Wideband 02 UEGO Air Fuel Ratio Kit

I’m currently running a duel fuel setup on my XJ6 S3 1982. It’s a simple open loop Impco mixer inline with the standard AFM.

I had the LPG and petrol mixtures set by someone who had a Snap on exhaust gas analyser, and set the CO level to around 2.5%. Unfortunately, after that I had to replace the airflow meter and retune both petrol and LPG. I’ve also moved a long way from the mechanic with the gas analyser and it’s hard to find anybody that can work with LPG. Currently there is a lockdown here and looks like it could last a long time.

So I’m looking at doing it myself, by using something like this;

Anybody used these type of gauges, if so can they recommend a specific model and whereabouts on the exhaust to fit the bung for the oxygen sensor?

p.s are these sensor any good?


I assume you run on petrol or(!) LPG, Zenit…?

2,5 % CO is above petrol spec - which varies, with market, from 0,5 up to 2%. I have nos specs for LPG, and the immediate question if you know how to adjust mixture(s). If so, the fuel ratio kit, if suitable for LPG is fair enough at a guess. It doesn’t mention is it also measures HC and NOx…?

xj6 85 Sov Europe (UK/NZ)

Hi Andrew,

I’ve seen AEM recommended by others, and would be happy to try them, but I have used Innovate wide-bands myself. Sensor positions indexed in the in upper portion (9 - 3 o’clock) are usually recommended, apparently to reduce the risk of condensation hitting the sensor, perhaps implying that they may not be as robust as narrow-band sensors. Sensors are also supposed to go after the turbo and before the cat, if either apply.

Thanks Colonial,

If I select the lambda mode can I switch between lpg and petrol without have to program into the gauge the fuel type? Is that the idea of lambda readings, I don’t have to define stoic points or multipliers?

Hi Frank,

Setup is dual fuel, i can switch between the two fuels while driving using an electronic switch inside the car.

I think the meter only displays AFR or Lambda values. No CO2, CO or HC.

I’ve worked out how to adjust the lpg mixture. For fine adjustments there is an air bypass screw on the mixer, similar in concept to the AFM screw adjuster. For larger adjustments I have to disassemble the mixer and insert leaning ring shims to let more air through, which is fairly straight forward.
And for WOT there is an adjuster on the gas going into the mixer.

Yes - measured lamba is expressing your deviation from stoic regardless of fuel type. See this:

Microsoft Word - 11-0053 LC-1_Manual_2.0.doc (innovatemotorsports.com)

Note on P8, if you want to use AFR instead of lamba in the display, you need to reset the fuel type accordingly.

If you search the interweb, you will easily find tables of %CO and AFR or lambda equivalence.


Thanks Simon,

Can the sensor be temporarily clamped into the exhaust tail pipe, as opposed to fitting an exhaust bung? Or would the results be questionable ?

Have you had a look at the exhaust manifolds? Some have ready made threaded inserts for O2 sensors. I’m not sure which models have it, but usually Series 3.

Thanks Nick,

I think some s3 had them and some didn’t. Will have a look, would make things simpler if it’s there.

Andrew - To be useful the sensor has to avoid contamination from outside air, so likely questionable. But taking a “can-do” approach (you have the wide-band kit, you’re in lock-down, the usual rules don’t apply) maybe you can arrange to run the sensor up the tailpipe as far as the sample tube of the gas analyser went. O2 sensors have quite high grade insulation on their wiring…

Further to Nick’s question, does your car have a cat converter?

Simon and others have well covered questions you raise, Zenit - but some additional remarks…

It needs to be further stressed that the Lambda technology is based on detecting the presence or absence of oxygen molecules (O2) in the exhaust. Ie, lambda=1 means complete combustion; all fuel is burnt - no O2 or residual fuel present. Old technology, carbs and EFI, is unable, neither quick nor precise enough, to maintain this throughout all variations of load and throttle settings. So ‘our’ ECUs reads O2 from the O2 sensor - and continuously fattens the mixture if O2 is detected, and leans out if no O2 is detected.

And also; lack of fueling precision dictates that engines are set up to run fat, primarily to avoid the mixture going lean at any time when throttle settings are altered. And during throttle variations the Lambda circuit is ‘disabled’ to avoid this. So when using AFR its necessary to ensure that your mixtures setting conforms to the prescribed CO petrol spec settings - usually somewhere between 0,5 - 1% for EFI, and 2 - 3% for carbs, ref Simon’s comments.

Also, since Lambda relies on O2; there must be no air contaminating the AFR sensor - the exhaust tail pipe is not really the safest place. The best place for the sensor is one of the bungs on the exhaust downpipe, originally used for exhaust back pressure testing, but also used for the O2 sensor - which of course has the same ‘air free’ requirements as the AFR sensor. As the O2 sensor must(!) be disconnected while fuel mixture is adjusted - and obvious solution is to replace the O2 sensor with the AFR one. However, the 02 sensor must be heated to some 400C to work - and is placed accordingly…

xj6 85 Sov Europe (UK/NZ)

Unfortunately, pipes run straight though. No cat or sensor in exhaust system.

There is not cat on this model, 1982 s3.

They should be on the cast iron headers, before the exhaust pipes.

Hi Nick,

Are you referring to the 2 plugged ports as in this picture with the rusty header?

Yes indeed, those are the ones.

I downloaded the manual for UEGO AFR kit that I’ve ordered. They recommend placing the sensor 18 inches downstream of the cylinder head’s exhaust port, and as far away as possible from exhaust tailpipe. So I won’t try the sensor in the tailpipe idea.

So would one sensor in one of the ports work ?
The manual for the kit says to place the sensor 18 inches from cylinder head exhaust ports, do you think the header would get too hot for the sensor, possibly damaging it?
It uses a Bosch LSU4.9 sensor.

My 3 Innovate WBO2 sensors are placed the same distance, they work fine. I only have them there while tuning, not all the time.