Mogwyth

New Member
Well coming to the end of my tether with this motor now. Despite an exhaustive session on Testbook, a complete wire by wire test of the engine loom and all the sensors no faults can be found!

For those not familiar with the problems, we fitted a recon. 4.6 back in October, did about 1,200 miles ok, but then fuel consumption dropped of and we found the motor isn't going closed loop on petrol (or LPG)and has a hesitancy on LPG when accelerating, overall fuel consumption is 15% down on what is should be.

We have replaced:

  • ECU
  • All four Lambdas.
  • MAF
  • Engine coolant sensor
  • Plug leads
  • Plugs
  • Throttle position Sensor
Only one fault code is showing, this is to do with the Secondary Air Injection which has been removed and opinion is this is not relevant. All sensor readings are within expected parameters except the Lambdas which are mostly reading low.
 
Have you tried John Curtis at the Range Rover Register - he used to run his own service garage and I`m pretty sure still does and I think is still President of the RRR - it isn`t much to join and there are people in there who have forgotten more about RR`s than I will ever know. They have local meets so there might be one nearer you where you could pick their brains over a pint. And I`d be amazed if they didn`t have a stand/display at Billing if you wanted an excuse to go ?
 
Sorry, I didn't see this post!

What type Lambdas do you have?? There are two different types - one has two wires coming out, the other has six usually. The one with two wires are both earths, as they produce there own power. The six wire requires power.

Although they will be responding, it is perfectly possible that either a bad earth, or wrong connection may mean that the wrong circuit is activating at the wrong time, giving the ecu the wrong info.

To test this theory, does it drive ok warm? Only stalling or playing up at idle? If so, then the low tension circuit is either badly earthed or incorrectly wired. The same goes for the other way around - runs like poop, but idles fine would put the high tension under scrutiny.

You also have to look into timing - it the system set up for the engine, capable of the horrsepower (a low horsepower will not run a high power engine well, resulting in power fade at speed) is the engine timed to bias petrol or LPG? lpg would be better!

Look into servicing, especially leads and plugs - must be well serviced to run LPG as it requires a perfect running engine to work effectively.
 
Well coming to the end of my tether with this motor now. Despite an exhaustive session on Testbook, a complete wire by wire test of the engine loom and all the sensors no faults can be found!

For those not familiar with the problems, we fitted a recon. 4.6 back in October, did about 1,200 miles ok, but then fuel consumption dropped of and we found the motor isn't going closed loop on petrol (or LPG)and has a hesitancy on LPG when accelerating, overall fuel consumption is 15% down on what is should be.

We have replaced:

  • ECU
  • All four Lambdas.
  • MAF
  • Engine coolant sensor
  • Plug leads
  • Plugs
  • Throttle position Sensor
Only one fault code is showing, this is to do with the Secondary Air Injection which has been removed and opinion is this is not relevant. All sensor readings are within expected parameters except the Lambdas which are mostly reading low.

Like Vagrent said - Never Give Up!!

Anyway, I'm not brilliant on the 4.6 but if it was an old 3.9 I might be thinking about doing a compression test; dry then wet it might indicate sticky valve or something. (I know it is a recent re-con engine but always worth a check). The 4.6 ignition timing is controlled by the ecu, have you scoped the pick up (can't remember if it is a cam or crank sensor). Thats about all I can offer. Good Luck.
 
Thanks Guys, I don't plan to give up yet although I am getting grief form the OH. Got no chance of Billing I am taking the family down to Essex for a family do.

Auto thanks for you response but I think if you re-read my post we have covered most things in your post.
 
The fact the engine isn't going closed loop is the key to your problem I believe.

Now not actually playing with it did you buy a genuine ecu temperature sensor or pattern part?

There should be some temperature test results for resistance at differing temperatures-possibly look at rave for these or google as I have in the past.

Did you replace a 4.6 with a 4.6 or have you gone to a 4.6 from 4.0? if so and I don't think it is, but has the fuelling re learnt.

When you have tried all the cheap possibilities possibly have ecu tested
 
Cheers Fanatic, all sensors are genuine. As for ECU testing that's how we ended up replacing it, in an an attempt to cure the problem it was sent for testing and possible repair and never seen again!

