Disco 2 Rough idle in a V8 D2a — tried MAF, O2, sparks, leads

MAF Diver said:

Thanks everyone! Here are some replies/next steps:

My mechanic said he looked visually but doesn't have a smoke machine. I tried making a homemade smoke machine with a cigar and a siphon, but it didn't work

Yes, tried a few tankfuls of premium 98 and still the same issue. (Running E10 94 most of the time, but also "premium" 95.)

Idle control valve: couldn't find a video or guide that covers how to remove and clean it.

Is it as simple as this? (1) disconnect battery, (2) disconnect hoses from ICV, (3) remove it, (4) clean with brake cleaner and wait 1h for it to dry, (5) refit everything and go?

Or should I just buy a new one?


For that first error code, how would I troubleshoot the wiring or ECM? (I've tried 6x different O2 sensors and all end up getting this error.)

Thanks, I'll try a fuel injector cleaner. They've got this in AUS for about £3.

Yeah, under load it seems fine.

Once warmed up and while driving, the OBD2 Car Scanner app reports fuel system status:
- ECU #1: Closed loop, using oxygen sensor feedback to determine fuel mix
- ECU #2: Closed loop, using oxygen sensor feedback to determine fuel mix

While stopped at a traffic light, fuel system status:
- ECU #1: Closed loop, using oxygen sensor feedback to determine fuel mix
- ECU #2: Open loop due to system failure

What is this second ECU?


Just to confirm, is that the vacuum leak smoke test the others are suggesting?


After unplugging the battery and driving around for 40 minutes:
Long term fuel % trim - Bank 1: 17.19 %
Long term fuel % trim - Bank 2: 17.97 %

Yeah, it wobbles sideways at idle! Not hugely but noticeably. Funnily, very occasionally it doesn't, especially when the system is reporting both ECU1 and ECU2 running closed loop. But it lasts only a minute and then it starts wobbling again.

My cheap OBD2 scanner app can't seem to report/show misfiring codes even if I had them. I'll borrow a more advanced tool, but that won't be for a week or so. I wonder if there's an iPhone app (can be paid) that can also show this? Shouldn't everything be the same over OBD2 for all devices?


Thanks, will do! Nobody on Facebook/forums near me has a smoke machine and my DIY one didn't work. But I'll manage!

Click to expand...

I can only "ASSume" the reader says ECU1 for bank one and ECU2 for bank 2. That does narrow it down a bit, if the fault is generally on bank 2.

Those Long term trims suggest it's running similarly on both banks, when driving at least.

Vacuum leaks affect idle sooo much more, because the engine side of the butterfly is at high vacuum when the throttle is closed, so the pressure differential sucks air through tiny holes. When driving, the vacuum drops and the air takes the easiest path - generally in through the MAF route and any small leaks make very little difference at this point.

*IF* that wobble is from an ignition or compression misfire at idle, it can cause the out of range O2 reading - if one cylinder isn't firing, it doesn't burn it's share of the oxygen, so there is too much in the exhaust on the bank, and the ECU *thinks* it needs to add more fuel... but it doesn't. And it doesn't matter how much fuel is in there, it can *only* read the O2 level. You can try running it with one spark plug lead unplugged at a time, and see which cylinder doesn't make it run worse - that's probably the misfiring one. It does depend if the misfire causes the wobble, or if the fuelling being altered causes the misfire - it can go either way.

A compression test is always a good place to start though. Always do the basics, before getting too complicated... But I'd have thought your mechanic would have done that as a start, if he's doing all the work? Maybe he's not the best option???

Vacuum leaks are still an option too - anything that gets in without the MAF seeing it, is a leak. Make sure the rocker cover gaskets and the dipstick seal correctly. Also check the PCV valve - the 4.0 V8 has some weird valve thing that ****s itself.

You can try something flammable for vacuum leak searches... but it can be dangerous! Yuo can try water in a mister - you can sometimes see or hear it get sucked in, but it doesn't change the idle much. If you use brake cleaner or ether etc, you run the risk of igniting it, but you can usually hear the idle change.

If you can, try to get a really still, clear audio, video clip up - I'd be keen to hear it running with the wobble, to decide it's its just rough, or if it's a single misfire. You can often hear a constant misfire near the tail pipe too.

