It's a given that narrowband O2 sensor (Lambda) output swings up and down within the 0-0.9v range, but everywhere (on the 'net including here) I am seeing people say this is due to the ECU in closed loop in effect firing the injectors rich/lean trying to maintain stochiometric mixture but overshooting to some extent.

Whilst this may be true in many cases, I am aware that in other cases it is not. I know this because I have monitored a narrowband sensor on a carbureted V8 (no ECU at all) and it also displays this behaviour.

The frequency follows engine rpm and rises to many times per second.

Does anyone know why?

  • Welcome to Motor Vehicle Maintenance & Repair! Sep 14, 2018 at 13:55

5 Answers 5


The O2 sensor changes its voltage for two reasons. One it's hot (and you have to wait until it gets hot - hence modern ones tend to be heated) and two, the gasses flowing past it. If you disconnect the computer and take a high impedance volt meter (most digital ones) you can measure the O2 sensor output directly.

The computer simply reads the voltage to determine the current air/fuel ratio (or lean vs. rich).

As for swinging back and forth. To some extent this is because exhaust flow is pulsing (for each cylinder). In theory, if you synchronize, you can actually tell which cylinder is lean vs. rich.


On an injected motor, as you said, the ECU adjusts the mixture based on rich / lean feedback from the O2 sensor, throttle and other inputs. On a carb system, this is controlled by adjusting the carb, to achieve the desired mixture (though obviously this doesn’t change dynamically while driving). Even without any closed loop feedback, the mixture will be rich at times and lean at others.

The difference is that, in the EFI system, the computer attempts to correct for those changes to get closer to the target ratio.

  • Well yeah, but doesn't really answer the question. This is not drifting lean/rich, it is a very regular pattern occuring in time with rpm. Once or twice per rev it looks like. Must be something to do with exhaust valves opening/closing I reckon.
    – Ian Cope
    Sep 14, 2018 at 17:09
  • @IanCope Even without feedback, it's going to be cyclical, but I'd expect the period to be fairly short on a V8. :/
    – 3Dave
    Sep 14, 2018 at 18:00
  • The question is why :-)
    – Ian Cope
    Sep 15, 2018 at 10:03

I am guessing you don't have this carb measured data available to share with us? How far downstream did you take lambda measurement?

The simple facts are this:

  • When there is no oxygen present at point of measurement the narrowband will read ~0.9v
  • When there is plenty of oxygen the narrowband will read ~0.1V

You say that the sensor is switching in time with RPM, I wonder if possibly a number of cylinders in this engine were running rich and others running lean. The closer to the exhaust manifold you measure I suspect the less mixing of the various cylinders exhaust gas will happen and this will result in you observing the different lambda mixtures from individually firing cylinders.

  • I don't have any data recorded no. The sensor is about 6 inches downstream of the collector. Craig I like your theory. It's a 2 barrel carb on a dual plane inlet so it could be the carb is not quite evenly balanced, and/or imperfect mixture distribution in the runners. If I play with the balance of the idle mixture maybe I will see if the fluctuation will increase or decrease (at idle anyway).
    – Ian Cope
    Sep 17, 2018 at 8:11
  • agreed, try that I suspect you will certainly see some change
    – mx5_craig
    Sep 17, 2018 at 10:04

The oxygen sensor reads the amount of oxygen and thats it. They don't cycle on there own without the level of oxygen cycling.

So the question should be "Why is the oxygen level cycling in the exhaust, if the o2 sensor it is not connected to anything at all?"

If you remove it and test it with a torch, I'm sure your voltage will behave the same way as this video


  • "So the question should be "Why is the oxygen level cycling in the exhaust, if the o2 sensor it is not connected to anything at all?" " - Assuming it's working correctly, yes I agree that's the question.
    – Ian Cope
    Sep 17, 2018 at 7:50

The sensor is pulsing because it reads the oxygen level of the air in the exhaust, which changes every time a cylinder fires and pushes gasses through the exhaust. The flow out of the engine follows the cylinders firing, it's not a constant stream, so it pulses with the engine.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .