BMW M5 Forum and M6 Forums banner

How to test the MAFs yourself and How MAFs work

300K views 500 replies 158 participants last post by  68FB 
#1 · (Edited)
The MAFs are vital to the proper function of the S62 engine. The engine can never work better than its MAFs.

The MAFs (Mass AirFlow meters) are located in the M5 and Z8 air intake tubes. The MAFs report to the Engine Electronics (DME aka ECU) the amount of air that is drawn into the engine, and what temperature this air is. The DME uses this to calculate how much fuel to inject.

Vital as the MAFs are, those have to be _really_ bad before you actually get a stored fault-code.

The functional test by the BMW tester just tests the MAFs at idle with the car standing still. IMH experience that says absolutely nothing about how they perform at WOT (Wide Open Throttle).

I think there is an economical (warranty) reason why BMW does not want the MAF fault code to trigger as soon as the MAFs start to be less-than-perfect, rather than only when there is extremely poor MAF function.

MAFs typically last 50000 miles before getting so contaminated it affects S62 engine performance.

There are two methods to test the MAFs yourself:

Method A - using the instrument cluster of the car *********

Read post #11 in:

http://www.m5board.com/vbulletin/showthread.php?t=1898

Follow those instructions, and select the mode for fuel consumption per hour.
With the engine fully warmed up, drive several WOT (wide open throttle) accelerations. Notice the fuel flow per hour numbers. If everything is ok the numbers should increase continously to about 140 litres per hour at WOT at 7000 RPM. (don´t worry about bouncing into the rev-limiter)

The fuel-flow is directly proportional to the MAF reading reported to the Engine Electronics, so if the Fuel-flow is correct, the MAFs are most likely correct too.

To be able to compare your fuel-flow readings with other people´s readings,
you need to correct the data for temperature, barometric pressure and altitude. This is easy: look at the table attached at the bottom of this sheet.

Look up the two factors matching your temperature, barometric pressure and altitude. Then multiply the reading from the instrument cluster with both those factors. The result will match what the fuel flow number would have been at sea-level, +20 deg C and normal barometric pressure.

Method B - using an OBD-II scan-tool *******************

1. Get a graphing and data-logging OBD-II scantool. If you already have a laptop PC, it will cost you about US$ 100 to buy one. Worth every cent. (I use the one called "ISO" from www.obd-2.com which I am very pleased with)

2. Plug it into the car and go for a drive

3. Look at the graph for CLV (calculated load value). This is basically the MAF signal expressed in percent of the expected maximum at each RPM.
When you accelerate WOT this number should immediately go to 100% and stay there until you release the throttle or hit the rev-limiter .

4. If the CLV reading it is low at all RPMs, try new air filters.

5. If it is ok at low RPMs, but lower at higher RPMs, try the MAF cleaning procedure and the retest immediately. If this improves the numbers you know 100% there is (or if the cleaning was very successful used to be) a MAF problem. This does not say it is the only problem, but the car can never run better than its MAFs.

Symptoms of bad MAFs ************************

In degree of malfunction of the MAFs possible symptoms are:

High Lambda values at idle (lean air-fuel-mixture)
Engine hesitation at high RPM
Engine misfire at high RPM
Engine going into limp-home mode at high RPM

MAF cleaning procedure ************************

http://www.louv.tv/cars/m5/MAF/
(by m5board member Louv)

How the MAFs work ***************************

The MAFs have a non-heated temperature sensor and heated temperature sensor. The latter is made of a very thin wire of platinum metal. The wire is heated to about 200 deg C, and its temperature is continously monitored and the heating current is controlled so as to keep the wire at this temperature. The more air that passes the wire on its way into the engine, the more the wire will be cooled. The electrical current needed to maintain the sensor wire at 200 deg C, is directly proportional to the amount of air that passes it, and also depends on the temperature of the air (hence the non-heated temperature sensor).

The electrical current and the air temperature are measured, and the corresponding air-flow is calculated and reported to the engine electronics.

How MAFs clean themseves ***********

Once in a while the MAFs will run a very high electrical current for a few seconds through the platinum sensor wire, so the wire gets red-hot (1000 deg C). This burns off much of the contaminants, and probably help keep the MAFs working much longer between manual cleaning or replacment intervals.

