BMW M5 Forum and M6 Forums banner

1 - 20 of 224 Posts

·
Registered
Joined
·
174 Posts
Discussion Starter #1
I start a new thread because one member suggested I do.
My first post for the issue is #50 of this thread.
Before going on please read the above.

Here is the answer from my indy for my inquiry. They use BMW tester(ISIS) and obtained for the SES light;
000090 DME lambda control function bank1 (These are translated from Japanese)
000091 DME lambda control function bank2

The emissions are measured at the tail pipe at idle with noload and CO and HC vary rather sporadically.
When OK, CO is less than 0.1% and HC is around 10ppm.
When dense , CO reaches 1-5% and HC 100-500ppm. These values changed every time they made measurement. Sometimes CO and HC are both high and in another time CO is low and only HC is high. Quite puzzling.

The replaced parts are;
secondary air shut-off valve
intake hoses and gaskets
brake booster hose
spark plugs
two ignition coils
CPS(bank2 exhaust)
crank shaft sensor
DME relay
water temp sensor (at radiator exit)
all four lambda sensors

Before the issue happened (scheduled maintenance);
T-stat (worn)

After clearing the codes the SES light is off, but comes back at the next engine start.

Any comments and advises are welcome.
 

·
Registered
Joined
·
9,927 Posts
Those are pretty simple codes to deal with. The diagnostic description is Pre Cat Oxygen sensor aging for both banks. You have replaced them and the code is still there so that tells me something else is preventing the O2 sensor from switching fast enough. The volts should go high to low and to their max very quickly. Yous are delayed and likely not going to the Max volts the DME expects.

Ask your shop to do a back pressure test on the exhaust. I am almost positive from your slow speeds you have significant carbon buildup in your cats. Don't let them just replace them, they can likely be cleaned from some driving. You will need some drives on the highway at higher RPM. A dozen one hour drives at 4000 rpm should get the cats hot enough to clean them out. So it will take a few months to do this. It will take 6 times before you notice much, and that might even get rid of the code but you have to keep doing it to get them clean.
Even after you get them clean you will need to get the cats good and hot to burn off what builds up once a month. These cars were not designed to trip around at 10 MPH.
 

·
Registered
Joined
·
4,488 Posts
sailor, the 90/91 codes are for fuel trims reaching the allowable limits, either negative or positive. The consensus on here seems to be it's specifically the high-load multiplicative adaption trim. Which is a bit surprising as the OP is obviously also running too rich at idle, which would indicate the additive adaptions should be maxed out too and no longer controlling, which should have brought in the AE/AF codes as well.

And wouldn't carboned-up cats have brought in the B2/B3 codes? Although I agree, under the circumstances, sooted-up cats seem likely.

Since there is no mention of the MAF's being replaced, it may be under-reading MAF's causing the 90/91 codes. Hyper, how old are the MAF's? Do you know how to do the MAF 140 litre/hr test using secret menu#4?

But bad MAF's wouldn't explain the rich idle. And the intermittent nature of the high HC and CO is strange. That seems to rule out the usual suspects like poor injector spray, contaminated cats, open t/s.

Maybe a saturated charcoal filter giving high readings when EVAP valve opens? I believe the valve only opens at idle. The DME probably ignores O2 voltages when EVAP is working as it expects rich readings.

Hyper, you could ask the shop to disable the EVAP system to see what the effect on HC and CO is.
 

·
Registered
Joined
·
9,927 Posts
I am assuming those are decimal codes considering the description he gave as lambda control. Suppose we should confirm how he got the definition and if they are hex codes or decimal.
You can't have a B2 or 3 with with a 5A or B (which is the 90/91 dec) because the 5A/B turns the pre cat O2 not ready, all four O2s must be ready to throw B2/3.
Besides the only time a car goes rich as he is describing at idle is when the vacuum control fails at the pressure regulator, I have to assume they checked that while dealing with the fuel reg. Maybe not though.

The odds on the evap being open every time the car is at idle is low.
 

·
Registered
Joined
·
4,488 Posts
The description "lambda control" makes me think those are hex codes.

And I figure an FPR with no vacuum giving 72 psi instead of 62 would only increase fuel flow by 8% (square root effect) which the adaptions should be able to handle easily. If the diaphram has failed that would certainly cause problems, but he's not getting constant high HC/CO every test. It passes one time and fails badly the next. That seems unusual.
 

