Affordable CO2 sensor?

[O2surplus]

Originally Posted by Hoppy

Does this work by shining a beam of IR light through a length of air gap, and measuring the absorption in the wave length band that CO2 absorbs? If so, it must require a flow of air through the air gap. And, if that's the case there would be a minimum volume of air around the sensor that would be needed to refresh the air in the gap so it can follow the fluctuations in CO2 content. Or, am I off on a tangent here?

Since reading that the sensor has to "warm up" a little while before it produces reliable readings- I'd bet your description of how the sensor works is accurate. Providing for gas flow through the sensor will be an important step to keep it working accurately. I'm wondering if the sensor and a very small fan or other air moving device, could be mounted remotely in a sampling tube. The sampling tube would connected at both ends to a chamber that has a open end submerged in tank water. The air trapped in the tube would be constantly recirculated through the sensor, while the Co2 concentration would be allowed to diffuse freely in and out of the chamber, Similar to how a ph drop checker functions.

[mistergreen]

Yes,
it uses IR (non-dispersive infrared (NDIR)) to read the CO2 spectrum. It lights up when plugged in. It has 2 air ports for air movement. I think the issue when I dunked it in water, the breather bag collapsed to plug up the air ports. I'm building a proper plastic box for it now with a breather bag membrane. I'll use rubber bands to hold the bag on. I'll have to find something heavy to glue to the box so it would sink.




From looking at the ppm in the air, it shifts from 400ppm - 500ppm. I'm thinking it'll shift back and forth as well in water. So I'll plan to take the average reading within a certain time frame like 10 seconds or so.

[mistergreen]

I made the box for this and threw in a bunch of washers to see how much weight it takes to sink the thing. It'll need a lot. Looks like this route is not a good one. I'm thinking suction cup is the way to go.

My only concern is there is quite a bit of air in the box but the membrane surface area is quite big so the gas exchange will be ok.

[mistergreen]

Things look good for a bit!


1701ppm outside water



I didn't see any noticeable leak so I went ahead with the experiment


It took a few minutes but the ppm started dropping. It dropped 1 ppm every 2-3 seconds. This is totally responsive in an aquarium where we're looking at 9ppm - 50ppm max. The ppm is a lot more stable in the container compared to air. There was a 2-3 ppm deviation sometimes.

I wanted to see where it would wind up but the readings stopped and notice there was a little water. The experiment ended at 1648ppm. I'm drying things out now. Hopefully there's no permanent damage. I'll need more silicone and thicker band next time.

[mistergreen]

whew, no permanent damage.

[O2surplus]

Nice work so far! How are you getting the ppm readings? Is the sensor connected to an Arduino? If you've already written a sketch to do the math, have you thought about adapting it to run on one of the "close enough" PAR meters, just for fun? One from the latest batch could be re-flashed without to much trouble and your arduino would be freed up again for other uses.

[mistergreen]

yup, it's connected through the arduino. It's outputting to the serial window for now.
Ultimately this application would work better as part of the DIY controller yes?
http://www.plantedtank.net/forums/sh...d.php?t=192436

It'll be more efficient as a group of devices with a single controller. You can tell the solenoid to turn off at a certain ppm and on at etc....
The settings would be unique to each tank I think.
But for now I'm building as a single app for testing purposes.

I'll have to find a way to build a better casing. I'm worried the drop in ppm is due to air leaking.

[O2surplus]

I like the idea of an aquarium "controller" per se, but I don't like the idea of tying in too many controls to one micro-controller. I would rather have a system that is comprised of "slave control modules"( each w/ it's own programming, that run stand alone and merely report their status to a main display), than a system of centralized control. Having all your functions tied in to a central controller sounds great until something malfunctions. Then your forced to take the entire system offline for repairs. The module based approach gets around this potential pit fall, because each module operates independent of a centralized controller, and can be repaired without having to take the entire system offline to do so. You could incorporate the CO2 sensor into a "stand alone- Ph control module" that reports it's status and receives commands from a central controller via the I2C communication protocol. Software could be written to allow the module to operate independently but with certain fail safes included to prevent unwanted malfunction.

[mistergreen]

Ah, makes sense. This sensor can communicate through i2c. I'm using uart at the moment.

[Hoppy]

When I tried one of those breather bags as a membrane for a drop checker I was disappointed in how slow it worked. The breather bag is designed to pass a very small amount of air per minute. If you used a membrane made for an oxygen sensor, which you can get at Perkin-Elmer (?), it should work a lot faster, but the membranes are extremely thin and easy to break. Avoiding small water or air leaks is a very difficult thing. You need a brainstorm session to come up with some kind of breakthrough for that. (And very good manual dexterity)

[mistergreen]

I read silicone rubber is a good gas permeable substance. Will have to test that out.

I can literally mold a membrane to a case but I'd be worry of the fumes affecting the electronics.

[Hoppy]

I think I read that most thin plastic films are gas permeable to some extent. But, to get much gas through them the film has to be very thin. Silicone would seem to be a good candidate, since it is one of the most "leaky" plastics used for hoses.

[mistergreen]

I'm running the sensor again. No leaks so far. I put a lid with large holes on thinking the membrane will need a little protection in the future like from rasping algae eaters. That's one thing I have to think about is designing something that you can easily replace the membrane. I can see it clog up overtime with algae etc...

Ok, So the ppm is dropping at a rate of 2.7 ppm per minute. Still pretty good in terms of responsiveness.


I'll run it till the ppm bottoms out.

[mistergreen]

The tub of tap water bottoms out at 970 ppmv <- per volume.

I just stuck it in a non-co2 aquarium and I'm watching the co2 rise as I write.... seem to be at 1009+ ppmv now.
So there is EXTRA CO2 in a low tech tank after all.

[mistergreen]

The reading in the 10G low tech tank kept rising to 2700+ppm! Somethings wrong.
I deduced that the membrane is facing the outflow of the filter. I suppose CO2 kept going through but can get out through normal diffusion.

So I placed the membrane face down. PPMs are falling. Interesting. This will help in the design.