Carbon monoxide detector signal conditioning

[Denis Anipchenko]

Hello everyone,

I am new on this forum and I hope that you will be able to help me. I am trying to build a carbon monoxide (CO) detector for classic cars using a hot wire as sensor. The idea is to expose a "heating element" to exhaust gas. The resistance change of the hot wire would be proportional to CO percentage in the exhaust gas. The idea is to be able to detect between 0% and 10% ( from 0 ppm to 100000 ppm) ±0.1 CO.

First of all please see Fig.1 which shows the wiring connections. the wire is heated by DC voltage (1.5V - 3.5A). Please see Fig.2 for the schematics of voltage supply and sensing chamber (sensor). I have added buck converter after AC to DC converter to be able to have a better control on voltage/current.
The output voltage of the sensor would be directed to ADC. To prevent its destruction, I thought to attenuate the signal with PI pod attenuator, Fig 3 . To be able to achieve required attenuation I needed circa 0.7 Ohm resistors, to do so I connected several 1 Ohm and 10 Ohm resistors.

Firstly I was using MCP3421 (datasheet attached) as ADC but I think it would be too sensitive for the "highly" technological for the hot wire sensor. So I have switched to the inbuilt Arduino uno ADC using 5V reference.

But I still cannot get a steady reading unless I start to average the readings, which would increase time response.

Could you please help me with some inexpensive signal conditioning? Or advice me how to improve the reading?

 

[Externet]

Explore using an automotive sensor, dedicated sensors into the tail pipe instead of a hot wire.

----> https://www.parallax.com/product/605-00007

 

[Denis Anipchenko]

Many thanks for your reply.

I have considered many solution for this uni project. Unfortunately the solution you are suggesting has a very limited range of detection, 0.1% CO or 1000 ppm.

I have found electrochemical sensor capable of detecting up to 100000 pmm of CO but they are way to expensive, about £190 per sensor.

Hence the reason to opt for a hot wire.

 

[hevans1944]

I have found electrochemical sensor capable of detecting up to 100000 pmm of CO but they are way to expensive, about £190 per sensor.

Well, it is what it is. The choice of sensor must of course consider acquisition cost, but the primary criterion should be the sensor capability. Low cost "sensors" that don't work waste our time and yours.

How does your "hot wire sensor" measure up for that? How does the resistance of a coil of wire exposed to CO gas vary with the concentration of the gas? What physical mechanism is involved? Is there a chemical reaction between material on the wire surface and CO gas that causes the surface resistance to change? If so, how does this interact with the bulk resistivity of the wire to allow you to distinguish resistance changes caused by CO gas concentration versus resistance changes caused by other environmental factors? If a chemical reaction is involved to produce a resistance change, is the reaction reversible and, if so, how? How does your coil of heated wire distinguish between resistance changes caused by other factors than CO concentration?

Finally, what is the expected sensitivity of your "hot wire" CO sensor, defined in parts per million CO versus resistance? Will your "hot wire" CO sensor require compensation for ambient temperature, humidity, water vapor in the exhaust gas, or hydrocarbon molecules in the sampled gas stream?

There are many CO-specific gas detectors available. There is a reason that those capable of a 0 to 200000 ppm measurement range are more expensive than a typical 0 to 1000 ppm sensor. Exhaust gas from an internal combustion engine is a filthy "soup" of chemicals. It is likely you will need to mechanically and chemically filter the exhaust gas to isolate the CO for measurement. Even then, there will be other gasses such as oxygen, nitrogen oxides, and carbon dioxide present that can interfere with the CO measurement.

You might want to investigate the Bosch oxygen sensor, now a commonplace component of emission-controlled automobile engines, to see how it was developed and why it is specific for oxygen. It is used by the computer in the car's ECU to set the optimum fuel-to-air ratio, Similar principles may be valid for CO measurements, although I am not yet aware of any.