·
Registered
Joined
·
9,927 Posts
Reality is the vacuum often compensates for the higher fuel pressure anyway. That said theory aside it has been reported many times that this failure does cause a really rich idle.
I see your point and now that I have reviewed the 00000 in front lead me to believe they are hex minus the 0x. I am reading occasionally it does not fail on HC and CO at idle but if these are hex codes then that is just a red herring anyway.
Back to the definition it is a fuel control fault in comparison to AF which is an air-fuel adaption fault. That in itself is funny because because a fuel fault is not lamda control, I am sure they just mean lamda error.

Anyway codes are sort of secondary I am assuming the techs have been complete because of the fees and have done all of the normal things. So I thought out of the box and other rarities that might not come to the techs mind and waited to see if the reported code would support something. I had two ideas and one is ruled out whichever those codes are, but the second idea is still plausible no matter if those are hex or dec. High exhaust back pressure and poor flow can cause a lot of strange symptoms.

Since the techs have done a lot it would be nice to know if they saw the proper fuel pressure readings ( I assume yes) and if they ever did a back pressure test. I am sure if they do a back pressure check they can figure it out from there.
 

·
Registered
Joined
·
174 Posts
Discussion Starter #7
Thank you very much, Sailor24 and 68FB. You are giving me the strength to cope with the issue.

It is a bit hard to catch you up because of the technical depth and language. Quite naturally English is not my native.

After I replaced the lambda sensors I took my car to high-speed run for about one hour at midnight. I primarily used 3rd and 4th gears to maintain high rpm, hoping the run would turn the SES off but in vain. It seems only an hour is not enough as you say.
Inside of my tail pipes have been clean with almost no soot. I often wondered why because I know my fuel consumption is poor. Maybe was the soot caught in the Cats?

As for the fuel pressure, my shop said it is 5 bar without vacuum and changes in accordance with negative pressure from the engine. Also the return line is OK and they are sure on the fuel supply system.

I know how to test MAFs because I did a lot of research on the forum (Remember I did it for idle?). But I did not do it for two reasons.
Firstly the performance of my car seems to be OK and the test requires tremendous speed. Secondly, as you say, it seems degraded MAFs will not make fuel rich.

My MAFs are more than 7 years old but my car’s odometer shows 43,000 km or 26,900 mile, if it is correct.

Regarding the back pressure test and the EVAP test, they said both were done without any anomalies.
 

·
Registered
Joined
·
9,927 Posts
Try this clear the codes, electrically unplug the MAFs drive the car for a while, the light will come back because of the MAFs being unplugged. After 300 - 500 miles have the codes read. Note how the car changes through the miles.
 

·
Registered
Joined
·
4,488 Posts
Inside of my tail pipes have been clean with almost no soot. I often wondered why because I know my fuel consumption is poor. Maybe was the soot caught in the Cats?

I know how to test MAFs because I did a lot of research on the forum (Remember I did it for idle?). But I did not do it for two reasons.
Firstly the performance of my car seems to be OK and the test requires tremendous speed. Secondly, as you say, it seems degraded MAFs will not make fuel rich.

My MAFs are more than 7 years old but my car’s odometer shows 43,000 km or 26,900 mile, if it is correct.

Regarding the back pressure test and the EVAP test, they said both were done without any anomalies.
No soot in the tailpipes! From my experience and from I've seen on here, that's rare for an M5. I wonder if the Japanese DME programming is even more emissions-strict than ours. Or maybe it's just that you never get aggressive enough with the gas pedal to get into open-loop operation of the engine when the air-fuel ratio gets very rich. And it indicates you don't burn any oil.

There's some good testing your shop could be doing to see what is happening at idle. (Maybe they are?) For example, they could be reading the fuel trims (lambda integrators, additive adaptations, and multiplicative adaptations) while the HC and CO is good and comparing it to when they are bad. And also reading the air flow measured by the MAF's, the coolant temp, and the intake air temp to see if any of these sensors are failing intermittently.

Has the shop given up? Are they waiting for you to tell them what to do? It can take a lot of testing and diagnosis to get to the bottom of this type of problem.