We basically need to rule some things out, or find some data to give clues on where to go. But basics should never be overlooked as a start. They're the foundation of a good running engine.

AllanWorms said:

You can try something flammable for vacuum leak searches... but it can be dangerous! Yuo can try water in a mister - you can sometimes see or hear it get sucked in, but it doesn't change the idle much. If you use brake cleaner or ether etc, you run the risk of igniting it, but you can usually hear the idle change.

Click to expand...

In the old days we used to spray WD40 or summat around suspected leaks. As you say the revs would pick up, sometimes more than somewhat!

Super interesting, thanks so much for your help!

Here's a video of the wobble at idle from a semi-warm start:



Not the worst it's been but gives an idea. Normally wobbles a lot more at traffic lights that the whole body moves.

Here's a video from the exhaust:


In progress:
- I bought a Penrite P26 carby cleaner and a white label fuel injector cleaner.
- I cleaned the idle position sensor with carby cleaner and no change. It had some black gunk in it that came out and it freed up a bit (wobbles better when shook).
- Trying to build another smoke machine in the meantime using an old soldering iron.

I got a new error code yesterday:
- P0102 — MAF low input (it's a brand new MAF though)

I cleared it and it didn't come back today. I think it's because I was doing aggressive acceleration yesterday.

In my Car Scanner app, I found some logs of misfires. I will check back in a few days because 256 sounds like a default value:

AllanWorms said:

Those Long term trims suggest it's running similarly on both banks, when driving at least.

Click to expand...

Long term trims after a few days of city driving:
- Bank 1: 20.31%
- Bank 2: 18.75%

Short term, I often notice:
- Bank 1: 25% trim
- Bank 2: open loop, 0% trim

AllanWorms said:

*IF* that wobble is from an ignition or compression misfire at idle, it can cause the out of range O2 reading - if one cylinder isn't firing, it doesn't burn it's share of the oxygen, so there is too much in the exhaust on the bank, and the ECU *thinks* it needs to add more fuel... but it doesn't. And it doesn't matter how much fuel is in there, it can *only* read the O2 level. You can try running it with one spark plug lead unplugged at a time, and see which cylinder doesn't make it run worse - that's probably the misfiring one. It does depend if the misfire causes the wobble, or if the fuelling being altered causes the misfire - it can go either way.

Click to expand...

Interesting!

When I run without a spark plug at a time: can I disconnect/reconnect safely while it's in idle, or should I shut engine off? Don't want to cause any damage to the plugs or ECU.

If it's a vacuum leak, how come:

1) The first error codes that appear are circuit malfunctions for the O2 sensors? First the bank 2 one, and then later the bank 1. And then only all the other O2 codes appear after some driving.

2) The right hand side O2 sensor is affected more than the left? It also often reads *slightly* lower voltages when driving. So when left is 0.03, right will be 0.02, left 0.06, right 0.02 etc.

AllanWorms said:

A compression test is always a good place to start though. Always do the basics, before getting too complicated... But I'd have thought your mechanic would have done that as a start, if he's doing all the work? Maybe he's not the best option???

Click to expand...

Don't think he did that. I'll try all the other stuff here before that, I reckon?

AllanWorms said:

Vacuum leaks are still an option too - anything that gets in without the MAF seeing it, is a leak. Make sure the rocker cover gaskets and the dipstick seal correctly. Also check the PCV valve - the 4.0 V8 has some weird valve thing that ****s itself.

Click to expand...

Good shout on the dipstick. I also heard the oil o-ring can be an issue, and the PCV valve yep, it's on the list to fix/clean! Maybe it's that because that's on the bank 2 side of the engine.

I tried opening the oil input while the engine was running and it was tightly in place that I couldn't and didn't want to put more force. So the vacuum there seems to be good?

AllanWorms said:

If you can, try to get a really still, clear audio, video clip up - I'd be keen to hear it running with the wobble, to decide it's its just rough, or if it's a single misfire. You can often hear a constant misfire near the tail pipe too.

Click to expand...

Thanks very much for your help! Videos above, I'm not sure what I should be listening for.

sierrafery said:

https://workshop-manuals.com/landro...omponent_information/specifications/page_587/

Click to expand...

This has been super helpful to read (modified link after browsing around)!

