Using a combustion analyzer on an engine

Gordan

I recently repowered a garden tractor with an LPG engine and I thought I'd check combustion to see if it needed tuning. This is just an air cooled engine with a simple venturi, no electronics at all, so you have a single screw to tune the mixture - just like a boiler, right? :-) Unlike a boiler, however, you're supposed to tune these to 14% CO2 and .5-.75% CO - that's 5000-7500 ppm! I stuck the probe inside the exhaust port as far as I could, and I can't get any readings that make sense. CO2 is reading at .2 - 2% (measured), O2 is reading at 20-18% (calculated based on lean mixture, I expect this to be wrong), and I know that I'm not anywhere close to that lean; the engine wouldn't run. The only reading that makes sense is CO (measured), but I'm not even trusting that one given how screwy the others are. Besides, the UEI SmartBell Plus that I'm using is only supposed to have a range up to 2000 ppm, even though it was reading out over 6000 ppm at times.



So... combustion being combustion, any ideas about why my CO2 reading is so off and how to improve it?

TDR

gas engine

Hello

Unlike a boiler exhaust flue. the exhaust on a gasoline / LP engine, is not constant pressure, there are times when the exhaust is flowing backward. If you look at the location of O2 sensors on cars, the location is important. If you put your probe, closer to the exhaust port, you will get different data. You could look at the summit racing online catalog, for fuel air testing equipment, and more information.

Thanks

TR

TDR

gas engine

Hello

Unlike a boiler exhaust flue. the exhaust on a gasoline / LP engine, is not constant pressure, there are times when the exhaust is flowing backward. If you look at the location of O2 sensors on cars, the location is important. If you put your probe, closer to the exhaust port, you will get different data. You could look at the summit racing online catalog, for fuel air testing equipment, and more information.

Thanks

TR

Gordan

Thanks...

I thought that perhaps this had something to do with exhaust pulses. So is it the pressure fluctuations themselves, rather than the air they draw in, that would throw the readings off? If it's just dilution, it seems like I should be able to adjust. Or simply find the peak CO2 and know that I'm stoichiometric, and then richen a bit...