They could also check if the high HC/NO is related to EVAP system ON/OFF. I still think it could be a saturated charcoal canister. I suspect their EVAP test may have just confirmed the purge valve is working.

And you can do the MAF test in second gear. That gets you up to about 110 kph.

It's hard to diagnose these problems from long distance without more information.
 

·
Registered
Joined
·
174 Posts
Discussion Starter #10
Try this clear the codes, electrically unplug the MAFs drive the car for a while, the light will come back because of the MAFs being unplugged. After 300 - 500 miles have the codes read. Note how the car changes through the miles.
What do you have in your mind? Would you give me some hints? I'm a bit scared driving with MAFs unplugged.
It gets the car into the secondary mode, doesn't it?
 

·
Registered
Joined
·
174 Posts
Discussion Starter #11
No soot in the tailpipes! From my experience and from I've seen on here, that's rare for an M5. I wonder if the Japanese DME programming is even more emissions-strict than ours. Or maybe it's just that you never get aggressive enough with the gas pedal to get into open-loop operation of the engine when the air-fuel ratio gets very rich. And it indicates you don't burn any oil.

There's some good testing your shop could be doing to see what is happening at idle. (Maybe they are?) For example, they could be reading the fuel trims (lambda integrators, additive adaptations, and multiplicative adaptations) while the HC and CO is good and comparing it to when they are bad. And also reading the air flow measured by the MAF's, the coolant temp, and the intake air temp to see if any of these sensors are failing intermittently.

Has the shop given up? Are they waiting for you to tell them what to do? It can take a lot of testing and diagnosis to get to the bottom of this type of problem.

They could also check if the high HC/NO is related to EVAP system ON/OFF. I still think it could be a saturated charcoal canister. I suspect their EVAP test may have just confirmed the purge valve is working.

And you can do the MAF test in second gear. That gets you up to about 110 kph.

It's hard to diagnose these problems from long distance without more information.

No soot in the tailpipe is a good sign, isn't it? It show the combustion is perfect:). Any way I remember last December, when I accelerated in the 3rd gear,a shock like a fuel-cut happened at 5500rpm. It was on a slight uphill on a highway. The engine did not stalled but would not go beyond 3000rpm after that. The engine was OK at the next day. Similar phenomena happened twice in last June in the 4th and the 2nd gear at lower rpms respectively while my shop killed the secondary air system. Do these mean something? What do you think?
My car is not an oil drinker.
BTW I use BARDAHL XTR 10w-60. How do you rate this oil?

I have no idea how to measure the fuel trims and a bit dubious they have ever measured STFT and LTFT. The shop does not specialized in BMW. My feeling is that they are very serious but more experienced in Mercedes, VW and Porsche. I once asked about the trim but they simply said the injector open times were around 2ms.

I talked with my dealer about the issue today. The dealer seems to have given up too. Since they have exactly the same codes as the shop, they thought it is associated with air leaks somewhere and end up in a wild-goose chase.

As for EVAP test I will ask the shop together with the measurement of STFT and LTFT.
Also I will try a MAFs test ASAP.

For high-speed runs I bought two cans of BMW petrol additive.

Regards.
 

·
Registered
Joined
·
9,927 Posts
If the car runs fine and the codes go it points to MAFs or air leaks. We then know that the car is getting the air the DME expects. If the car runs poor then you can rule out MAFs and air leaks above the throttle plates and look for something that would prevent the car from having the correct mixture., ie valves in the wrong place (vanos) poor fuel delivery, possibly an air leak below the throttle plates or ICV system although you don' have the classic stall when you take the gas off. Exhaust leak or pressure, it just narrows down the window to look in.
It is safe the car goes into factory Alpha-N. Where do you think the tuners got the idea? At this point I am guessing you have an air leak below the throttles or in the ICV system. The lack of soot tells me your car is running too lean which is what the codes indicate. You will likely only have to run unplugged for a bit because the car may not run well.
 

·
Registered
Joined
·
174 Posts
Discussion Starter #13
At Sunday midnight I did a MAF test.
Here is the result.
Measured Corrected(by the Excel chart given by DavidS: Temp;28C, pressure;1013mbar,alt;guesstimated to 50m)
2nd gear till fuel cut(7000 rpm) 138.5 L/H 143.1 L/H

I also measured with the 3rd gear and obtained higher value but I am less confident. I wish I had some one in my back seat to read those figures.
Now for the three consecutive nights I drove my car at high speed using lower gears possible and the tailpipes are accumulating soot rapidly.
 

