#300 How to measure CO2 and VOC with Arduino. Which one is better?

do you sometimes feel tired during meetings or do you sometimes have a headache after work do you want to change that then this video can be interesting for you because we will measure harmful gases in your environment which can trigger both greasy youtubers here is the guy with the Swiss accent with a new episode and fresh ideas around sensors and microcontrollers remember if you subscribe you will always sit in the first row in this video we will focus on indoor climate and we will concentrate on gases where the primary cause is humans not machines or lousy building materials we will look at co2 and its influence on indoor air quality and your well being and how it can be measured we will see the relationship between our co2 sensors and global warming we will use a second way to assess indoor air quality vo C or E co2 and we will build sensors which transfers the measured values via MQTT into graph Anna let's start we all know what fresh air is and we all remember the difference when entering a meeting room packed with people but how can we measure this difference the easiest way is to concentrate on gases which are produced by humans the first gas that comes into our mind is carbon dioxide or co2 these days this is a famous gas in global warming discussions and we will later learn why its significance for indoors climate comes from the following fact humans need oxygen to live and produce co2 by breathing we change the percentage of the different parts of the air around us but what are the components of fresh air according to NASA outside air consists of 78% nitrogen 21% oxygen and 1% others including co2 so co2 does not seem to be a problem because it is quite rare besides that co2 has no odor and also no color so why to bother low oxygen levels are dangerous for humans as we see here we could measure oxygen levels to assess this danger for example in our car a so called lambda probe measures oxygens in the exhaust gases to check if the engine burned fuel correctly but these sensors are not used for indoor applications maybe they are too complicated too big or need too much energy I do not know fortunately we can go the route of so-called secondary indicators because we inhale oxygen and exhale co2 we can use this gas as an indicator of the oxygen level at least in our offices or living rooms the more co2 the more people consuming oxygen and the lower its level as soon as we let fresh air in more oxygen and less co2 will be present the co2 level therefore is a perfect indicator of the indoor climate simple unfortunately co2 sensors are quite expensive and also have a slow reaction time that's why a secondary indicator is used for the same purpose transpiration of humans humans do not only exhale co2 while sweating we also create h2 ethanol and other gases in small quantities they are part of volatile organic compounds short VOCs from a chemical perspective they are entirely different from co2 but because one of their sources human bodies is the same we can use them also as a secondary indicator sensors that measure these gases are cheaper and react much faster than co2 sensors but because humans are not the only so of BOC they respond differently as we will later see so we can say humans are the primary user of oxygen and the primary source of co2 and VOCs in typical office or living rooms if we monitor co2 or VLC we either can warn people to open the window before they lose performance or before they get a headache or we can switch on the automatic ventilation the next question arises how do we measure co2 or VOCs as said before co2 is other less and also has no color this is not entirely true if we look at the whole spectrum co2 heavily absorbs waves with a wavelength of four point two six micrometers for a comparison green color has a wavelength of 0.5 and infrared light for remote controls uses a wavelength of 0.9 micrometers so we need sensors that detect light in the four micrometer range we know this range from our flirt terminal imaging cameras they work from 3 to 12 micrometers and here we see the relevance of co2 for our climate the visible sunlight passes the atmosphere without substantial dampening and reaches the ground here it warms the soil and is converted into heat as we saw co2 blocks this radiation from escaping back to space so the heat stays with us which is OK for me in November in Switzerland when temperatures are low a little warmer would be welcome coming back to our topic all current co2 sensors work like that they consist of a particular IR ty out a distance where the air can enter and an IR detector usually with a filter for four point to 6 micrometers in front of the sensor that's all but because the co2 levels are minimal these parts must be pretty boost in high-precision here I have two such sensors the left from the Swiss sensors specialists sin Syrian and the right one from the Chinese manufacturer Winson both are not cheap interestingly the sin Syrian Center is less expensive at a cheeky than on Aliexpress the SCP 30 works on I square C and the MH C 19 on serial both work with 3.3 volts and are 5 volt tolerant and for both we find libraries I have a third one from a German sensors company since air