1996 328is Complete DIY MAF Test, easy and simple

[berkeleyEE]

The Mass Air Flow (MAF) sensor is an important and expensive sensor in the M52 engine.
It's often blamed for cars not running right or not running efficiently.

Since the MAF is a simple sensor, just a hot-wire anemometer ( ref: http://forums.roadfly.com/forums/bmw...5790040-1.html and maybe Wikipedia ) it would seem that testing a MAF for go/no-go operation should be simple.
Here's how I did just that on my 4-pin (13 62 1 703 275) MAF:

First, this is an electrical test performed off the vehicle so you need some basic equipment.
- a source of 12VDC power, anything in the range of 10V to 15V at 1/2 Amp should do,
- a voltmeter,
- a heat gun: you don't need the heat, just the blowing wind.

Hook-up will be easier with some grabbers instead of just wires, and some sockets that fit the pins in the MAF connector. I found some repackaged Molex crimpable sockets, size 0.093 inches, in a local electronics supply store (what Radio Shack used to be) that worked for me.
The MAF connector has 4 pins and they're clearly numbered on the interior of the connector.

The test setup is simple:
- set the power supply to 12V to 13V, then turn it off until you're ready to measure,
- connect supply GND to MAF pin 1
- connect supply POS to MAF pin 3
- connect the voltmeter NEG to MAF pin 4
- connect the voltmeter POS to MAF pin 2

MAF_test_lo.jpg . . . MAF_test_HI.jpg . . . MAF_test_closeup.jpg

Turn on the meter, then turn on the power supply.
With no moving air, I found the MAF to consume about 150mA of current yet the output was pretty much zero volts.
But blowing fast air with the heat gun caused the current consumption to increase to 250mA and the output was well over 1 Volt, sometimes 2 Volts Clearly this MAF is operable.

In my car, I measure the MAF output to be about 0.75V at idle and well over 1V with a light foot on the accelerator. Power is supplied to the MAF with the ignition in both ACCY and RUN.

Of course, this DIY is descriptive only, I don't recommend you do this procedure without professional assistance or training. If you mis-read pin numbers or short contacts together, it's your own fault.

[328 Power 04]

this is how you test your car from an engineer's (or engineer wannabe) point of view obviously.... most of us just disconnect the MAF and look for differences, which might not work as well.


VERY VERY good job with the above work!

[328istothebone]

hahaha thats my thought exactly!
That'st what i thought this post was going to consist of! haha

[LaloHao]

Sorry for res'sing thread but i had to say
Thanks!

You saved me having to pay a s#%t mechanic (i don't know any good one from where i live), i rather fix it myself

[TRF434]

First, this is an electrical test performed off the vehicle so you need some basic equipment.
- a source of 12VDC power, anything in the range of 10V to 15V at 1/2 Amp should do,
- a voltmeter,
- a heat gun: you don't need the heat, just the blowing wind.

Hi. Using the 12V output of an ATX power supply is safe? I do know some basics of electronics and electricity, but i don't know if the ATX is an overkill. Thank you.

[Eric93se]

An ATX power supply should be fine, check the voltage before you do the test because it depends on whether or not you have the power supply setup as a bench unit or not (sense wire is connected, etc).

There's nothing wrong with this test, but MAFS can work at low range and fail to measure high range and if you care about performance this is an issue. We would need to know the working range of the sensor, what is the top voltage output (12v? that's just a guess) and then we would have to force more air through the MAFS, using something like a leaf blower.

[ed323i]

Great! Thanks a lot for this info.

Another option would be to use INPA and look at the live values. It would be nice if we could collect a database of reference numbers. I gave it a first try in an earlier thread, but I didn't find enough values.
I did find some examples at idle 700 rpm (youtube INPA sources in this post):
M52B20: 10.5 - 10.75 kg/h
M52B25: 11.5 - 11.75 kg/h
M52B28: 10.5 kg/h (I suspect this one has a vacuum leak, as it seems too low)
M52B28: 13-14 kg/h (idle at 750 rpm though, probably because A/C on or alternator at high load or something, so at 700rpm would be a little lower)

Perhaps other people with INPA can share their idle 700 rpm values (engine warmed up).
You can find it in INPA, live values, analog values 1 under "Total air use of air mass flow meter [kg/h]"

What would be really cool if some people could upload their TestO (a free logging program) log files, one at idle 700 rpm, and one at WOT going from 700rpm to 6500 rpm in, say, 2nd gear. Then we can compare the complete MAF curve.. One at idle (which should be stable) and one at WOT going through the rev range (which should be increasing). That way a large part of the range of the MAF is tested. Low loads are very difficult to compare though, because you can't be sure how far the gas pedal is depressed.. Although, one could specify the TPS value and try and keep the TPS value equal during the acceleration.. That way you could also compare low-load and medium-load accelerations MAF values. If the curve of your MAF is virtually identical to the curve of known-good MAFs, then you can be 99% sure your MAF is in perfect shape.

