Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

Lambda sensor

Understanding AFR

Forum Posts



Tech Articles

Discussion and questions related to the course Understanding AFR

= Resolved threads


Ok so I'm a bit confused. I know we should use wide band lambda sensor, but Ive heard everywhere about boss welding it to manifold. how exactly. Its a £200 sensor and your gonna weld it to some manifold which is going to take some time.

Explanations greatfully received.


I just get my local exhaust shop to weld in an extra bung, $40 and 30 minutes later I'm tuning.

I need a picture of this lol as im defo confused. I know a fair bit about cars but the concept of this as a temporary tune thing has got me

Cool links cheers. But It confirmed what I previosuly thought. If your purchasing one to fit in your own car for tuning purposes thats all good, but if you are a pro tuner and tune dozens of different variants of vehicle this would mean A: you would need to have a link exhaust pipe prepared in your stock for every manufacturer of performance vehicle which is costly and B: is the customer gonna want you to drill hole in their floor which is also very time consuming.

That was what my question was all about - the logistics of using one in a commercial setting.


There are numerous ways to get sensors into the exhaust flow, the best way would be for the customer to prepare their exhaust as part of their tuning preparation. Another option would be a sniffer up the tail pipe.

You are over thinking this and over complicating it. Lets look at the initial problem you raised the sensor into the manifold. You simply get your local friendly exhaust shop to fit a bung into the exhaust pipe, problems solved. I posted the links so you could see you weld a bung into the pipe and not the sensor into a manifold.

Now lets talk about drilling holes in floor pans. I have never drilled a hole in any floor pan so I could fit a sensor. There are various length leads and even on my 4wd (Toyota Troopcarrier) I can run a lead through a grommet already in the firewall and hook it up to the meter. If push comes to shove I'd grab a longer lead and run it under the door sill and up through the window to the meter, problem solved and no drilling of floor pans. The exact same thing would happen if you were using a sniffer pipe stuck in the end of the exhaust pipe with a sensor attached to it.

As part of our tuning service at the shop I owned, we had a charge to fit an O2 sensor weld bung and plug to the exhaust prior to tuning. This was just a standard job we performed on any car before it went on the dyno and it means that any time the car comes back for touch up work, the bung is already there. I've also never needed to drill a hole in a firewall or floor pan to get an O2 sensor into the cabin - For temporary installation just for tuning this isn't necessary anyway, but in every situation so far I've been able to repurpose a factory grommet for the task.

Thanks Andre & guys

Ok so to update this thread based on my current understanding of lamda sesnor placement. Hope you guys don't mind but this is probably my stickiest point of understanding at the moment.

So we have welded a boss into the cars exhaust approx 500mm back from the cat so as to not confue readings , which would happen if attached to tailpipe as Andre said due to ambient gases, but of course also due to the cat in between changing any real values coming from the engine.

But here is my confusion. We place our lambda sensor connected to our lm2 in the bung we have welded into exhaust and obviously use this to take afr readings whilst tuning. But what happens to the actual cars lambda sensor which is needed to control the running of the car whilst tuning? Is our welded bung a secondary one fitted in unison to the original? If so how do we stop the bung after using so as it is sealed?

Am I on the right track here?

Martin UK

Have you placed your bung after the cat or before? It should be before but I'm not sure if I'm reading what you've wrote wrong as it sounds like you've put it after.

When tuning you should disable the cars closed loop fuel control and zero and fuel trims stored in the ECU.

Once you finish tuning you should always remember to reactivate it again.

No Chris it would be before the cat by about 500 mm. just getting it straight before I take on first client as dont wanna look stupid lol.

The rest I think I'll be ok with as I'm doing all the courses. But at no point in the courses does it explain in detail about the process of Lambda sensor fitment in a commercial setting.



Placement depends on the exhaust (manifold/piping) itself. Each port feeds into the collector so you want the sensor far enough away from the collector so the exhaust is a steady flow not a series of individual pulses from each exhaust port (which is what will happen at low revs if you have the sensor to close to the collector). Having said that there are some cars (mostly 4 cylinder Japanese cars that I have seen) with the narrow band fitted immediately after the collector.

When you purchase an LM2 you get a bung and a plug for each sensor the LM2 comes with. So initially you could simply use them. Any decent performance exhaust shop should have bungs and plugs in stock so that when you are finished tuning you remove your O2 sensor and fit a plug into the bung. If they don't have a bung available make up your own or purchase some hex key bungs that suit the thread and put never seize on it as you are putting it in.