How MAFs fail *********************

With time the MAF platinum-wire gets fouled with contaminants settling from the rapidly passing intake-air. These contaminants acts as thermal insulation on the wire. The effect is that the wire will not be cooled as much as it should by the air-flow, and thus reports too low air-flow to the engine electronics. The engine electronics in turn uses this too low air-flow number when calculating how much fuel to inject. The fuel injection will be to low, and the air-fuel mixture in the cylinders of the engine will be lean.

Many M5 owners have reported K&N oiled intake air filters to cause MAF contamination.

What to do if the L/H fuel flow numbers are low?
************************************

1. Have you corrected for altitude/barometric pressure/temperature and humidity?

2. Numbers still low? Most of the time, with a L/H reading very low, either the fuel pump/fuel pressure regulator and/or the MAFs are shot.

Most likely the MAFs.

To check the MAFs for severe contamination:

CHECKING MAFS
************

Either run the BMW DIS / GT-1 tester procedure for MAF testing at idle and inreceased idle RPM, or use this (IMHO far more accurate and reliable) procedure to measure MAF output voltages with engine off:

(Better still use both methods: road test L/H, and direct voltage. Then compare the results.)

MAF TESTING - DIRECT VOLTAGE METHOD
********************************

0. Remove tie-wraps holding wire protection tubing in place at the MAF connectors. Pull tube back and find the applicable wires. Connect to these using insulation piercing needle-and-hook type test leads(or just plain needle type test leads if you don´t mind risking to prick yourself).

Make ceratin to have the mulitmeter set to DC volts and that you have not connected to the AMPs connectors on it. (The signals you are connecting to go straight to the DME so be careful not to short-circit anything.)

1. Ignition on
2. Engine not running
3. On each MAF measure DC voltage between
pin 3 (ground, brown wire) and pin 5 (signal, yellow wire)
4. Reading should be 1.000 +- 0.040 V. The closer to 1.000 V the better. the dirtier the MAF the higher the voltage will be.
5. If reading in step 4 is off by more than 0.020 V then measure voltage between pin 3 and pin 4 (+5V reference from DME, yellow/green wire). This reading should be 5.000 +- 0.200 V.
6. If not check the battery voltage, and connect a charger if needed and redo the tests.

7. Put the tubing back in place and secure it with new black 2.5 mm wide tie-wraps. (Make sure you have narrow enough tie-wraps to fit on hand before cutting the old ones.) To avoid sharp edges, after the tie-wraps are in place cut the excess with a flush (one side only beveled) pair of cutting pliers.

If you don´t have that kind of cutters, just cut the tie-wraps with any sharp cutter and the file the sharp corners. Only takes a minute, and might save some of your blood next time you do something in that area on the engine compartment.

The reading in step 4 should be exactly 1/5 of the reading in step 5.

If this procedure gives a reading in step 4 that is outside tolerance the MAF is definitely contaminated or otherwise damaged.

FUEL PRESSURE
************

The other plausible possibility is the fuel pressure.

To measure the fuel pressure is very easy: he will need a BMW adapter for fuel pressure sensing (BMW special tool 13 5 220. This is a kit containing 13 5 221 and 13 5 222. This is common for all E39 5-series, so the dealer probably has it already) that connects to th efuel pressure regulator, and to the pressure test sensor on the DIS.

The procedure is very easy and is described in detail in TIS 13 31 029.

The thing to be aware of is that BMW states the fuel pressure as the difference between the absolute pressures in the fuel rail and in the intake tract (downstream of the throttle butterflies).

At idle the intake ports will be at -0.5 BAR. Just measure the atmosphere-relative pressure of the fuel-rail, and then add 0.5 BAR to get the reading relative to the intake ports.

In other words: hook the fuel test port up to the DIS, read the (athmosphere relative) pressure the normal way, and if it is 4.50 +- 0.05 BAR everything is fine so far.

Then disconnect the vacuum hose from the fuel pressure regulator, plug the hose, leave the fuel-regulator vacuum port open, start the engine again, and measure the fuel-rail pressure once more. It should now be 5.00 +- 0.05 BAR.