·
Registered
Joined
·
174 Posts
Discussion Starter #14
If the car runs fine and the codes go it points to MAFs or air leaks. We then know that the car is getting the air the DME expects. If the car runs poor then you can rule out MAFs and air leaks above the throttle plates and look for something that would prevent the car from having the correct mixture., ie valves in the wrong place (vanos) poor fuel delivery, possibly an air leak below the throttle plates or ICV system although you don' have the classic stall when you take the gas off. Exhaust leak or pressure, it just narrows down the window to look in.
It is safe the car goes into factory Alpha-N. Where do you think the tuners got the idea? At this point I am guessing you have an air leak below the throttles or in the ICV system. The lack of soot tells me your car is running too lean which is what the codes indicate. You will likely only have to run unplugged for a bit because the car may not run well.
Sorry the quality of the picture. I only have a hard copy. My shop disassembled the engine up to this point and also checked the ICV. They replaced O rings and connecting rubber pipes.
 

Attachments

·
Registered
Joined
·
4,488 Posts
At Sunday midnight I did a MAF test.
Here is the result.
Measured Corrected(by the Excel chart given by DavidS: Temp;28C, pressure;1013mbar,alt;guesstimated to 50m)
2nd gear till fuel cut(7000 rpm) 138.5 L/H 143.1 L/H

I also measured with the 3rd gear and obtained higher value but I am less confident. I wish I had some one in my back seat to read those figures.
Now for the three consecutive nights I drove my car at high speed using lower gears possible and the tailpipes are accumulating soot rapidly.
Those numbers look good, indicating the MAF's and the fuel delivery system are working well (assuming you had the MAF's connected).
I'd still like to see fuel trims to confirm it.
 

·
Registered
Joined
·
4,488 Posts
Sorry the quality of the picture. I only have a hard copy. My shop disassembled the engine up to this point and also checked the ICV. They replaced O rings and connecting rubber pipes.
Taking all that stuff apart makes me think the shop was looking for a below-throttle body vacuum leak. So maybe they read the fuel trims and they figured the 90/91 codes were caused by severe lean running? I say "severe" because the fuel trims will correct for moderate vacuum leaks.

I'm just guessing though.
 

·
Registered
Joined
·
4,488 Posts
I did some digging into what causes the 90/91 codes and found that they trigger when the O2 sensors see a large, rapid change in the O2 content of the exhaust - a change so quick the relatively fast-acting (minutes) additive adaptations can't change quickly enough to correct for it. This drives the lambda integrators to their limits of 1.25 or 0.75.
So I am now convinced the 90/91 codes and the random passing and failing of the tailpipe testing are caused by the same problem, something changing randomly and quickly.

So what can cause a sudden change in AFR? The only things I can think of are the EVAP system switching ON and OFF during idle, a bad fuel pump changing speed erratically, or the MAF signal jumping around at idle.

Hyper, once again it comes down to your shop doing some monitoring during idling, namely fuel trims, fuel pressure, and air flow.

Regarding what brings in the 90/91 code, I found some posts on the board where people had provided the code freeze frames from INPA and also shown the analogue 2 data. Every time the codes were caused by the lambda integrators getting to 1.25 or 0.75.

I duplicated this by pulling both MAF's about 1/2" out of the tubes to force them to read low, about 15 kg/h. This forced the lambda integrators to quickly drive to 1.25 to correct for the sudden low reading and when they did the 90/91 codes came in. In the INPA freeze frame below, the lambda integrators are called "lambda regulation factor".

I also noticed that when air flow dropped to about 14 kg/h, the DME finally said "that can't be right" and started substituting values of around 20 kg/h instead. But it didn't set a MAF code. Probably needed more convincing the MAF's had gone bad.

EDIT: I forgot to mention, it has always bugged me that the 90/91 code doesn't tell you if the mixture offset is rich or lean. Turns out it does if you look at the INPA freeze frame below, specifically the line "upper regulator limit".