Yeah cheers Michael. I actually found the bungs on a performance website tonight. Before I'd only seen the boss but now I see the stop bung it makes more sense.

Nice when the lightbulb comes on isnt it.

Martin UK

I think it is important that you should take some more time testing/experimenting on your own/your shops car before you take in a customers vehicle. You need to be fully comfortable with all the tools you are using and comfortable with the software before you take money for your services.

It might be an idea for Andre & Ben to do a course that goes into detail on setting up tuning hardware? How to properly strap a car on the dyno etc.?

@ HS Engineering, we were actually discussing this earlier in the week. A dyno operation course might be quite valuable. I know the move from hub dyno to rolling road has been quite a learning curve for me personally and there is a lot of knowledge I can share in regard to this.

It's on the list :)

I would expect AFR readings to not be as accurate with a sniffer than with one welded in the bung, however, I have been taught that in practice this is not the case and the readings do not differ enough to worry about. There are several forum posts claiming both arguments to be true. My EFI-U instructor taught there was no difference between readings.

I get the feeling @ conceptune was hinting at this in his earlier post and if there is a difference, should we strictly tune via welded sensor over tailpipe sniffer if accuracy is the goal or .....?

It depends on the type of tune that you are doing, if it's tweaking a running car then a sniffer is fine as the exhaust assessing be travelling fast enough to avoid environmental contamination. However, if it's a full tune the lower speed gasses around idle and low rpm areas can sometimes have their exhaust gasses contaminated my the pulses of the exhaust pipe pulling in clean air, I can't remember the proper name of it just now, making a false lean condition

I've tuned using a tailpipe sniffer as well as a welded O2 bung and my preference is the welded bung. The tailpipe kit in most instances works pretty well but you often see poor AFR readings at idle and low rpm where the exhaust flow is low and the sensor reading can be affected by ambient air. You can also struggle to get a tailpipe sniffer to fit into some exhaust systems that make an abrupt 90 degree turn right after the exhaust tip.

That being said, a tailpipe kit is normally pretty effective for reflashing where you're going to predominantly be interested in measuring AGR during WOT ramp runs.

What about the role of the catalyst? Are we not seeing enough of a difference in O2 molecules being affected before or after the cat? I was curious how much of a difference the post cat AFR accuracy was to the exhaust exiting the engine. I see how low air flow and the introduction of ambient air could also affect the readings but the catalyst efficiency was my target question.

Instead of posting another question, I'll add it here as well;

STFT & LTFT: Are the stft actually commanding adjustments that the ecu is making or is it just taking an error reading for the ltft to adjust for and is the stock O2 sensor calibrated for a stoichiometric ratio or does it vary a Rich/lean reading from the programmed AFR? Can the O2 sensor voltage be offset to work with whatever AFR is "in the box" to give an updated fuel trim?

What good is the stock O2 sensor if we are changing our AFR outside of stoichiometric? This concept is clouding my brain and keeping me from moving forward until I understand it so I thought I'd bring it up again.

There is a lot of dicsussion about the catalyst affecting the accuracy of the lambda reading if your wideband sensor is post cat. Years ago I tested this with dual Motec PLMs and LSU 4.2 sensors mounted pre and post cat and the difference in reading required me to use three decimal places to be able to see it. At that point I deemed the difference to be small enough that it wasn't going to significantly affect my tuning. Also consider that the lambda sensor will be affected by exhaust pressure which is almost certainly going to be higher pre cat than post - So what data do you want to rely on? Unfortunately we live in an imperfect world and compromise is the name of the game.

Most OE ECUs use narrowband sensors that are only useful at stoic. These sensors are used for the factory closed loop operation and the STFT is the instantaneous result of whatever the ECU is applying to the injector output in order to swing the AFR back and forth across stoic. These sensors work on a 0-1 volt output and generally the ECU will switch at about 0.45 volts. On this basis they are of no use other than closed loop operation at stoic. You can't target anything other than stoic with a narrowband sensor as it simply can't read accurately anywhere else.

Some OE application however do use a wideband sensor and hence can control closed loop fuelling across a wider range of AFR targets. In cruise and idle conditions though it's important to remember that for catalyst efficiency the AFR must swing back and forth across stoic.

We usually reply within 12hrs (often sooner)

Need Help?

Need help choosing a course?

Experiencing website difficulties?

Or need to contact us for any other reason?