Hmmm... those videos sound more like it's rapidly pulsing up and down? rather than misfiring??? Does that sound about right?

It's a bit hard to hear over the belt/fan noise etc.

Those long term trims are pretty up there - Pretty sure a Disco 2 only goes to +/- 25% for long and short.

Are yours definitely positive? Do they show + or - before them at all?

If they aren't negative, the long is approaching max additive on both banks, and the Bank one of 25 short term, means it might even climb more. They does point to them reading lean and the system adding more fuel.

Where did this MAF come from? Cheapo (or even expensive fakes!) aren't generally the best option. Some of them are no better than the "default" setting when the MAF is unplugged... You should be able to log it's reading on a fullthrottle run (in 2nd gear say) and compare it's max reading with the max expected HP at the peak power RPM. There are calculators online to see the max reading

The circuit malfunction for th O2's generally means either open circuit, or possibly reading out of range. More specific reader sometimes give much better codes (like VCDS for VW's gives useful codes, where generic scanners give generic codes). I'm also not experienced specifically with the D2.

There could be an intermittant wiring fault, but it could also be the reading going out of range (off the scale).

I don't suppose you know anybody with an Oscilloscope to read the output voltage direct? Or an exhaust sniffer to confirm if the Lambdas read about what it really coming out the tailpipe?

Can you give more info about *when* the problems started? There might be something silly affecting things: Just yesterday, on my D1, I hade arough idle, and and HT lead had dropped onto an injector plug - the EMF from the lead was causing enough voltage in the injector for it to open and dump an extra injection of fuel. Moved the lead awy, and it improved immediatly :-D

AllanWorms said:

Hmmm... those videos sound more like it's rapidly pulsing up and down? rather than misfiring??? Does that sound about right?

It's a bit hard to hear over the belt/fan noise etc.

Those long term trims are pretty up there - Pretty sure a Disco 2 only goes to +/- 25% for long and short.

Are yours definitely positive? Do they show + or - before them at all?

If they aren't negative, the long is approaching max additive on both banks, and the Bank one of 25 short term, means it might even climb more. They does point to them reading lean and the system adding more fuel.

Where did this MAF come from? Cheapo (or even expensive fakes!) aren't generally the best option. Some of them are no better than the "default" setting when the MAF is unplugged... You should be able to log it's reading on a fullthrottle run (in 2nd gear say) and compare it's max reading with the max expected HP at the peak power RPM. There are calculators online to see the max reading

The circuit malfunction for th O2's generally means either open circuit, or possibly reading out of range. More specific reader sometimes give much better codes (like VCDS for VW's gives useful codes, where generic scanners give generic codes). I'm also not experienced specifically with the D2.

There could be an intermittant wiring fault, but it could also be the reading going out of range (off the scale).

I don't suppose you know anybody with an Oscilloscope to read the output voltage direct? Or an exhaust sniffer to confirm if the Lambdas read about what it really coming out the tailpipe?

Can you give more info about *when* the problems started? There might be something silly affecting things: Just yesterday, on my D1, I hade arough idle, and and HT lead had dropped onto an injector plug - the EMF from the lead was causing enough voltage in the injector for it to open and dump an extra injection of fuel. Moved the lead awy, and it improved immediatly :-D

Click to expand...

I do seem to remember even ages ago that the positioning of the spark plug leads on Rover V8s was critical, so sounds like a good call.
Is there any mileage in disconnecting the MAF to see how it runs on default? It works like a dream on a TD5 10p engine! Totally aware there is no other thing in common, or not much, between these two engines!

Thanks for the reply!

AllanWorms said:

Hmmm... those videos sound more like it's rapidly pulsing up and down? rather than misfiring??? Does that sound about right?

Click to expand...

Yeah, I wouldn't know the difference. Driver turning into a mechanic due to D2 purchase here

AllanWorms said:

It's a bit hard to hear over the belt/fan noise etc.

Click to expand...

I took another video from the back, maybe helps?

AllanWorms said:

Those long term trims are pretty up there - Pretty sure a Disco 2 only goes to +/- 25% for long and short.

Are yours definitely positive? Do they show + or - before them at all?

Click to expand...

Short answer: yes, probably positive.

Long answer: no negative symbol before it, but when the short term trims go negative, they'll get a negative sign! I've never seen the long term dip negative though.