(The difference between these two readings is due to the vacuum pressure in the intake tract.)

BTW, this would be a good time to measure the vacuum too while you are at it as you have access to the hose anyway.

Warranty *************************

As the MAFs influence the air-fuel mixture and cause high Lambda values (lean mixture), it is possible to fail emissions testing due to non-perfect MAF sensors. When the vehicle fails emissions testing the Emissions Components Warranty may be applicable.

In the us see:

http://www.epa.gov/obd/warranties.htm

The mandatory warranty covers everthing relating to emissions for the first two years, and the OBD, ECU (DME in BMW language) and the Catalytic converters for eight years.

So in the US the MAFs are covered for at least 2 years by law, but not for eight years.

David

UPDATE NOTE 2006-Oct-24:
I have updated the Excel-file, so calculations are now automatic. Just enter your numbers. Much more conveinient to use.

UPDATE NOTE 2006-OCT-25:
I have fixed a bug affecting calculations in metric units in the Excel file. Current version is 3. Please, do not use ver 2.
 

Attachments

See less See more
#3 ·
Great post! Good job. Thanks! :cheers:
 
#5 · (Edited)
That was really great! And I appreciate the effort. However, if one does suspect the Mafs going bad wouldn't it be easy to just replace them instead of cleaning them?

I know they are expensive like $380 ea. but why wouldn't BMW just clean them if they are still under warranty. It would be cheaper for BMW. Can the Mafs or any part of them wear out and need to be replaced?

I heard of a few people trying to clean them and their M5 still didn't run correctly until they were replaced. Then Waa-Laa runs fine...

I am asking as a layman.
 
#6 ·
Thanks, everybody!

The BMW TIS (workshop manual) very specifically forbids any kind of disassembly or cleaning of the MAFs. The reason for this, I assume, is that Bosch cannot guarantee the function of the MAFs if handled. Since the MAFs are critical to good emissions levels, there might be regulatory matters involved too.

On the other hand the BMW test procedure for MAFs is a joke. You can have pretty much MAF-contamination induced high-RPM power loss, and still not get a MAF fault code, and pass the BMW idle-speed MAF test.

David
 
#7 ·
I think there might be financial considerations involved. The MAFs surely don´t cost BMW hundereds of dollars each.

Over the lifetime of the vehicle, if BMW can sell a pair of MAFs every 50000 miles it adds up nicely.

Similar to why you can´t buy a set of syncronization rings (syncros) for the gear-box. You have to buy a new gear-box instead...

David
 
#8 ·
DavidS said:
1. Get a graphing and data-logging OBD-II scantool. If you already have a laptop PC, it will cost you about US$ 100 to buy one. Worth every cent. (I use the one called "ISO" from www.odb-2.com which I am very pleased with)

David
I think the link is http://www.obd-2.com/ is anyone is wondering...

- Kin
 
#9 ·
Kin Mak said:
I think the link is http://www.obd-2.com/ is anyone is wondering...

- Kin
Oops.. You are absolutely right! Big Thanks for pointing it out.
I have corrected the link in the first post.

David
 
#10 ·
Good post, thx DavidS :)
 
#12 ·
DavidS said:
The BMW TIS (workshop manual) very specifically forbids any kind of disassembly or cleaning of the MAFs. The reason for this, I assume, is that Bosch cannot guarantee the function of the MAFs if handled.
David
Great Info David
Reference your cmt above about BMW TIS not being specific enough to disassemble / clean MAFs.....was wondering if you knew if Bosch had those functions documented for use at their service facilities. :confused2

I remember a past problem I had with a Volvo Turbo diesel that Volvo could not fix. They sent me to VW (OEM for engine) for tests.

VW then sent me to local Bosch dealer who eventually fixed the injector system. Turned out to be a faulty gasket assembly that Bosch used in the assembly. :crying:

I can't remember if Bosch has repair facilities in the US but they have many such facilities here in Germany.

Ken
 
#13 ·
David,

Excellent post. Because I am at 48k miles I might start looking at the cleaning issue in the very near future.

