10000.0;    // this has to be tuned 10K Ohm

Do I have to messure the sensor and adjust the variable or do I have to tune the resistance? If I have to do the first thing, when do I have to messure it? In warm state and clean air with a multimeter?

alexsh1

@epierre Did you check out the airbeam, which is based on a more expensive sensor Shinyei PPD60PV?
http://www.takingspace.org/airbeam-technical-specifications-operation-performance/

They made a step-by-step manual about building the same on Shinyei PPD42NS. I took their code and stripped a few parts and this is what I am left with:

#include  <SoftwareSerial.h>
#include  <FlexiTimer2.h>

int pin = 3;

volatile  double  rawParticalCount;
volatile  double  totalParticles  = 0;
volatile  double  particleCountToDisplay  = 0;
volatile  double  ratio = 0;
volatile  uint16_t  timeCounter = 0;

#define         numberOfPeaksRecording                            5
volatile        uint32_t        previousPeaks[numberOfPeaksRecording];
volatile        uint32_t        sumOfPreviousPeaks          = 0;
volatile        uint32_t        instantGoal                         = 0;
volatile        int32_t         delta                                     = 0;
volatile        uint32_t        slowMovingAverage             = 0;
volatile  boolean readyToSendData = false;

void  setup() {
    Serial.begin(115200);
    pinMode(pin,INPUT);
    
    FlexiTimer2::set(1,1.0/10000,readPin);
    FlexiTimer2::start();
    
}
void  loop()  {
   
    
    if(readyToSendData){
                Serial.print(rawParticalCount,  DEC);
                Serial.print("  Raw Particle  Count (0-10000) ");
                Serial.print(ratio, DEC);
                Serial.print("  Ratio (0-100%)  ");
                Serial.print(particleCountToDisplay,  DEC); 
                Serial.print("  Particle  Count");
                Serial.println("");
                             
                readyToSendData = false;
    }
    
}

void  readPin(){
    if(digitalRead(pin) ==  LOW){
        rawParticalCount++;
    }    
    timeCounter++;
    if  (timeCounter  ==  10000)
    {
        timeCounter=0;
           
        //Changes are made  here  based on  Chris Nafis's code: http://www.howmuchsnow.com/arduino/airquality/grovedust/
        ratio = rawParticalCount/100.0; 
        //Convert to  percentage, the shinyei reads 10milliseconds  to  90milliseconds  duration  for particles.  Basing on 10milliseconds, smallest  particle  assumingly  from  specification sheet.
        //FlexiTimer2,  reads 10,000  readings  per second, which would be  1 reading per 100 microseconds. 100 readings  would be  10  milliseconds. Since Shinyei runs  at minimal 10  millisecond range.  I divided 10,000  readings  by  100 to  get 100.
        //Good  example would be  rawPArticalCount  is  5000  half  of  the 10,000  readings were  active. 5000/100  would be  50  which translate to  50% low pulse occupancy.
        totalParticles  = (1.1*pow(ratio,3)-3.8*pow(ratio,2)+520*ratio+0.62);
        rawParticalCount  = 0;
          
                //  shift counters  over, code  adapted from  template  provided  by  Mike  Taylor  and Joshua  Schapiro  from  Carnegie  Mellon  University's  CREATE  Lab
                for (uint8_t  i = 0;  i < (numberOfPeaksRecording-1); i++)  {
                        previousPeaks[i]  = previousPeaks[i+1];
                }
                previousPeaks[numberOfPeaksRecording  - 1]  = totalParticles;
                sumOfPreviousPeaks  = 0;
                for (uint8_t  i = 0;  i < numberOfPeaksRecording; i++)  {
                        sumOfPreviousPeaks  +=  previousPeaks[i];
                }
                instantGoal = 2*sumOfPreviousPeaks;
                
                delta = instantGoal - slowMovingAverage;
                
                if  (delta  < -5000){
                        slowMovingAverage = slowMovingAverage - 250;
                } else  if  (delta  < -2500){
                        slowMovingAverage = slowMovingAverage - 120;
                } else  if(delta  < -1200){
                        slowMovingAverage = slowMovingAverage - 60;
                } else  if(delta  < -500){
                        slowMovingAverage = slowMovingAverage - 25;
                } else  if(delta  < -5){
                        slowMovingAverage = slowMovingAverage - 5;
                } else  if(delta  < -1){
                        slowMovingAverage = slowMovingAverage - 1;
                } else  if(delta  > 5000) {
                        slowMovingAverage = slowMovingAverage + 500;
                } else  if(delta  > 2500){
                        slowMovingAverage = slowMovingAverage + 250;
                } else  if(delta  > 1200){
                        slowMovingAverage = slowMovingAverage + 120;
                } else  if(delta  > 500){
                        slowMovingAverage = slowMovingAverage + 50;
                } else  if(delta  > 5){
                        slowMovingAverage = slowMovingAverage + 5;
                } else  if(delta  > 1){
                        slowMovingAverage = slowMovingAverage + 1;
                }
                
                particleCountToDisplay  = slowMovingAverage;
        readyToSendData = true;
    }
}

I have not adopted it for MySensors yet.
I like moving average they use, but the values do not make sense to me:

0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  53470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 27.0000000000  Ratio (0-100%)  53970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  54470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 21.6200008392  Ratio (0-100%)  54970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  55470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 24.2800006866  Ratio (0-100%)  55970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  56470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 24.1200008392  Ratio (0-100%)  56970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  57470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 22.3199996948  Ratio (0-100%)  57970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  58470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 20.3600006103  Ratio (0-100%)  58970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  59470.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 17.6299991607  Ratio (0-100%)  59970.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  60220.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 19.0499992370  Ratio (0-100%)  60720.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 0.0000000000  Ratio (0-100%)  60715.0000000000  Particle  Count
0.0000000000  Raw Particle  Count (0-10000) 20.6599998474  Ratio (0-100%)  61215.0000000000  Particle  Count

epierre

@alexsh1 said:

I took their code and stripped a few parts and this is what I am left with:

@alexsh1 the PPDN42 is for 1 micron and 2.5 micron , this one is for 0.5 micron

airbeam has standard code: https://github.com/HabitatMap/AirCastingAndroidClient/blob/master/arduino/aircasting/aircasting_shinyeiPPD42NS.ino

or I've not seen ?

alexsh1

@epierre this is not the air beam code. They introduced "do it yourself" consept similar to the airbeam but with a different sensor (shinyei ppd42ns)

rollercontainer

Hey guys, I am struggeling to get the MQ135 to work. Is there a step by step howto? Or can someone answer my question from 3 days ago?
I would appreciate it very much.

alexsh1

@rollercontainer I followed David Gironi method to calibrate - it is described as follows:

Before you can use the sensor, it has to be calibrated. For this, connect the sensor to your circuit and leave it powered on for 12-24 h to burn it in. Then put it into outside air, preferably at 20°C/35% rel. hum. (humidity is not so crucial). Read out the calibration value as such

float rzero = gasSensor.getRZero();
Wait until the value has somewhat settled (30min-1h). Remember, this is an ADC measurement so you might not want to wait some time between reading the sensor and also do some averaging. Once you have determined your RZero, put it into the MQ135.h. Note: Different sensors will likely have different RZero!

#define RZERO 76.63

Congrats, you have calibrated the sensor and can now read the CO2 ppm value in your sketch

float ppm = gasSensor.getPPM();

Air Quality Sensor

17

11.7k

11.2k

113.0k