[TRF434]

Yes. I'm building an ADS interface myself but due to time constraints I'm getting there slowly. In the meantime I'm diagnosing my car the old-fashioned way.

It'd be cool if the experts on this matter do a write-up on a "standardized" test procedure so everyone can collect data following the same criteria for better analysis and comparison.

[ed323i]

Yeah, that's a cool idea. For a lot of test procedures the Bentley manual is quite good, but the MAF test procedure in that manual isn't this extensive/complete.

Also, specifically for this topic: The 328i MAF is used in:
* M52B25 323i
* M52B28 328i
* S52B32 M3
* Z3's with these (single Vanos) engines

So, the test procedure also applies to those E36s (and also single Vanos E38 728i, E39 523i and 528i)

[TRF434]

Yeah, that's a cool idea. For a lot of test procedures the Bentley manual is quite good, but the MAF test procedure in that manual isn't this extensive/complete.

Also, specifically for this topic: The 328i MAF is used in:
* M52B25 323i
* M52B28 328i
* S52B32 M3
* Z3's with these (single Vanos) engines

So, the test procedure also applies to those E36s (and also single Vanos E38 728i, E39 523i and 528i)

Is the procedure/results different for the M50B25TU?

[ed323i]

Here's a list of cars in which this MAF was used:
http://www.realoem.com/bmw/enUS/part...&q=13621703275

The M50B25TU used a different part, but I think it also uses Siemens electronics (MS40 DME if I'm correct), and hence the MAF should also be Siemens/VDO.. So, probably would be comparable, the procedure, but no guarantees.
The M50 non-Vanos used a Bosch MAF and is of a different type, so for that model I'm certain you'll need another procedure.

[TRF434]

Here's a list of cars in which this MAF was used:
http://www.realoem.com/bmw/enUS/part...&q=13621703275

The M50B25TU used a different part, but I think it also uses Siemens electronics (MS40 DME if I'm correct), and hence the MAF should also be Siemens/VDO.. So, probably would be comparable, the procedure, but no guarantees.
The M50 non-Vanos used a Bosch MAF and is of a different type, so for that model I'm certain you'll need another procedure.

No. The M50B25TU engine uses the Bosch Motronic M3.3.1 and thus a different MAF. I thought the test was universal, sorry for the misunderstanding.

[ed323i]

I think you're right, for all M50B25TU's. At least the European M50B20TU did use Siemens MS40.1, but also the Motronics 3.3.1 (both as of 9/92 until 12/94, when the M52 replaced the M50TU in Europe) . So, it seems both were used. See http://www.realoem.com/bmw/enUS/show...86#12141703942 . Even discovered that some German cars had MS40.0 for non-Vanos M50. Probably all European and US M50B25TU's are Motronics 3.3.1 though.. Didn't manage to find examples where the MS40.1 was used.

[tjm3]

I agree that this described testing will tell you if there is a signal being output from the MAF sensor but a good, reliable signal depends on a couple of other parameters. The very first thing is the correlation between the AMOUNT of air passing through the MAF providing the correct voltages as output. For an outright test you must know the amount of air going through the unit along with a way to check the output signal using either factory data OR, if you can measure the airflow, multiple readings at various airflow amounts to create a valid table. Otherwise there is just an unverified number that doesn't really translate to anything. The other major issue untested in this procedure is the speed or reaction time of the output signal from the MAF in response to changing airflow. Most diagnostic procedures will advise to scope the output voltage vs. time and look for response time, smoothness and dropouts. Even these won't alway find a bad sensor. Known good readings need to be recorded because factory info is lacking.

But, if there is a signal that can be had through a scan tool in grams/sec or lbs/hour or whatever, then that number along with a couple of other engine parameters like RPM, displacement and air temperature can be plugged into a Volumetric Efficiency calculator, any number of which you can find in a google search. Using those numbers a good VE calculator will estimate the airflow that should be passing through the engine for that displacement and RPM and then try to match the amount of air going through the MAF and be able to tell you something isn't right. Good airflow through the MAF matching the airflow estimated through the engine = good or high VE most likely over 80% in most cases. MAF not working = VE dropping much lower. This works, and nothing needs to be dismantled for testing. And, using the same VE calculator any two or more engines can be compared for good or bad MAFs using the scan tool data from each.