unfortunately I was not able to get it running also because no Arduino library exists but before we start to use them we have to understand their limitations the sincere ian provides an accuracy of plus minus 30 ppm plus 3% and the Winson plus minus 50 ppm plus 3% which is as we will see good enough for our application sincere ian guarantees a lifetime of 15 years and wins and says more than 5 years why is this important because these sensors typically are for cotton sensors in equipment like ventilation and nobody cares about them but the ventilation easily can malfunction if the measured values start to be wrong and here we enter an essential area co2 sensors have to be calibrated from time to time fortunately we can use a trick the outside air at sea level currently contains around 400 ppm of co2 I say currently because as we see here it is changing over the millennia so we might have to calibrate our sensors to a higher value in the future more important this value is reduced by about 3 percent for every thousand feet or 300 meters altitude if we want paste measurements we have to do this compensation already now and we see trick to avoid global warming just move to a place above 3000 meters then your co2 levels are at 300 ppm problem solved choking a sight you can easily calibrate your sensor by putting it outside your home and note down the value to calibrate it but most of the sensors are fixed inside buildings to calibrate those sensors you can assume that the lowest measured level is around 400 ppm and adjust the readings accordingly both sensors do this automatically if you select the function but it can take a few days till this calibration is done and the sensors display accurate numbers the SCDS 30 also have a feature to compensate for pressure you either have a pressure sensor available and use its values or you enter an average value or also the elevation of your location to get closer to reality for the MH c 19 you have to do it in your sketch it always calibrates to 400 ppm another fact you have to pay attention after startup the values are unprecedented in door air monitoring but it can be a problem if you want to do measurements for experiments where values change quickly co2 sensors are no low-power devices according to the datasheet the SC d 30 consumes 19 milliampere and the MH C 19 consumes more than 60 milli amperes with Peaks up to 150 milli amperes and both should not be switched off so battery-operated devices would need a massive battery coming now to the vo C centers also here I have two sensors the SGP 30 from sin Syrian and the EM ICS 6814 from SG x sensor tech the SGP is 30 from sin syrian measures 2 Casas h2 ethanol the M ECS 6814 has three sensors and claims to detect many more cases but it has analog outputs and it's interfacing is quite complicated this is why I did not use it for my tests both sensors use a so called hot plate with a metal oxide surface to detect these gases fortunately the sgp 30 is tiny so the heating of the hot plate only takes 48 milli ampere the m ICS 6814 consumes around 100 milli ampere these are also not sensors for battery-operated devices the sgp 30 runs on only 1.8 volts but the breakout part has a voltage regulator and level shifters and works with 3.3 and 5 volts but what can we do with these two raw measurements the sgp 30 uses these raw measurements and outputs two values one is called t vo c or total vo c and the other ii co2 or co2 equivalent which is an artificial number that should somehow match the co2 values we will lay to see if this is true or not in the next step i built three sensors to with esp8266 and one with an ESP 32 why because i wanted to connect them to my graph on a dashboard of course so all sensors send and qtt messages to node red which feeds it into in flux DB in the end Ravana integrates the values into other weather and indoors data and shows nicely looking graphs you find videos on this channel explaining how to create such diagrams with your Raspberry Pi here are the measurements for the last two days all sensors were in my lab at nearly the same place about 50 centimeters away from me yesterday I came to the lab before 8 a.m. and the SD g30 had an issue the night before during the day I was a lot in my lap but not always and after sixteen o'clock I had a Skype call and had to close the door of my lab to not disturb my wife the first observation is that all three values correspond somehow the blue one is the e co 2 around 18 o'clock and also this morning it shows higher values than the real co2 sensors please do not think that I'm inert and therefore personal hygiene is the reason for that difference tishka for example could contribute more in this respect if you do not believe me try to wash your cut and do not complain about the scratches the MHC 19 seems to react a bit faster but maybe I did not use exactly the same sampling strategy for both sensors and be aware we are at the beginning of the measurements and the calibration process is not finished yet I will leave the sensors on and see how they behave in the future but looking at this table it becomes clear that we do not need a high precision to qualify indoor air quality this might be different for other applications looking at the curves we see how fast the values