3.5 V8 with a 3.9 injection conversion runing to rich

This site contains affiliate links for which LandyZone may be compensated if you make a purchase.

Awkward

Member
Posts
60
Hi, I have a Defender 3.5 V8 with a 3.9 Hotwire injection system that was running like a dream until it was left standing for 6 months. Took it for a MOT and the Co2 on tickover is 3 time over the limit. Found the MAF sensor was not working so got it replaced, stepup motor is ok, has a new engine temperature sensor, has a clean air filter. Just wondering if the throttle sensor would make it run rich on tickover...?
Something very strange is if I start it and don’t touch the Excelerator and take a Co2 reading it’s well within the limits but as soon as the engine is reved up the Co2 goes off the scale... Any ideas anyone please.... Andy
 
If the TPS is out of spec or needs adjusting, does the ecu think the throttle is more open than it really is? It might then add more fuel because it thinks you're trying to accelerate?
 
That was my next item to replace but I can’t find one anywhere. If you know of one or if anybody else knows of one for sale can you please let me know, it would be much appreciated....
 

Attachments

  • E3CBDE49-873F-4DD7-9D2C-6A9EDD5071E2.jpeg
    E3CBDE49-873F-4DD7-9D2C-6A9EDD5071E2.jpeg
    301.4 KB · Views: 205
I don't know sorry, but can you test it? I'm assuming it's a simple potentiometer so you should be able to connect up to a multimeter, ideally an analogue one and see if it's in spec?
 
Il try and find some figures on the old interweb and see what the readings should be.... Fingers crossed...... Thanks
 
Have you checked if you have lambda control? if the O2 sensors are duff and the ecu isn't getting the right input it will go in to limp mode which is a really rich basic fuel map. If it is related to lambda control, it will be fine while cold and also at wide open throttle. Once warm it will possibly not idle because of an over rich fuel condition. Which tune select resistor is installed?

Limp mode can be triggered by a number of sensor errors. Speed sensor, Air flow meter, O2 sensors, throttle pot, fuel and coolant thermistors. Luckily almost all of it can be tested with a multimeter and if you do a search you should find most of what you need, I know over the years I have variously posted all sorts of info for dealing with misbehaving hotwire efi.
 
Hi Kev, mine is a very early injection system with no lambda, O2 or speed sensor’s. I’ve checked the readings off the MAF (& set the O2 screw), throttle sensor, water temp sensor top of engine, fuel temp gage on front of fuel rail, fuel pressure, step up motor and all are as should be, even swapped 2 ECU’s around with no joy.... Just taken the injectors out to be tested incase they are leaking, can’t think of what else it could be....
 
Ok first of all, all EFIs require a speed sensor, early engines it was called the speedometer which has a transducer as the speedo drive or added to the drive just for the EFI so two types of vehicle speed sensor both are known to fail.

You say you have replaced the coolant (water) temp sensor, hopefully that was the one for the EFI and not the dash gauge, as the EFI temp sensor is known to fail and while the engine idles ok when the engine is hot, but the engine is really lumpy when driving the car, I had a couple sensors fail in my Austin Rover Days and it isn’t gradual but instant failure.

It may be something simple, as it’s strange that an engine that’s running “like a dream” and then laid up has the issue u have 6 months later, usually it’s ‘it won’t start’ although altering the o2 screw is interesting, that’s set in the factory and then capped never to be touched again it normal use.

EFI diagram fo the disco but is typicals :)
 

Attachments

  • FCE8071A-100A-413F-B897-2E2F6AFA010F.jpeg
    FCE8071A-100A-413F-B897-2E2F6AFA010F.jpeg
    393.4 KB · Views: 397
Last edited:
The 3.9 injection conversion was done before I baught the Landrover, when I got it the exhaust was clean with just about zero soot and the spark blugs where the perfect biscuit brown. Have done a huge amount of research and aperantly if you run the early set up with no lambda or speed sensor the ECU will stop looking for these items once it can’t find them but you do need to reset the ECU... Unfortunately I don’t know if the injection system parts even came off the same vehicle but apparently they do have to be synced with each other the same as they do when leaving the factory. My MAF was just over 1 volt out which mean s it probably came off a Rover car that runs on different values... Not 100% sure of course as there are so many threads giving different info but when I get the injectors back I will run every sanareo once more just incase Iv missed something and let you guys know the outcome.... Any more info out there will be gratefully received ....
 
I don't think you understand this injection system. 14CUX needs speed transducer - it is part of the idle control circuit. The components of the injection system do not need to be synced to anything, the ECU isn't that clever. The only clever thing it can do is look at O2 sensors and apply fuel trim. The ECU will always look for lambda if it has a cat tune select resistor in the loom.