Mark
 
#14 ·
You guys are not gonna beleive this! The local delaer tried to clean my MAF's and didnt work, so they are replacing both as we speak!!!!! Sonce i had them replaced 8 months ago... they are still under warranty!
 
#15 ·
excellent write up DavidS, all the concise info in one post! :M5thumbs:
 
#17 ·
Great write up!

I went out to try this because I've been skeptical of my engine's performance lately. What I notice most often is a "fallout", where I loose power under heavy acceleration often right at 4000 RPM. I also notice stumbling under low RPM (2500-3000) heavy acceleration.

I followed this recommendation. " With the engine fully warmed up, drive several WOT (wide open throttle) accelerations. Notice the fuel flow per hour numbers. If everything is ok the numbers should increase continously to about 140 litres per hour at WOT at 7000 RPM."

Conditions: wet roads, 62 degrees F, DSC off, Sport on.

I performed several WOT accelerations in 2nd gear from a rolling "start", but only saw the l/h values push into the range of 90 l/h once at about 7000 RPM. Most often the values that were present with acceleration I have grown to accept as "strong" were in the range of 80-88 l/h. These observations were in the 6000-6500 RPM range. That's waaaaaaay off from David's observations.

I have never driven another M5 so I have no comparison to make in that regard. I'll try again with dry roads just to verify there isn't something going on with the DSC, but it sure doesn't feel like it. I have recently cleaned the MAFS, but observed no change after doing so.

Has anyone else had similar observations of 140 l/h? :confused:

:cheers:
 
#19 ·
I have also never driven another M5 back to back for comparison and I tried the tests also (fantastic tips by the way :) ).

A few WOT's in 2nd got up to around 100, one of them went up to 113; however, I have not kept the throttle open once up in the high revs (perhaps 140 is registered when after several seconds of being near 7000 rpm?) ... I also don't have so much road space to do this on the way home from work (not to mention people looking out of their cars at me as if I'm a complete idiot!).

I'm really looking forward to a few people trying this to get a good benchmark and to enable fine tuning of the testing :)

Regards,

Gary.

OldRanger said:
Great write up!

I performed several WOT accelerations in 2nd gear from a rolling "start", but only saw the l/h values push into the range of 90 l/h once at about 7000 RPM. Most often the values that were present with acceleration I have grown to accept as "strong" were in the range of 80-88 l/h. These observations were in the 6000-6500 RPM range. That's waaaaaaay off from David's observations.

I have never driven another M5 so I have no comparison to make in that regard. I'll try again with dry roads just to verify there isn't something going on with the DSC, but it sure doesn't feel like it. I have recently cleaned the MAFS, but observed no change after doing so.

Has anyone else had similar observations of 140 l/h? :confused:

:cheers:
 
#20 · (Edited)
Going from 5k to 7 k RPM is enough

Gentlemen,

There is no need to do the test accelerations from low RPM. Going several times from 5k to 7K RPM and back down gives good data. You just need to be accelerating long enough for the Vanos units to adjust the camshafts to favourable WOT positions. I think that takes half a second or so.

The figure i posted, 140 l/h, is in 2nd gear, about 132 l/h in 3rd gears, about +10 deg C, dry, normal atmospheric pressure. Oil at +90 deg C. I think the reason for the higher number in 2nd gear is that the DME enriches the air-fuel mixture during transient accelerations, and that the accelerations performed are shorter in 2nd than in 3rd, and therefor get more enriched. Anyway the numbers I have seen correlate with outside temperature and atmospheric pressure. On top of that there seems to bee perhaps +- 3-4 l/h noise.

There are very many factors that influence the fule flow number, all that factors that would influence engine performance on a dyno test plus the uncertainity of the car´s fuel flow meter.

Perhaps different fuels with different compositions have different energy densities, such that different volumes of fuel match different masses of air?

I posted both proceures to allow a ballpark estimate, which i believe is far more accurate and relevant than the BMW Tester checking the MAF output signal at idle.

If you look at the fuel flow figure at idle, you will se a figure of about 4-5 l/h. This is just 3% of the peak flow, and hence judging the MAFs performance at WOT at such low airflows that correspond to that fuel-flow is surely prone to error.