INPA2.jpg
 

·
Registered
Joined
·
174 Posts
Discussion Starter #18
I did some digging into what causes the 90/91 codes and found that they trigger when the O2 sensors see a large, rapid change in the O2 content of the exhaust - a change so quick the relatively fast-acting (minutes) additive adaptations can't change quickly enough to correct for it. This drives the lambda integrators to their limits of 1.25 or 0.75.
So I am now convinced the 90/91 codes and the random passing and failing of the tailpipe testing are caused by the same problem, something changing randomly and quickly.

So what can cause a sudden change in AFR? The only things I can think of are the EVAP system switching ON and OFF during idle, a bad fuel pump changing speed erratically, or the MAF signal jumping around at idle.

Hyper, once again it comes down to your shop doing some monitoring during idling, namely fuel trims, fuel pressure, and air flow.

Regarding what brings in the 90/91 code, I found some posts on the board where people had provided the code freeze frames from INPA and also shown the analogue 2 data. Every time the codes were caused by the lambda integrators getting to 1.25 or 0.75.

I duplicated this by pulling both MAF's about 1/2" out of the tubes to force them to read low, about 15 kg/h. This forced the lambda integrators to quickly drive to 1.25 to correct for the sudden low reading and when they did the 90/91 codes came in. In the INPA freeze frame below, the lambda integrators are called "lambda regulation factor".

I also noticed that when air flow dropped to about 14 kg/h, the DME finally said "that can't be right" and started substituting values of around 20 kg/h instead. But it didn't set a MAF code. Probably needed more convincing the MAF's had gone bad.

EDIT: I forgot to mention, it has always bugged me that the 90/91 code doesn't tell you if the mixture offset is rich or lean. Turns out it does if you look at the INPA freeze frame below, specifically the line "upper regulator limit".

View attachment 463874
What a great insight! Appreciated, 68FB.
However I still am having hard time catching you up. Let me ask some questions.

'Additive adaptation' means LTFT? Does the lambda integrator mean the maximum correction quantity for LTFT?
Also could the shop measure the fuel trims with the software which spat the codes 90/91?
Or do they need INPA or Analogue2, or whatever software?

When it comes to how to look at the INPA result you have shown, it is a mystery to me.

I have been puzzled recently by fuel consumption issues.

First, after I had the lambda sensors replaced I went home from the shop with the great expectation for fuel consumption.
Actually the value, 6.1km/L, was worse than it used to be(6.9km/L). The distance between my house and the shop is 20 miles, half highway and half city. Why do the new sensors give worse number?

After that the fuel consumption on the on-board computer sticks around 4.6km/L plus/minus 0.1. For the high speed MAF test the average speed jumps from 18km/h to 22km/h(I did not reset it), but the fuel consumption value has kept the value. That is quite puzzling because my understanding is the higher the average speed the better the fuel consumption (to a certain degree).
 

·
Registered
Joined
·
4,488 Posts
'Additive adaptation' means LTFT? Does the lambda integrator mean the maximum correction quantity for LTFT?
Also could the shop measure the fuel trims with the software which spat the codes 90/91?
Or do they need INPA or Analogue2, or whatever software?

When it comes to how to look at the INPA result you have shown, it is a mystery to me.

I have been puzzled recently by fuel consumption issues.

First, after I had the lambda sensors replaced I went home from the shop with the great expectation for fuel consumption.
Actually the value, 6.1km/L, was worse than it used to be(6.9km/L). The distance between my house and the shop is 20 miles, half highway and half city. Why do the new sensors give worse number?

After that the fuel consumption on the on-board computer sticks around 4.6km/L plus/minus 0.1. For the high speed MAF test the average speed jumps from 18km/h to 22km/h(I did not reset it), but the fuel consumption value has kept the value. That is quite puzzling because my understanding is the higher the average speed the better the fuel consumption (to a certain degree).
Additive adaptions are neither STFT or LTFT. They are a BMW (or maybe German?) trim that does fine corrections for fuelling errors at idle and low loads. They may make a small contribution to the LTFT.

The closest BMW thing to LTFT is the multiplicative adaption, which corrects errors in fuelling at higher loads. The multiplicative adaptions are close to but usually slightly less than the OBDII LTFT's.

Lambda integrators are exactly like STFT's, they change rapidly in response to the O2 sensors indicating the AFR has gone rich or lean.

