K6JRF at the mic K6JRF's Page
formerly W6FZC

My Mercedes Benz
S500 Coupe

K6JRF's MB S500 Cpe
(Updated: Jul 23, 2012)

Analyze and Troubleshoot "Check Engine" MIL and Electronic Control Units (ECU)!

MAF Contamination & Lambda

A few days ago while using my tech's AST Retreiver scanner at his shop, I checked "live data" from the LH-SFI (N4/1) module and found some questionable readings. The ON-OFF ratio (Lambda) was reading from 65 - 75% with the center around 70%. The MAF (Mass Air Flow) air mass was measured low, about 18 Kgm/h.*  There were NO CODES in the Diag or LH-SFI Modules. This menu item discusses what steps I took to diagnose and resolve the problem.

What is the cause . . .
At first glance you might think that it's the O2 sensor since it's not at "50%" point, but you would be wrong in this instance. Many O2 sensors are replaced for $125 a pop but that does not correct the problem. Normally an O2 sweeps a voltage range from .1V to .8V, from lean to rich. If the ECU gets a lean signal from the O2 it will richen the fuel mixture to compensate. And if it gets a rich signal, it will lean out the mixture. The Fuel Computer (LH-SFI) is looking for the ideal voltage signal of .45V from the O2 sensor or at least the 'average' to be about 0.45V (Menu#4). So the O2 signals ranges up and down around the 'perfect' 0.45V signal.

The readout showed that indeed the O2 sensor was ranging from lean to rich, so that appeared normal. But the MAF reading was very low . . at least as I remembered the former readings so that prompted me to go on to do some experiments. Jumping ahead, this problem may have different causes, but in almost every Mercedes in the '90s, the most likely cause is the Mass Air Flow (MAF) sensor!

MAF Functions . . .
The MAF (B2/2) is mounted in the center of the engine under the air cleaner. My S500 Cpe employs the "hot wire" type sensor. A constant voltage is applied to the heated wire. This wire is positioned in the air stream and is heated by the electrical current that the voltage produces. As air flows across it, it cools down. The heated wire is a positive temperature coefficient (PTC) resistor, meaning that the resistance drops when the temperature drops. The drop in resistance allows more current to flow through it in order to maintain the programmed temperature. This changing voltage is sent to the fuel computer (LH-SFI). The voltage level is interpreted as airflow and the voltage changes are interpreted as change in the airflow. Straightforward but elegant stuff!

Mercedes has a cleaning cycle built into the MAF circuit. When the key is turned off a larger current is sent through this hot wire causing it to glow red hot, thus burning off most contaminants. It usually does a good job for most contaminants but if it's contaminated by "silicone" then it's resistant to this burn cycle. Or in other words, it's not usable until it's replaced. It can also become silicone contaminated via the silicone that's used on the circuitry in these sensors because it migrates and then sticks to wire surface.

It can also become contaminated by use of sprays that contain silicone! If you look at MENU#22, you'll see that the 'tire-shine' contains silicone and, most likely is the cause of my contaminated MAF. As you see from the pictures, the silicone was applied to most 'everything' including the air breather tubes and the air box itself. A very short path from there to the MAF!  The other 'co-contaminate is the filter oil used in the K&N filters. Read the second footnote at the end of this article.

What Happens. . .
The MAF sensor can also be contaminated from a variety of other sources such as: dirt, oil for vapor, spider webs, potting compound from the sensor itself and the filter oil used in reusable air filters.**

When a MAF sensor is contaminated, it skews the transfer function such that the sensor overestimates air flow at idle and that causes the fuel system to go rich and it underestimates air flow at high air flows and that causes fuel system to go lean. This means long-term fuel adjustments will learn lean corrections at idle and learn rich corrections at higher airflows. So my observations at idle showing 70% Lambda and 18 Kgm/h air mass now make sense.

Contaminated vs Good MAF

The figure shows a normal MAF compared to a silicone contaminated MAF. The differences are a low air flow (20-40 Kgm/h) where it gives higher readings and at high airflow (300 - 400 Kgm/h) where is gives lower readings.