An ecu reset is really easy - just pull the plug off the ecu to remove power, re-attach and job done. If you prefer, you can always disconnect the battery wait a second or two then reconnect.

Your MAF setting - this can vary and is best set using a gas analyser.

Have you considered if the problem is something else - a leaking gasket for example?

Based on my own trials with the venerable injected RV8 I doubt the injectors are the problem.

Let us know what's what when you get the injectors back and then maybe we'll see if we can troubleshoot it from there but be prepared for some frustration along the way. Also, it might be useful to know where in the country you are.
 
Hi Kev I’m down in Dorset.
I have a (Blue) resister and Iv descoverd the ECU is pre 14CUX.
Iv been in touch with a Rover V8 specialist that builds these & other engines with carbs and EFI systems including the ECU internals / programming. U can buy parts or complete just turn the key engines from them.
They gave me the list of the things I had to check and how to adjust the verious parts and this did make a huge difference when done until the spark plugs got sooted up.
I told them I had no speed transducer no cat & no lambda sensor and they said it didn’t matter I didn’t need them and if I did have them they wouldn’t play a part in causing my vehicle to over fuel once the excelerater had been touched.
When starting the vehicle if you don’t touch the excelerator the HC & CO levels are spot on, soon as you just lightly touch the excelerator the HC stay the same but the CO goes up to 8%.
If you turn the CO adjusting screw on the MAF when the vehicle is ticking over it does alter the CO levels until the excelerator has been touched.
After the excelerator has been touched this screw has no effect on the CO levels (even on tickover) until the engine has been turned off & restarted.
I have tried this with 3 MAF sensors and they all do the same.
Definitely no air leaks and air filter is as new...
I don’t think it’s the injectors ether because I’m thinking if they where leaking or giving a horrendous spray pattern I would have over fueling at tickover but it’s another box ticked.
 
Hi,

I'm not sure this makes complete sense. You said this was a 3.9 efi conversion. AFAIK all 3.9's had hotwire injection, this would be the 14 CUX ecu and lucas hotwire air flow meter. Can you post a photo of the ecu and air flow meter, that should tell the tale.

If you do have a 14cux ecu, the tune select resistor is wrong for non-cat tune.There was never a blue tune select resistor for the hotwire system. I can only conclude you don't have a 3.9 injection system or somebody has been messing around with it.

Ignoring all of that and anything to do with lambda control, the basics are present and correct - fuel, spark, BANG! At idle, you say all is good then when you open the throttle you get over fuelling. So you need to look at things that tell the ecu how much fuel to pour in. Air flow meter is primary input. Throttle position would be next together with engine speed. Other potential culprits could be coolant and fuel temperature sensors.

If you don't have a hotwire air flow meter I can't help you as I don't know how to test the early flapper units. Next, throttle potentiometer, carefully check the wiring and the connector, then get the multimeter out and check the throttle pot for resistance and scaling - I have posted how to do this so do SEARCH.

Temperature thermistors can again be checked with a multimeter, check the resistance values. If you want to fool the ecu, bridge the connector with a paper clip, ecu will think engine really hot so won't add any extra fuel by way of cold start enrichment.

How's your ignition system? Check the ecu has a good clean run input from the coil, comes off coil negative. Is the ignition timing correct? You might want to think about maybe doing some other mechanical checks/tests and I would say maybe look at vacuum gauge test and a compression test. This will indicate if there are possible issues with valve train.

HTH
 
I have that link and gone through it already.
The resistor should be green but I was told it wouldn’t cause the problem that I have.
When I originally bought the vehicle it ran perfectly, I have the original MOT slips that show the CO HC readings. The exsaust was clean as a whistle, pulled like a train.
 
Please post up pics of:
A: your ECU
B:your air flow meter
If possible, can you post the following:
CO trim voltage
Throttle pot closed voltage.
Fuel rail thermistor resistance
Coolant temp. thermistor resistance.

Where is your ignition timing set to? and what compression ratio engine is this on. I understand this is a 3.5 not a 3.9 - is this correct?
I have re-read the first post and I would strongly suspect throttle potentiometer is out of adjustment and causing a fault putting the ecu in to limp mode which is a very rich tune that could easily account for the high emissions.
 
Got bad signal can’t upload the pics, the ECU looks like a14CUX but the serial numbers aren’t right for a 14CUX. I had to send a picture of the internals to the V8 engine builders so the could work out which software it was using. (They called it a X22 early software)
The MAF is off a 3.9 setup I’ve checked the part numbers.
Iv done everything you have suggested and all is fine, all the values are correct.
When it’s on tickover after touching the excelerator and overfueling all the sensors are giving the same readings as when it’s on tickover running properly with the correct CO at .80%
 
Back
Top