Say an air leak develops downstream of the MAF's. Now not enough fuel is being injected to give stoichiometric combustion. The O2 sensors see excess O2 in the exhaust and their output immediately drops below 0.45V and stays there when it should be switching rapidly from 0.1 to 0.9V. After a few seconds, the lambda integrators react by increasing from 1.00 to say 1.10 to increase the injection time. The DME monitors the lambda integrators and if they stay >1.00 for a few minutes, it starts increasing the additive adaptions to permanently increase the injection time. Eventually the increasing additive adaptions will add enough injection time (fractions of a millisecond) to bring the air-fuel mixture back to stoichiometric. At this point, the O2 sensors start switching up and down between 0.1 and 0.9V, the lambda integrators return to 1.00 and a new equilibrium has been established.

If the air leak is small, it will have little to no effect on the multiplicative adaptions, since a small leak is insignificant at the higher airflows this adaption deals with.

If the shop can read the BMW hex codes, they will be able to read these fuel trims.

Doing the MAF test gives high speeds, but running at WOT gives terrible fuel consumption as you can imagine! I wouldn't worry too much about your mileage. It's just a symptom of the problem that needs fixed.
 

·
Registered
Joined
·
174 Posts
Discussion Starter #20
Additive adaptions are neither STFT or LTFT. They are a BMW (or maybe German?) trim that does fine corrections for fuelling errors at idle and low loads. They may make a small contribution to the LTFT.

The closest BMW thing to LTFT is the multiplicative adaption, which corrects errors in fuelling at higher loads. The multiplicative adaptions are close to but usually slightly less than the OBDII LTFT's.

Lambda integrators are exactly like STFT's, they change rapidly in response to the O2 sensors indicating the AFR has gone rich or lean.

Say an air leak develops downstream of the MAF's. Now not enough fuel is being injected to give stoichiometric combustion. The O2 sensors see excess O2 in the exhaust and their output immediately drops below 0.45V and stays there when it should be switching rapidly from 0.1 to 0.9V. After a few seconds, the lambda integrators react by increasing from 1.00 to say 1.10 to increase the injection time. The DME monitors the lambda integrators and if they stay >1.00 for a few minutes, it starts increasing the additive adaptions to permanently increase the injection time. Eventually the increasing additive adaptions will add enough injection time (fractions of a millisecond) to bring the air-fuel mixture back to stoichiometric. At this point, the O2 sensors start switching up and down between 0.1 and 0.9V, the lambda integrators return to 1.00 and a new equilibrium has been established.

If the air leak is small, it will have little to no effect on the multiplicative adaptions, since a small leak is insignificant at the higher airflows this adaption deals with.

If the shop can read the BMW hex codes, they will be able to read these fuel trims.

Doing the MAF test gives high speeds, but running at WOT gives terrible fuel consumption as you can imagine! I wouldn't worry too much about your mileage. It's just a symptom of the problem that needs fixed.
Thank you for the explanation.

The additive adaptation, the multiplicative adaptions and the lambda integrators are BMW's (or German) terms and the STFT and the LTFT are the OBDII terms? Are you saying they are similar but different in their definitions?
The additive adaptations seem to correspond (in a sense) average values and the the lambda integrators are deviations around them. Also the 90/91 codes are telling the DME could no longer adjust fuel trims correctly. That is my understanding.
If fuel trim measurements are done what comes out from instruments depends on which(BMW-based or OBDII-based)software you use?

The photo below is the diagnosis I got from my dealer. Perhaps the shop's looks the same. Sorry for the flash masking some part. The three characters right after 'DME' mean 'lambda', the following six 'control', followed by two symbolic kanjis which mean 'function'. The three right in front of '1' are 'bank'.
I do not know whether the codes are hex or decimal this is how the diagnosis looks. Very simple.

Since it took whole month for the shop to disassemble/assemble the engine, I would like to exclude air leaks somewhere. Does that point the EVAP or MAF? What about the engine temp sensor, the EGR valve or the carbon on the intake valves?
Although the MAF seems OK at WOT, there could be a nonlinearity near the idle? Does the unplugging clarify this?

I sent a mail to the shop asking for the test, waiting for their reply. I really wish to solve the issue soon.
 

Attachments

1 - 20 of 224 Posts
Top