VOC Sensor Update (TGS8100)

[admin]

Almost 10 months after i bought the TGS8100 sensors i finally managed to find time and start testing if and how they work
This is so far my test bench. I plan to test it for the next couple of weeks and see how it reacts to various gases.
Any feedback or experience from fellows that are using already VOC sensors is highly appreciated!


The VOC sensor is the TGS8100 from Figaro
It is a promising little sensor with quite a low power consumption, which shouldn't affect all too much the temperature reading (so they say, but i need to test);

There is already a breakout board for this sensor from Unitronic, 
http://en.unitronic.de/news/news-details...ality.html
Though it costs 60 euro, and my target is 30.
All in all it will be a small board that can be connected to the existing CO2/Dust monitors via a couple additional wires - so easy upgrade.
But since the sensor will be driven by an Atmega328, this leaves the possibility also for standalone operation, and - e.g. connecting some RGBLED Strip to have a standalone VOC meter.

Still need to check when exactly VOC levels change though

[escheers]

(02-22-2017, 05:24 AM)admin Wrote: Almost 10 months after i bought the TGS8100 sensors i finally managed to find time and start testing if and how they work
This is so far my test bench. I plan to test it for the next couple of weeks and see how it reacts to various gases.
Any feedback or experience from fellows that are using already VOC sensors is highly appreciated!


The VOC sensor is the TGS8100 from Figaro
It is a promising little sensor with quite a low power consumption, which shouldn't affect all too much the temperature reading (so they say, but i need to test);

There is already a breakout board for this sensor from Unitronic, 
http://en.unitronic.de/news/news-details...ality.html
Though it costs 60 euro, and my target is 30.
All in all it will be a small board that can be connected to the existing CO2/Dust monitors via a couple additional wires - so easy upgrade.
But since the sensor will be driven by an Atmega328, this leaves the possibility also for standalone operation, and - e.g. connecting some RGBLED Strip to have a standalone VOC meter.

Still need to check when exactly VOC levels change though

promising update!

[admin]

Quite interesting articles from two different companies
AMS - who doesn't sell CO2 sensors - http://ams.com/eng/content/view/download/387555
CO2Sensor - who doesn't sell VOC sensors - http://www.co2meter.com/blogs/news/51851...difference

First one claims measuring CO2 is not enough as VOCs are the source of danger
Second one claims VOCs are related to CO2, but there may be also other sources
))
Both agree though - that a VOC sensor needs some calibration from time to time, and can be used to roughly estimate the CO2 level.
Fact is, that is a vast amount of commercial devices that only measure only VOC and refer that it is related to CO2. Of course for them it is much more convenient, as VOC sensors are much much cheaper than CO2 sensor, yet they sell them in the range of 100-200 eur.

I hope that soon you will have the ability to measure both and decide for yourself.
For sure there will be also a cheaper version of the (now Air Quality) Monitor, that would measure only VOCs

[escheers]

AMS is an interesting article. I will follow this thread about your research.
Also I think of buying an extra meter for measuring outside, having a lot of houses with wood stoves and fireplaces around here. Ventilation is sometimes impossible because of the smell.
I'm curious about the outsite dust PM25 and PM10

[admin]

i am progressing slowly with understanding how to use this sensor, but it is quite weird
since the output depends to some level on temperature and humidity, compared to the optical CO2 based sensors, you cannot just put it outside and compare to the inside values (especially where there are 15 degrees difference)
So it took me almost a week of testing before i realize something quite weird
the chart below shows what happens when i open the door (going down, close it, going up, then again open and close it)
there is a 15% difference in the output
And this all happens in an isolated room where the CO2 ppm is ~420-430, temperature is 22 degrees (an additional heater works, so that there is no diffrence when i open the door)
Which shows that using the voc sensor you can measure even beyond what the CO2 sensor measures (e.g. even cleaner air)

[admin]

Short update on the timeline for the VOC sensor

Delivery:
- i expect to have the first sample PCBs by end of March
- The first week of April i will do some mass testing to see how individual sensors behave
- if everything is alright i expect to start shipping from mid of April

Results:
- So far it is clear that compared with a CO2 sensor, the VOC sensor output does not get beyond 1800 ppm from human activity
- on the other hand, spraying some deodorant increases the reading much more, since the sensor reacts on ethanol and simillar gases
- i am yet to test it against some cigarette smoke. The problem is that here it is forbidden to smoke inside public areas, and since i am not smoker it is a bit hard to find a closed space to test it

Extending existing devices:
- To extend existing devices it is enough to connect the VOC sensor with 4 wires + some splitter - i will include them
- most probbably it wouldn't fit into the existing boxes, so i will also create some small enclosure, that can be glued to the main box

[escheers]

Good news, sounds promising ?

[admin]

boards have arrived several days ago, yet, as was busy testing the battery-operated board and have not yet sent them to assembly. Anyway, today i finally got some interesting breakthrough.,.. in the last month i kept wondering why, when comparing to the CO2 sensor, the VOC sensor seems to saturate at about 1200 ppm, while others (Velux/Rehau Raumluft Sensor) claim it works till 1800.
I even bought a Rehau Raumluft Sensor, based on the ams sensor, yet, it still delivers too weird values, seems to miss also temperature and humidity compensation

But today i finally managed to make this graph. Green is the CO2 sensor (CM1102) and Red is the VOC sensor. until 19h, the room was open, then we closed the door. Until 19.40, where i again opened it and at 20.30 closed it.
You can see how the VOC sensor responds quite rapidly, compared to the CO2 sensor (which has similar response time to the CDM7160 and CM1106, though i need to test this). And we are not speaking about some averaging algorithms, as this happens in the course of 30+ minutes. So apparently something changes, though i do not know why so fast. The other sensors were in the boxes and this one is bare, so this may be also a reason

[admin]

and one from yesterday
what is not yet exactly clear is to what level CO2 and VOC do corellate (and if they should)
as the VOC sensor measures rather Ethanol and go some level CO. And Ethanol is exhaled when we breathe, and also from many other sources.

also i am not sure if it is possible to create a good initial calibration, which would be the base for a good correlation, as it seems like the outdoor CO2 in my city does not fall below 420-430, and when i was in my country house it was 400, and the VOC sensor was measuring indeed more than 10% better results.. for 5% diffrence in CO2

it is not visible on the charts below as i have alligned them

[admin]

i am sorry, but the factory delayed my order for the sample boards too much so i am still waiting for them, to do some testing with several sensors together