can go above 1000 ppm when I close the door this is not yet a problem but I was nearly alone in the room only dishka helped me with co2 production but because of her typical low activity behavior she does not contribute a lot this changes if you for example have a class of 20 people in school or 15 people in a small meeting room then these values can increase fast I think each classroom with closed windows should have such a sensor because school is hard enough and bad air quality should not be the reason for sleepiness I think other causes like late gaming or using apps on the smartphone and for the elder drinking too much alcohol during nights contributes enough to sleepiness of children in school maybe one of you guys creates a cool project which then could be used to build such a meter by the students I would create the video for the Assembly promised but coming back to the sensors first we go outside and we see if they really show around 400 ppm that seems to be okay the VLC sensor also measures 400 ppm but I suspect that the algorithm adjusts values below 400 ppm but how can we create co2 the simplest way is to use our body and exhale the air into a small container fortunately co2 has a bigger mass than air and therefore stays at the bottom of the char even if we do not close it by the way fire extinguishers use this effect if you want to protect a data center against fire a sprinkler would probably not be a good idea because water quickly causes a lot of damage in such areas therefore they use co2 to fill up the whole room as seen before it creeps below the oxygen and displaces it without oxygen the fire stops problem solved without collateral damage as you see my exhaled air contains a lot of co2 and it takes a few seconds until the co2 sensors read these high values the VLC sensor reacts much faster this can be useful for specific applications where speed is essential next we do the solder test with all sensors the co2 sensors are not very much impressed the vo see sensor shows very high numbers for e co 2 and t vo see and if we put a banana close to the VLC sensor it also indicates higher e co2 values but the air around the banana and the sort of fumes do not contain more co2 this effect is also visible in the diagram where we see the measurements of a proper co2 and a dlc sensor placed in a room of a shul building during pizza cooking the VLC measurements were very high I'm pretty sure pizza order does not trigger complains about drowsiness or sleepiness at least not with young people summarized we learned why co2 could be used to assess indoor air quality we saw the difference between a primary and a secondary indicator and when they can be used instead of primary indicators we saw how co2 is measured and that the same effect can isolate the earth and increase its surface temperature we used a different secondary indicator to assess indoor air quality vo C or the calculated E co2 it measures gasses in human transpiration we built sensor notes to transfer the measured values into gravano comparisons of the three sensors showed that the measured values were different the two 302 sensors were quite close and the VLC sensor in some situations was off by 30% and more playing with the sensors showed that it is easy to produce and store co2 exhale in a container and because co2 is heavier than air it stays there for a while we saw that soldering or placing a banana in the proximity of a BOC sensor made its values increase co2 sensors stayed roughly the same co2 sensors are more expensive than VLC sensors so you decide which one is good for your application I will stick with both because of the soldering the sin Syrian co2 sensor is much more expensive than the sensor from Vinson for a maker project the Jeepers is definitely good enough for industrial applications a trusted supplier for sure is the better choice by the way since Irian just presented a new co2 sensor based on acoustics no more on IR light it should even be smaller and under current sensor and microphones probably are also less sensitive to dirtying than optics maybe I will get one from them for a test one last thing air conditioning and heating or vehicles consume a lot of energy for conventional cars this is not a huge problem because they produce a lot of extra heat if you use electricity however this is no more the case air conditioning can eat up to 30% of the overall energy consumption therefore knowing when to exchange the air inside the car can lead to smaller batteries or bigger range I hope this video was useful or at least interesting for you if true please consider supporting the channel to secure its future existence you find the links in the description thank you bye

Channel: Andreas Spiess

Views: 68,272

Rating: 4.9431458 out of 5

Keywords: arduino, arduino project, beginners, diy, do-it-yourself, eevblog, electronics, esp32, esp32 datasheet, esp32 project, esp32 tutorial, esp32 weather station, esp8266, esp8266 datasheet, esp8266 project, greatscott, guide, hack, hobby, how to, iot, lorawan, nodemcu, project, simple, smart home, ttgo, wemos, wifi, CO2, CO2Sensors, Sensirion, SCD30, measuring CO2 tutorial, indoor, indoor climate, ventilation

Id: FL0L-nic9Vw

Length: 21min 12sec (1272 seconds)

Published: Sun Dec 01 2019