To double-check the readings I got with the AST Retreiver, I hooked up my Lambda Tester (M0039) and verified the 70% Lambda value but couldn't measure the MAF volume since it doesn't have the capability. So now, it's time to see if I can fix it.

Fixing The Problem. . .
To confirm this disconnect the MAF sensor connector. It's located on the drivers side and takes a quarter turn to loosen it. With the car running, looking at the Lambda tester, M0039, this should force the car into the learned AZ/EL precoded fuel management mode and drop to 50% area. In this mode, airflow is generated by using rpm and throttle position instead of reading the MAF sensor. If the lean drivability symptoms go away, the MAF sensor is probably contaminated. If the lean drivability symptoms do not go away the MAF sensor is probably not contaminated and it might well be the O2 sensor.

Mercedes says that you should replace the MAF since it cannot be repaired. With the cost of about $250, it's a fairly large repair bill. This problem can be fixed and it can be done by just about everybody. Put it this way, what do you have to lose to try this method?? Not much, just a few dollars for the cleaner.

When you look at the MAF what you see with their naked eye is very little. It is nearly impossible to determine if a MAF sensor's thermistor is dirty or clean from looking at it without magnification and good lighting. When viewed under a microscope, this is what you might see.

MAF sensor closeup

What you need to do is go to your local car parts store and buy a can of CRC's "MAF Sensor Cleaner". Price is apx $5.99 for 11oz can.

CRC brand MAF sensor Cleaner An alternate cleaner is a non-chlorinated brake cleaner that leaves no residue. I found a cleaner that carries the Auto-Zone brand that is guaranteed to leave no residue. Don't use carburetor cleaner b/c it will leave a chemical residue.

Disconnect the NEGATIVE battery cable and remove the MAF from the car. Spray the inside of the MAF with the sensor cleaner thoroughly being careful to clean the sides as well. Be careful to spray the center wire since it's very thin and if it breaks, you will have to replace the MAF. Clean it well and let it air dry for 20 to 30 minutes. You can use compressed air to remove any remaining particles but make sure you bleed the pressure down to 70 psi so as to not break the wire.

When it's dry put it back into the vehicle and reconnect the NEGATIVE battery cable. You will need to drive the vehicle several miles to let the fuel computer "relearn" your specific driving condition but it will adjust quickly since the MAF is now properly sensing the air flow. The fuel computer will have to readjust over a longer term but that will happen;  it just takes some driving time. Of course you could reset the fuel computer to 'nominal' values using the scanner.

While putting the car back together, I noticed that the IAT sensor (B17/7) was also heavily contaminated. So it was cleaned and appeared perfect after the cleaning.

Results. . .
After cleaning the MAF and IAT, Lambda tests were back at "50%" point (as measured on the M0039) at both idle and 2500rpm; see MENU#4 for Lambda test methods. This morning (3/27), I measured the same parameters using the AST Retreiver. Air mass measured 23.2 Kgm/h, an increase of apx 25+% which shows that the MAF was contaminated! The scanner also confirmed the M0039 ON/OFF ratio 50% oscillating measurements as correct. So, the experiment was a success! To say I'm pleased is an understatement!

This a worthwhile procedure which may just save you the cost of a MAF or an IAT or both!

Comments. . .
This section was written after reading the number of threads concerning the effectiveness various filters and MAF contamination. The information provided puts a finishing touch on my observations.

The 'real-world-test' described in this link is the best that I've seen and shows test results of using paper, foam and cotton composition air filters in a real auto, '92 Miata, driven on real roads in the Pacific Northwest for 500 miles for each filter. Be sure to read the FLOW and FILTRATION tests. Their conclusions say it all; must reading!

The second is a combination link showing all the recent discussions on K&N and paper filters. Also good reading.
*  The AST Retreiver's standard measurement units for Air Mass is "kilo grams/hour", abbreviated as "Kgm/h".
**  I use K&N filters in my car and they claim that their filter oil does not 'fly-off' even in high air flows and it can NOT contaminate an MAF.   Referencing K&N's installation sheet under "Performance Hints", K&N says "Caution: Extreme fine dust, agriculture or off-road use, will pull the oil from the element. Frequent re-oiling of the element's clean side may be required". For more info, click here to see K&N Tests.

Send me Email Icon with your comments