The way to sort this out is to get it on Testbook and drive it up the road whilst watching live data.From the fault it does sound very much like its running lean in the part of the fuel map where its supposed to be closed loop.Motronic seems to have a very narrow range of injector pulse width for a given rpm/throttle opening/MAF reading.If the MAF reading is too low it simply cant provide enough fuel to make power,switch the oxy sensors up to rich and get them cycling.The ecu will soon give up and stop trying to adjust the short term fuel trims and go open loop.This ends up with low power or spitting back/hesitation etc.
Whacking the throttle wide open just tells the ecu to go into the area of the fuel map where it ignores the oxy sensors and runs open loop - hence being able to drive through the fault.
I may be wrong about the low MAF reading,but someone who knows what to look for on live data displays will fix this for you.
 
What do you say about the 0.35 fuelling? If this is the lambda value, it's rich not lean.

I had MAFs go and you notice at all engine speeds, and you get pinking. If he has tried a different MAF, then two shouldn't have the same problem, unless it's the wiring.

With the hot surface/hot wire type MAF, the surfaces get coated and chemically reacted because they are hot and slowly get more insulated, so they don't get the airflow cooling the wire/surface as much as when new. This will give a lean mixture because the MAF surface is not being cooled by the air and the resistance will be higher (hotter surface/wire), so the ECU "sees" a lower airflow than is actually there.

Don't MAFs loose their sensor ability to detect airflow slowly over time, not in such a short period, unless he's running without an airfilter?

Are there some sensors that the ECU can work out if they are wrong, e.g. lambda sensors, but with the temperature sensor, what else does it have to say that it has failed, the MAF sensor also? When and how does the ECU recognise a fault?

Will the MAF sensor stop it going closed loop? I thought it was only dependent on temperature sensor and throttle position/rate of change of throttle.
 
Have you pulled a plug or two? If it's rich or lean, you'll soon tell.

The pinking (MAF failure or TPS) is due to the lean mixture. Lean mixtures are more likely to pre-ignite than rich mixtures, hence the need to keep mixture stoichiometric or rich.

Temperature sensor failure - open-circuit gives rich mixture.

Lean mixtures run hotter - clean plugs (shiny white porcelain insulator)
Rich mixtures run cooler - dirty plugs (plug cannot sustain self-cleaning temperature)

When I've had a MAF failure - giving pinking, the fuel economy was unaffected, just didn't like the throttle being opened too quickly, could cruise easily.
 
Final point, does the ECU use lambda sensor during cruise as the primary sensor, but during acceleration, the MAF and TPS become the primary sensors?
 
Right replaced MAF yet again and also replaced Engine Coolant Sensor.

Made no difference.

I have now got a diagnostic setup on the laptop. Only faults showing are P0304 Cylinder 4 Misfire Detected (happened this morning) and P0413 Secondary Air Injection System Switching Valve A Circuit Open, could this be the problem? The old engine had SAI, this one doesn't I had been led to believe it wouldn't be an issue but........ and if it is what can be done about it.

Still getting to grips with the software (PCMScan) but:

O2 sensors front two (same readings on both):
Capture2.jpg



Rear 2 (same readings on both):
Capture2a.jpg


Fuelling shows: Open Loop Fault.
 
Last edited:
Have you got an SAI switching valve?

Don't plumb it in for the air, try just connecting one up so it is there electrically, but doing nothing. It might make the computer happy, instead of having a problem.

The next problem that could occur is if the fault sensing side of the ECU actually measures the performance of the SAI. If you put one in place, will it then throw a code to say performance "out of range" - you would have to look at the fault code catalogue for that ECU to see if it might be a possibility.
 
Last edited:
Having a look at the rear sensors, they should have a time between sensor transitions.

Have you replaced the rear sensors? - sorry just checked previous post and you have.

If the rear sensors are telling the ECU it's lean, the rear sensors might shift the long term fuel trims, which you say need resetting every few hundred miles.

If the rear sensors cannot get into the sensor transition range, then it might not go closed loop.

You will have to check on rich/lean for the high voltage - might be rich.

The feedback from the SAI valve might need to confirm that it has closed to ensure closed-loop operation occurs.
 
Last edited:
Mogwyth,I'm afraid your figures dont really mean too much as they are just a snapshot in time,you really need to watch in live data whilst a helper drives the car to create the fault.I think you will find that they will go from a base value say of 1 in steps up to 1.2 - 1 and then freeze because they cant go any further,then the ecu will default to open loop.This will happen over a few seconds during acceleration.
Raving man you need to study a bit more,the downstream oxy sensors do nothing at all to fuel trims,all they do is monitor how well the cats are working.Under closed loop fuelling at cruise they should just hover at around 420mV output.If they cycle like the upstream sensors should be then they are just telling the cats are having no effect on the exhaust gases.
The .35 value you asked about would not be a lambda reading as there is no way that the narrow band sensors can read and give a plausible value to a mixture that rich.
The last point I would make is that LR products are well known for being difficult to get generic diagnostic tools to work on.That value of .35 could be a real value,it could be total rubbish.Where the data cannot be double checked with either a multimeter,scope or Testbook you have to take it with a pinch of salt.
 
Thanks.

Yes as eightinavee says the rear O2 sensors are just a check.

I will try and get out and get some live data.

Regards

Bill
 
Theoretically they are just a check, but with some cars, eventhough manufacturers will say all they do is check the catalytic converter efficiency, they still act on the fuel trims. You may be completely correct on this system, as there are differences between the systems.

On the live data, it would be worth checking the fuel trims after they are reset, and then see what happens over time as the fuel trims are shifted again.

If the primary O2 sensors stay within range, then they obviously aren't affecting the fuel trims, which would only leave the secondary sensors that are causing the problem.