I think the fuel-flow method is very useful if you do a few accelerations, then clean the MAFs and then immediately repeat the accelerations. If you get a relevant improvement you can safely assume the MAFs are directly involved.

David
 
#21 ·
Re: Going from 5k to 7 k RPM is enough

Thanks for the follow-up everyone!

DavidS said:
Perhaps different fuels with different compositions have different energy densities, such that different volumes of fuel match different masses of air?
Hmmm. I didn't think much about the different fuel compositions. There are obviously many additional factors affecting the readings I haven't considered. The tank of fuel I'm using right now was filled up from nearly empty with a Sunoco 93 octane fuel. To the point of other variables affecting the other side of the equation (air density), it is quite humid here (>90%) and raining, temperature yesterday and today is about 15C. I'm operating the vehicle at an elevation of 250m according to the onboard GPS. I suspect this is accurate within 10m.

DavidS said:
If you look at the fuel flow figure at idle, you will se a figure of about 4-5 l/h. This is just 3% of the peak flow, and hence judging the MAFs performance at WOT at such low airflows that correspond to that fuel-flow is surely prone to error.
Interesting observation to note here that at idle with a warmed up engine (end of commute to work), the fuel consumption of my engine is indicated at 1.7 - 1.9 l/h. This is very close to yesterday's 1.8-2.0 l/h at idle. In order to get readings of 4-5 l/h under no-load conditions, I had to raise RPM to 1500-1750.

Gary...I'm in the same boat as you, and I did not keep WOT once I got up to 7000RPM. I'm chicken :D
 
#22 · (Edited)
Another datapoint

I agree about the contributing factors. If it is of interest, here is another datapoint:

M5: March 2000
79000 km

Intake air filters: about 2000 km

Fuel: Statoil 98 octane (European octane number)

Oil: Castrol TWS about 12000 km

Air:
+10.3 deg C
+12.0 deg C indicated by car OAT
95% RH
1008 mbar absolute

Coolant 81 deg C
oil 91 deg C after runs

2nd gear

1st run 135 l/h at revlimiter
2nd run 132 l/h at revlimiter

Will check warmed-up idle fuel flow on next drive...

David
 
#23 ·
Yet another datapoint. This is on a 2000 M5 with 49.6K miles. Stock air filters(changed 7K miles ago). Oil temp 97C, running Shell 91 at 70F, heading East:M5thumbs: .

The highest I saw was 120L/H in 3rd gear at 7K. It fluctuated very rapidly between 113 and 120, but I never saw anything above 120L/H. I will re-run in 2nd gear.
Initial no load fuel consumption is 1.2 to 1.7L/H after the car is warm. AC on causes a jump to the 1.8 to 2.7L/H range. Fully cold the consumption is more like 4.5L/H.
 
#24 ·
Great!

Did you run it WOT until the rev limiter kicked in?

David
 
#25 ·
kinghurl said:
Initial no load fuel consumption is 1.2 to 1.7L/H after the car is warm. AC on causes a jump to the 1.8 to 2.7L/H range. Fully cold the consumption is more like 4.5L/H.
I can confirm that I've seen L/H numbers nearly equivalent to kinghurl's under 'no-load' conditions like idle or coasting. Upon further observation, the range of numbers I've seen at idle is 1.4 to 1.9 L/H when the engine is fully warmed up. (2000 M5 with stock filters in good condition)

Unlike the other data points, I have not seen WOT consumption numbers exceed 90 L/H. I did run WOT right up to 7000, but backed off as soon as that RPM was attained.

As a side note, I've decided to have the fuel filter changed next week when I have it in for follow up service on a starter noise they couldn't diagnose last visit. I figure at almost 59,000 miles it's probably worthwhile.

:cheers:
 
#26 ·
DavidS said:
Great!

Did you run it WOT until the rev limiter kicked in?

David
Yes I did.
Just did another run this afternoon and it was a tad warmer- 75F. Oil temp at the start was 96C and 104C at the end. In 2nd gear at WOT, I saw 132 briefly and it appeared to fluctuate between 120 and 127 for most of the time between 7K and limiter. That's about a 6% off of your original post. Where did the 140L/H number originate?
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top