AllanWorms said:

If they aren't negative, the long is approaching max additive on both banks, and the Bank one of 25 short term, means it might even climb more. They does point to them reading lean and the system adding more fuel.

Click to expand...

Cool, sounds like a consistent narrative, right?

AllanWorms said:

Where did this MAF come from? Cheapo (or even expensive fakes!) aren't generally the best option. Some of them are no better than the "default" setting when the MAF is unplugged... You should be able to log it's reading on a fullthrottle run (in 2nd gear say) and compare it's max reading with the max expected HP at the peak power RPM. There are calculators online to see the max reading

Click to expand...

I tried 2x brand new Bosch ones from reputable stores!

OK, cool, that's a good shout. I'll try that tomorrow. To confirm: put in second, full throttle on highway, record MAF kg/h throughout?

How many kg/h does yours read at idle?

AllanWorms said:

There could be an intermittant wiring fault, but it could also be the reading going out of range (off the scale).

Click to expand...

Yes, you're right! I looked this up in the manual better and it's exactly that, plus more:

P0130 - O2 sensor circuit malfunction (bank 1, sensor 1) - Front sensor LH bank stoichiometric ratio outside operating band
P0150 - O2 sensor circuit malfunction (bank 2, sensor 1) - Front sensor RH bank stoichiometric ratio outside operating band

P0171 - System too lean (bank 1) - Multiplication injector adaptive fuelling - lean limit exceeded LH bank
P0174 - System too lean (bank 2) - Multiplication injector adaptive fuelling - lean limit exceeded RH bank

P1171 - System too lean (bank 1) - Additive injector adaptive fuelling - lean limit exceeded LH bank
P1174 - System too lean (bank 2) - Additive injector adaptive fuelling - lean limit exceeded RH bank

P0102 - Mass or volume air flow low input - MAF signal < minimum threshold, which is speed dependent

AllanWorms said:

I don't suppose you know anybody with an Oscilloscope to read the output voltage direct? Or an exhaust sniffer to confirm if the Lambdas read about what it really coming out the tailpipe?

Click to expand...

Not that I know of. What kind of people would have an exhaust sniffer and who'd give me readings for free?

Environmental groups?

AllanWorms said:

Can you give more info about *when* the problems started? There might be something silly affecting things: Just yesterday, on my D1, I hade arough idle, and and HT lead had dropped onto an injector plug - the EMF from the lead was causing enough voltage in the injector for it to open and dump an extra injection of fuel. Moved the lead awy, and it improved immediatly :-D

Click to expand...

I bought it like this, sadly!

I put new leads and plus, no change, for example.

Stanleysteamer said:

I do seem to remember even ages ago that the positioning of the spark plug leads on Rover V8s was critical, so sounds like a good call.

Click to expand...

Crazy!

Stanleysteamer said:

Is there any mileage in disconnecting the MAF to see how it runs on default?

Click to expand...

Interesting idea. Will throw up errors, no? But I've tried like 4x different MAFs (2 used, 2 new), it can't be that?

I've never touched a D2, just D1s, and stuff that works similar to a D2 (BMW and VW stuff etc).

It does sound more like pulsing - a rapid hunt, rather than a misfire. Usually misfire make a slight pop or a slight pause in the exhaust. It could be related to where the richness is coming from.

So, based on those code descriptions you included, it thinks the system is too lean and is adding fuel, but can't add enough without hitting the +25% trim limit. Bank 2 hits the fault limit first. The fact it's doing on both banks fairly evenly suggests a fairly systemic problem.

This means either:

the fuel system can't deliver the required amount, and it IS still running lean.
OR
It's increasing the fuelling, but the O2 readings are still reading leaner than whats actually supplied.

You can unplug each O2, one at a time and see which code comes up - just to verify they are plugged in correctly. If they are swapped, it gets confused really fast. While I see there is a code for trnsposed sensors, other brands don't always bring up thise codes when they probably should.

At this point, I'd check the fuel pressure and regulator are working. The high vacuum at idle shoud reduce the fuel pressure, compared to throttle open. (the Vacuum at idle sucks on the injectors, so the pressure is reduced to compensate - referenced to the manifold vacuum, rather than atmospheric).

