Converting percentage O2 to Lambda

[diycerb]

Hi,

Playing with an old gas analyser at the moment that gives readings of CO, CO2, ppmHC and %O2.

Anyone know how to convert from %O2 to a lambda reading.

Cheers,
Rich.

[striker 20v]

for small band lambda voltage read out (means milivolt meter on the sensor wire GREY and chassis ground the volt meter has infinity resistance so you can do this with the engine running)

you'll get
450- mv = to lean to run on most engines
450mv on lean out (don't try that)
500mv is best economy but not with high comp aplications
550mv around stoich, thats what you want for nice economy
700mv richer side of stoich no economy but nice power
850mv SAFE high compression high rpm rich side (regarding detonation) my 160hp 4age is around there
900+ is pig rich and you'll need to watch for borewash at that stage

basicly it

just take in mind that most smallband lambda's are fluctuating like hell so it could be pretty hard to understand what the meter is saying. no throttle alternations help to smooth it out, it cabn be perfectly normal for the meter to show the whole band from 100mv to over 900mv in a second or two, so don't get scared over nothing.

DAkota digital makes nice dashmount meters with a computer inside to smoothen out the signal so you can read what it says. those are also very quick reacting for small bands

have fun

[eliot]

Narrowbands are better than nothing, but Wideband lambda's are very affordable now and i would strongly recomend using one of them instead.
For example, I have a wideband in one side and a narrow in the otherside (v8) - when on boost (12afr) i get a nice steady readout on the narrow, but on cruise i have it slightly lean, say high 14's approaching 15's - the narrowband doesn't even register a reading - which would make you panic if that was all you had to go on!

[diycerb]

Thanks for the replies

What I'm trying to do is use the O2 sensor in the gas analyser to calculate a reading of lambda. I'm not using a normal narrow band sensor or Bosch LSU equivalent. The O2 sensor in the gas analyser must be a wide range as it will read up to 21%. I expect its still zirconia cell but more like an industrial cell. Its an old Crypton analyser.

I've made circuits before that work off of a narrow band lambda sensor, 0-1v in lights a series of 10 LEDs. Very pretty but useless.

I've been given the gas analyser with a new O2 sensor in it, my idea was to make a small circuit to convert the O2 percent reading to lambda and then I've got myself a cheap wideband sensor. Very simplistic I know but its a start.

I have got a bosch LSU4 wideband sensor but the price of the readers puts me off. Another idea was to hook the LSU4 to a microcontroller and display the O2/lambda reading on a laptop screen. Laptop program can then do the calibration and display/log the results. For this job I need to find output characteristics of the wideband sensor I beleive its non linear and current derived???

Anyway anyone know how to convert percentage O2 to lambda?? Got a feeling lambda = 1 (14.7/1) is about 2% O2.

Cheers,
Rich

[rev-erend]

I have the info for CO (but not O2)..

0.1 % = 14.71 AFR
1% = 14.10 AFR
1.5% = 13.93 AFR
2% = 13.76 AFR
2.5 % = 13.55 AFR
3% = 13.37 AFR
4% = 12.99 AFR
5% = 12.63 AFR
6% = 12.24 AFR

I should get out more

[diycerb]

Thanks Alan, its a start, I'm sure it will be useful. Didn't think it was going to be this difficult!!!
Rich.