It was an upgrade from 4 to 4.6 but I doubt it's the learning as for the first 1200 miles it was fine.
 
Cheers Fanatic, all sensors are genuine. As for ECU testing that's how we ended up replacing it, in an an attempt to cure the problem it was sent for testing and possible repair and never seen again!

It was an upgrade from 4 to 4.6 but I doubt it's the learning as for the first 1200 miles it was fine.


As a thought you have checked temp sender wiring to ecu plug is intact and low resistance with meter, i't strikes me as mighty strange that it does not go close loop-almost as if a wire had chafed through

Irish rover may know of a solution also
 
Last edited:
Have they tried to disable to the gas system to ensure it is not the gas ecu causing the fault one thing I forgot to test was does it go closed loop when on gas?

Just to help on the idea front when Bill came over to me we found that the lambdas were not switching at all but were only sending a flat signal back to the ecu but were receiving there ecu feed voltage.

From mechanical point the engine runs very well so I can say that mechanically all is well.
 
Last edited:
If your lambdas are mainly reading low at idle, then they are not responding in the correct way. This usually is down to a break in the wire. When you test, you inadvertently earth them in order to test. This will hide a break.

Start by testing the new wire on the sensor - not near the join, but near the sensor as these are new and can be presumed good - then the other side of the join - but not at the join. If the reading is the same, then work you way back up the wire.

When at idle, you SHOULD have a high reading - as they are effectively acting in the same way a choke does - they lean out when driving and warm, by changing the circuit.

You effectively are running lean all the time, so there is either a broken wire or the sensors are gone, as they are new, I would check for wire breaks or chaffs first. It they are not thermal powered, then you can bench test with a power source, and a hot air gun.
 
This is what we suspect but every sensor and wire has been tested and shows correct.

Just a thought and I'm sure you checked ecu supply voltage and wiring- a most strange fault.

Failing that is is possible gas conversion ecu is causing error, I'd be tempted to warm it up, remove engine ecu and see what reading you get at ecu plug with ignition on on temp sensor wiring
 
Last edited:
Stu, no they haven't, when I suggested they didn't quite laugh but said it can't, all the LPG ECU does is switch the petrol injectors on and off and fuel the LPG injectors based on the signals form the main ECU.

Just an extra note you can make the Lambdas go rich by restricting the air flow.
 
But as we found out when I played with it if you disconnect the lambdas they are not switching and a false signal is being given I do believe the gas mite be the problem
 
As a suggestion the gas system is piggybacked on petrol ecu wiring, therefore if you have tested and changed the petrol side and they are fault free.

I'd suggest it is looking at time to disable the gas system, depending how much loom you have on gas system it may be easier to link it out there-rather than pulling loom apart to get to splices.

If you have or can find a wiring diagram for gas ecu it would make things quicker.
Or you could pull appropriate wire to lambda and see if the values out are correct when not connected into ecu wiring
 
Well I will investigate this further.

Interesting I have done 101.4 miles of steady 60-65 mph motorway work today on LPG with the lambdas disconnected for a total of 21.75 litres which I make 21.66mpg, although I still have the hesitation! What is happening is the ECU is now using a default value of .45 for the lambdas instead of .1 they were sending to it.
 
Snore all you want mate, if it's meant as a sarcastic comment come out and say what you mean or don't bother reading it if it bores you!
 
LPG ECUs are easy to check.

You need to go to a dealer for the LPG system and ask them to swap the LPG ECU.

How old are the LPG injectors? They could be malfunctioning.

Lambda sensors and ECU work together in a feedback loop, where the signal/mixture, if rich, tells the ECU to lean out, which changes the mixture leaner, which the lambda sensor picks up, showing the voltage output and when it goes past 0 volts, which tells the ECU to richen, which the lambda sensor monitoring the gases feeds back through this cycling voltage and therefore mixture. If all sensors are working, crank, lambda, temp, airflow, etc, and wiring, you need to look at the ECU again.

It could be the second ECU that has a problem, even though it has been replaced.
I know of someone who replaced a coil, and the new one only worked for a short period.

If the main ECU is bad, the LPG ECU won't be able to make any difference. The main ECU controls the fueling with respect to lambda feedback.
 

Similar threads