I'd also check for exhaust leaks affecting the O2 sensors. (Which I mentioned already, and you may have already checked).One way to do this, is to get the engine up to temp, then clear the ECU fuel trims (Don't know how on the d2 - some cars do it when you clear codes, the D1, you take power off the ECU for a few seconds). Then start it up, with a rag etc restricting the tail pipe and watch the trims. If there is a leak prior to the O2, the exhaust pulses can suck air in. The restriction means that air is less liekly to get sucked it, but the trims can take a while to adjust. If the shorts sit a bit more normal for a while, perhaps there is a leak letting in air.

Holding the rag on the exhaust can get very tiresome, but it can also get hot and smelly. Sit upwind and find something like vicegrips etc to hold the rag in a way you can change hands. Or see it you can get enough layers over the tailpipe, with a hose clamp over the outside. You need enough pressure that the engine can idle fine, but the exahust stays under pressure - I usually hold it.

You can also check for leaks with a stethoscope, with the head taken off and a thin, long metal pipe attached - as you move it along exhaust flanges etc, you hear the gasses blow over the pipe end.

You can aslo have an injector faulty/blocked, which makes the whole bank read lean. and the fuelling gets ramped up on that bank... meaning 3 cylinders on the bank run richer, one is still lean and so it can't reach the limit - that's less liekly to be both banks though.

Thanks! This is all really great advice.

I ran a DIY smoke machine through it and found some vacuum leaks coming deep from somewhere. I made a new thread for that with videos: https://www.landyzone.co.uk/land-ro...oke-machine-d2a-4l-v8-with-rough-idle.393398/

Do you think I should focus on only that? Or some of the other procedures you mention below?

AllanWorms said:

It does sound more like pulsing - a rapid hunt, rather than a misfire. Usually misfire make a slight pop or a slight pause in the exhaust. It could be related to where the richness is coming from.

Click to expand...

Would a vacuum leak like this cause the pulsating?

AllanWorms said:

So, based on those code descriptions you included, it thinks the system is too lean and is adding fuel, but can't add enough without hitting the +25% trim limit. Bank 2 hits the fault limit first. The fact it's doing on both banks fairly evenly suggests a fairly systemic problem.

Click to expand...

The smoke coming out more on the bank 2 side a hint of anything here?

AllanWorms said:

You can unplug each O2, one at a time and see which code comes up - just to verify they are plugged in correctly. If they are swapped, it gets confused really fast. While I see there is a code for trnsposed sensors, other brands don't always bring up thise codes when they probably should.

Click to expand...

Good shout this! I didn't think to check if they've been swapped at ECU.

AllanWorms said:

At this point, I'd check the fuel pressure and regulator are working. The high vacuum at idle shoud reduce the fuel pressure, compared to throttle open. (the Vacuum at idle sucks on the injectors, so the pressure is reduced to compensate - referenced to the manifold vacuum, rather than atmospheric).

Click to expand...

How would I check those and assess if working?

AllanWorms said:

I'd also check for exhaust leaks affecting the O2 sensors. ... One way to do this, is to get the engine up to temp, then clear the ECU fuel trims Then start it up, with a rag etc restricting the tail pipe and watch the trims. ...

Click to expand...

This is a great idea.

AllanWorms said:

You can also check for leaks with a stethoscope, with the head taken off and a thin, long metal pipe attached - as you move it along exhaust flanges etc, you hear the gasses blow over the pipe end.

Click to expand...

Taking off head seems like a huge job, eh?

AllanWorms said:

You can aslo have an injector faulty/blocked ...

Click to expand...

Can only check/clean injector spray with those $300 machines right?

If nothing made it out the oil filler... it suggests a fairly massive intake leak - maybe the gasket was screwed up last time it was off?
Perhaps an injector O ring?
I would 100% start with lifting the plenum, while carefully watching for anything moving as I did it. I'd expect some pretty obvious though, based on the amount of smoke making it out, and there no massive pressure its under.


I had wondered about rocker cover gaskets and valley pan gasket, but if the oil filler isn't getting enough flow to smoke, then it seems it's not getting that far anyway.

AllanWorms said:

If nothing made it out the oil filler... it suggests a fairly massive intake leak - maybe the gasket was screwed up last time it was off?

Click to expand...

Thanks! Oh wow, that's interesting to know. Yeah good point that it didn't make it to the oil filler.