If the primary sensors start to show that the fuel trims are shifting, but cannot keep them within the normal range, you would have to then ask, what is causing the fuel trims to shift?

As far as I know only the lambda sensor data is used to modify the fuel trims.

Shouldn't the voltages be cycling around 0.5V, as a normal voltage average?

Did you have the engine working at normal temperature? It's almost like the thermostat is stuck open, over-cooling the engine and preventing closed-loop operation, or fuel pressure regulator stuck and giving too much pressure. Again only the fact that open-loop operation is occuring will be seen.

If a catalytic converter converts NOx and HC to N2, H2O and CO2, then the secondary sensors should also cycle, around the 0.5V, but not quite the same as the primaries. Perhaps they do settle at a lower voltage, but 0.615V is too high.

With cruising, this should almost show up just revving the engine and holding it at different engine speeds in neutral. Whilst there is no load, you still have to open the throttle and the engine management and sensors should all be doing what they do normally. On the MOT test, they do the "fast idle" between 2500 to 3000 rpm to check that the lambda sensors and engine management are working properly controlling the emissions.

On the 0.35, did the garage stick an exhaust gas analyser on it and look at what the emissions were? An independent lambda monitor/gas analyser such as using a rolling road does make it easier to see if the primary sensors are even in the right ball-park.
 
Last edited:
If you can get an independent gas analyser to look at the gases when the trims have shifted, see if the CO has increased greater than 0.2% (and lambda less than 0.97) - rich and see what the primary sensors and secondary sensors are telling you.

On the voltage, it does depend on the type of sensor. Bill you would have to let us know the exact make an model of the sensors to ensure we can look up what they should be doing.
 
With the sensors, if you got the secondary sensors fitted where the primary sensors should be, and the primaries fitted where the secondaries should be, then depending on the sensors if they are zirconia or titania type, and they could cause a problem for the ECU.

The ECU might not know what you have done and may try to adjust the fuel trims with the data it is getting, but the data will be wrong.
 
Last edited:
If you are really unlucky with the engine build, there might be silicon poisoning of the lambda sensors. That screws them up pretty quickly.

Did anyone squirt a load of WD40?
 
The lambda sensors and ECUs changed for different years. What year is your ECU for and do you have the correct lambda sensors for that year of ECU?

You did say about going from SAI type to non-SAI type.
 
Last edited:
Raving man I'm not sure your posts are helping.downstream sensors are only used for monitoring what the cats are doing,it is the job of the upstream sensors to tell the ecu what the gases are doing fresh from the combustion chambers.Fuel trims alter on ALL closed loop petrol engines as a result of all the variables,which is why they are so useful.The ecu has a setpoint programmed in it which the ecu will work away from in the short term - in whichever direction it needs to,setting long term trims after all the drive cycles have been done.
Using an exhaust gas analyser is not a big help as they lag,being quite slow to respond - which is why I suggested using live data on Testbook or similar.An analyser will tell you nothing about fuel trims,just the make up of the gases coming out of the pipe - not how they were made.All of the sensors on a Motronic DII are 0-1v Zirconia so will do the same job,from memory it would be hard to swop the upstream sensor wires side to side,cant remember about the downstream - but they wouldnt affect the trims anyway.
 
Sorry that you don't feel my contribution is helping.

With any engine diagnosis you have a load of sensors that tell the ECU what is happening and then the ECU supposedly does it's calculations for fuelling and ignition timing, operating the injectors and coils.

Beyond that either all components are working well or one or more has failed, or the wrong component has been fitted or not (if someone is adapting the system to a different engine or vehicle) or the wires between each of the components of the system as a whole.

If Mogwth has checked every single component - new or working within spec, then we get to an ECU or wiring fault.

If Mogwth doesn't know the specs, then I am prompting him to look at each component and the spec that it should have, or if something has happened or has been done - like WD40 on the lambda sensors.

Put it all together and it works....shouldn't it?

If it comes down to the ECU, you could hypothesise that the ECU is setting the wrong fuel trims, causing a problem for itself, which wouldn't be the first time it has happened.

Human beings have the same problem, they also create unnecessary problems for themselves, but that's life.
 
You could look at the EGR valve to see if that has failed.

It could be open too much, knackering the combustion, of course eightinavee can correct me if I am wrong. With too much EGR it won't matter what the lambda sensor says, because poor combustion will always result in high HC/CO and more O2 because of the incomplete burn.

Isn't around 2,000 rpm on cruise the range where EGR happens? This should normally be controlled. If it uncontrolled.....what happens?

Wouldn't this give full power on petrol and LPG.

Many engines require the EGR valves to be cleaned because they get stuck - on petrol and diesel engines.

And if the O2 is constantly high, won't that mean that the ECU will keep correcting the mixture to richer and richer, trying desperately to do what it is programmed to do, but won't know that the EGR valve is stuck open, unless the fault detection system is absolutely fully operational and able to detect such a failure.

With the fault detection, would it show up a misfire because of the poor combustion? and then will it also say that it isn't actually going closed loop?

With the LPG, if the regulator / valves aren't working properly, then the LPG ECU might need recalibrating. If it is too far out, it could start to affect the petrol ECU.

All I can do is ask questions and give hypotheses and make suggestions, until some feedback confirms one of these to be appropriate, then we can expand in that area.

And if you think that replacing a part will make it work, I know of someone who had a problem, replaced a coil with a brand new one, and the same problem appeared - the new coil was duff as well. Moral - double check everything.

With lambda sensors, it might tell you what the voltage output is from the direct drive data, but then if it is rich or lean, it won't tell you WHY!!!!!
 
Last edited:

Similar threads