CO2 levels

Try 0.038%.

I think you're confusing carbon monoxide & carbon dioxide (neither of which are present in the atmosphere in levels anything like 2%)

Tim

It makes sense at that level. What was confusing me was that the gastester needs calibrating to 2 in normal atmosphere before sticking the probe up the exhaust pipe so I assumed that atmospheric co2 was 2%. I also didn't link the 380 odd ppm as 0.038%. Told you I knew I was being thick :-)

I believe you are thinking of the gunsons tester, which does calibrate at 2 percent and reads in CO2, the reason for the discrepancy, which I read somewhere on their literature is : It doesn't actually measure CO2 !! It measures something else and takes that to give a reference level (oxygen IIRC) With care, (the unit must not be knocked about) and using a steady power supply, the one I had gave fairly consistent readings, but it was not accurate enough for modern vehicle checking, something with a carb. or early injection yes.

As someone else has said, you're also confusing CO and CO2. Your gas tester measures CO (carbon monoxide) not CO2 (carbon dioxide). If you have 2% CO in the atmosphere that you are inhaling, if you're not dead now you soon will be.

Yes, it's a Gunsons tester. I've just had a look at the box and, as Steve Firth says, it measures CO, not CO2. My car failed it's mot due to a CO reading of 0.9% and, after changing the inlet temperature sensor, it's reading 0.3 (I might well have got the decimal point in the wrong place to add to my catalog of errors) so it's at least given me guidence that it's worth taking in for another reading since it's gotten better. There's also a small hole in the back box that needs fixing first and that's likely to be sucking some air in.

It still takes me back to the original query but on CO not CO2 but presumably the CO gets converted to CO2 in the burning process.

I had one of those, a green box with a fragile hose connector at the top (IIRC).

It was useless, readings were all over the place, and that was just while "calibrating" the thing.

David

Um, CO is present in the normal atmosphere in pretty negligible quantities. Normal atmospheric level (according to wiki) is 0.1 ppmv. (parts per million my volume) which = 0.00001%

The CO in your exhaust is produced by incomplete combustion of the fuel, and these days, mostly converted to CO2 in your catalytic converter.

You *can* burn CO but it's not a normal fuel these days (it used to be in town gas supplies).

Tim

I'd agree with that - completely useless. Wideband O2 sensor kits ain't that much more expensive these days and far more use.

I wonder if there's much mileage in a DIY circuit attached to an O2 sensor from a random car. For example:

Assuming they're all standard zirconia sensors, do they have the same electrical behaviour (ie the same voltages at same O2 concentration)? So no need to calibrate?

Theo

I've a feeling that is for a narrow band O2 sensor, as fitted to production cars. They can only measure air fuel ratios close to stoichiometric (14.7:1). If you have an older car which is designed to run at a varying AF ratio, they won't be much use. Or for looking at a fault on a modern one. For that you need a wideband, which will measure AF ratios from approx 9:1 - 20:1. On a non cat. car at least, the lambda reading has a pretty direct relationship to the CO output.

There is at least one self assembly kit made by Tech Edge in Oz - but it may be more expensive than some ready made units.

is the URL I've got for them, but it appears down. Hope they are ok - I've got one. Used with the MegaSquirt in my old Rover.

Ah, that's interesting. There do seem to be a few wideband projects around, for example:

That's still up. There's a list of users of the Bosch sensor on that site, but they all seem to be US vehicles. I wonder if any UK vehicles use it?

Theo

and

I'm not sure any production vehicle uses a wideband sensor?

The big mistake with those is to connect them to the vehicle battery. As the load on the battery varies (fan cutting in/out, etc) the voltage varies by enough to throw the calibration out.

Use a standalone battery or even better, a mains-powered 12V with stabilised output.

I used a Gunson set up like this to get my recalcitrant single-cylinder two-stroke motorcyle exhaust into the recommended CO range, after a frustrating session trying to set the idle air and throttle stop by ear.

Terry Fields

Is that what you found, ISTR it was meant to be used from the battery of the vehicle ?

I did not have alternative power supplies, and I probably would not have thought of using one anyway.

Did it work ? I think I gave up after about an hour. This was around

1996-8 ish, so a while ago.

For balancing twin carbs, I found Gunson's 'Carbalancer' (?) good.

David

[...]

The instructions with the Gunson's one I had strongly advised using a second battery.

Chris

OK, but did it do the job for you ?

David

I used the battery pack I use for jump starting. The first time I used the tester I didn't top up the battery pack and did notice the reading drifting lower as I measured it. That was to confirm I got a similar reading to what the mot station did. Then I got a new sensor and replaced it but topped up the battery pack for the next time and it seemed pretty stable, down from 0.9% to 0.3%.