Rocker cover gaskets and the valley gasket replaced by the mechanic, but I had the rough idle before that work.

AllanWorms said:

Perhaps an injector O ring?

Click to expand...

I hope it’s this and not deeper (like a gasket).

AllanWorms said:

I would 100% start with lifting the plenum, while carefully watching for anything moving as I did it. I'd expect some pretty obvious though, based on the amount of smoke making it out, and there no massive pressure its under.

Click to expand...

I will try! Some plastic in my smoke machine melted last time so I have to build a new one

AllanWorms said:

I had wondered about rocker cover gaskets and valley pan gasket, but if the oil filler isn't getting enough flow to smoke, then it seems it's not getting that far anyway.

Click to expand...

Oh that's interesting! Thanks for this. I hope you're right. So then can't be the inlet manifold gasket either?

Could be the intake gasket. It seems to be pretty early in the intake path... and it seems to be quite a leak!

Thanks mate! How could I verify without taking the whole engine apart?

And is there a risk that if I replace the intake gasket, then run smoke again, something later in the intake part crops up? Would be especially punishing if that something is something deeper and I have to re-do it all.

Already am regretting not changing the deeper inlet manifold gasket (valley) when the rocker gaskets and upper intake manifold gasket was done :|

Again, I'm not a Discovery 2 nutter... although I DO want a V8 manual one! "sensile family adventure wagon".

The plenum should be easy enough to remove? The valley cover is a bit more major, ut should still be doable in a weekend (unless you have a Mrs, and kids - it could be a 6 week job then!).

May as well take some leave from work and send your stress levels thourhg the roof, working on it!

Solved! Thanks for all your help, I have some updates and replies.

First off, some replies:

marjon said:

Idle control valve sticky?
As well as a +1 for checking for air leaks in the system before the O2 sensor and after the MAF

Click to expand...

I cleaned the idle air controller and it didn't help. Great call on air leaks, turns out there were many.

biketeacherdave said:

Smoke test the engine you have a massive vacum leak

Click to expand...

Yep, you were on point. Smoke test showed it.

AllanWorms said:

If nothing made it out the oil filler... it suggests a fairly massive intake leak - maybe the gasket was screwed up last time it was off?
Perhaps an injector O ring?

Click to expand...

You were also right! I replaced the injector orings. The old ones were super brittle.

AllanWorms said:

Could be the intake gasket. It seems to be pretty early in the intake path... and it seems to be quite a leak!

Click to expand...

And right again here probably. I also replaced the lower intake manifold gasket and the two silicone end caps.

Solution:

It was hard to see once I opened it up, but replaced 16x injector o-rings and also did the valley gasket for good measure.

That sorted most of the vacuum leak and the rough idle! Revs went down a bit too, and O2 codes went away, apart from these two:

• P1171 - System too lean (bank 1) - Additive injector adaptive fuelling - lean limit exceeded LH bank
• P1174 - System too lean (bank 2) - Additive injector adaptive fuelling - lean limit exceeded RH bank

My long term O2 trim readings are now:

• Bank 1: 5.47%
• Bank 2: 5.47%

MAF air flow rate is still low at idle: ~20kg/h

I've still got some excess air coming in:

1. from the throttle body spring (videos)
2. from brake booster to intake manifold connector (you can see it in this threads videos)
3. throttle body to plastic airbox connector (a new one... gonna use some plumber's tape to seal that off)
4. also the breather hose for the charcoal canister (but that's normal apparently)

I'm hoping that fixing #2 and #3 will remove the codes and that I won't have to replace my throttle body (#1).

What would you suggest?

Glad your getting somewhere with this that's a good effort. Those p codes are identical for each bank so start looking at what they share ie maf sensor and or air leak at the airbox/intake hoses etc.

MAF Diver said:

• P1171 - System too lean (bank 1) - Additive injector adaptive fuelling - lean limit exceeded LH bank
• P1174 - System too lean (bank 2) - Additive injector adaptive fuelling - lean limit exceeded RH bank

Click to expand...

As it’s an adaptive fueling code, does it need a drive cycle, or just driving a bit so it can learn that maybe it’s all ok now? Thought.
But lean would say it’s a leak after all sensors.
Continue fixing the leaks then see what you are left with code wise.
But good progress

J