Mass Airflow Sensor talk

[snow663]

Hey guys I haven't been here in a while but I am still having an issue with my e36 hesitating, bucking, surging. I'm gonna try and keep this short so bear with me here. Everything I've seen so far indicates my car is running lean. I've checked my coils with an ocilloscope and indeed all six coils were putting out 50,000 volts and I can also say my boots and plugs are good. But a high voltage spike after the burn line tells me it's not getting enough fuel as the gap resistance increases too soon. All fuel pressure tests came out perfectly so I started looking at sensors. The MAF particularly. I do not have any experience with mass airflow sensors so this is new territory for me. I was expecting to see a low voltage signal from the MAF but it seems to be in the right voltage range. However there seems to be a lot of noise and I feel that this noise is excessive however I am not qualified to say so. I will have attached my throttle snap data from the oscilloscope hooked up to the MAF. If anyone can shed some light on this for me I'd be forever grateful. I don't know if the photo will be clear enough i may have to attach another one

Sent from my Z799VL using Tapatalk
009.png006.png003.png010.png

[dworthy]

Have you cleaned the MAF with the proper cleaner? While there should be some small spikes, it seems that you have quite a few so this is throwing the DME off a bit.

Also have you had the intake smoke tested to ensure you don't have a vacuum leak as that can cause a lean condition as well.

Lastly knowing that this is OBD-I you have no codes, correct?

[snow663]

Have you cleaned the MAF with the proper cleaner? While there should be some small spikes, it seems that you have quite a few so this is throwing the DME off a bit.

Also have you had the intake smoke tested to ensure you don't have a vacuum leak as that can cause a lean condition as well.

Lastly knowing that this is OBD-I you have no codes, correct?

I have cleaned the MAF but it seems like the only improvement comes from resetting the DME after a couple pulls it goes back to running lean and rough. I have smoked the intake and found no leaks. Even the valve cover gaskets are excellent. I've put a lot of effort into trying to solve this problem. And no matter what I do it still lacks power, idles rough, and hesitates coming off a Dead stop at 1800 rpm, I've put all new sensors on it, knock, IAT, ECT, O2, Cam & Crank position, TPS IACV, coils, boots, plugs, and even a DME and I've also tested the vanos operation with the valve cover off using compressed air.

Sent from my Z799VL using Tapatalk

[dworthy]

Humm, would you have a buddy who can loan you an MAF to ensure that is what causing this?

The only other thing that can come to mind is your sensors. What brand did you purchase, for I have seen one with "Made in China" fail straight out the box.

[snow663]

Humm, would you have a buddy who can loan you an MAF to ensure that is what causing this?

The only other thing that can come to mind is your sensors. What brand did you purchase, for I have seen one with "Made in China" fail straight out the box.

I actually bought my parts from pelican or my local parts store, I can say they are OEM and I have another MAF that I got out of a junker. It runs the same with either MAF installed but runs pig rich with the MAF disconnected. Also I forgot to mention there are no codes.

Sent from my Z799VL using Tapatalk

[snow663]

Here's some interesting info. When the MAF is COLD as in vehicle has been off for 12 hours the MAF signal is all over the place. The waveform has twice the amplitude and is super fuzzy. I was unable to catch this in photo. It only lasted a few seconds on cold start before the waveform continues like the photo I posted originally.

Sent from my Z799VL using Tapatalk

[dworthy]

What is the IAT (Intake Air Temp) in the MAF reporting? I am wondering if that is playing a role in all of this as well.

The bottom line is while you should have small spikes, the large ones worry me as they look pretty constant up/down at a set time.

[MarcoZandrini]

Darin, I don’t think the MAF has the IAT. It’s probably located in the intake.

[dworthy]

Darin, I don’t think the MAF has the IAT. It’s probably located in the intake.

True, as I was thinking AFM at the time. The M-50 has it in the intake of course.

[snow663]

It's a hot film meter it has 4 wires power, ground, Signal +, and Signal -, I've checked with a multi meter and all the connections seem to be good. Could it be noise from something else interfering with it.

Sent from my Z799VL using Tapatalk

[dworthy]

If both MAF's are showing similar noise, then I would check the wiring. Make sure every connector/splice is clean and it wouldn't be a bad idea to do a voltage drop test.

[snow663]

If both MAF's are showing similar noise, then I would check the wiring. Make sure every connector/splice is clean and it wouldn't be a bad idea to do a voltage drop test.

That's kinda what I thought but I haven't scoped out my backup MAF yet. I'll have to swap MAFs and scope it again. Now if I zoom way in on the spikes they don't have a flat point. It's like an acute triangle and it's a perfectly straight line up to a fine point then straight back down. If it were a shoddy connection I'd expect the spikes to be rougher with a flat point am I right? The spike event takes up a whole 15 microseconds.

I have an idea tho I can scope out the power and ground connections indidually and look for and dropouts or spikes in the supply current.

Sent from my Z799VL using Tapatalk

[dworthy]

Okay, lets see what the back-up MAF brings, and if it is the same or similar then we may have to start looking elsewhere if the wiring is ruled out.

I am hitting the edge of my expertise here, so let's learn together on this.

[snow663]

Okay, lets see what the back-up MAF brings, and if it is the same or similar then we may have to start looking elsewhere if the wiring is ruled out.

I am hitting the edge of my expertise here, so let's learn together on this.

That's why I'm here man I want to learn. Things like this are still scarce on the web. And that means if we figure this out we can be the first ones with data to show for verifiable results. Try looking up a bad MAF waveform lol.

Sent from my Z799VL using Tapatalk

[snow663]

I was able to find the bosch datasheets, after looking at these i almost want to say this is an intermittent failure of the output stage transistor or the OP AMP.

Datasheet.PNG
Design and function The sensor element comprises a ceramic substrate containing the following thick-film resistors which have been applied using silk-screen printing techniques: Air-temperature-sensor resistor Rϑ, heater resistor RH, sensor resistor RS, and trimmer resistor R1. The heater resistor RH maintains the platinum metallic-film resistor RS at a constant temperature above that of the incoming air. The two resistors are in close thermal contact. The temperature of the incoming air influences the resistor Rϑ with which the trimmer resistor R1 is connected in series. Throughout the complete operating-temperature range it compensates for the bridge circuit’s temperature sensitivity. Together with R2 and Rϑ, R1 forms one arm of the bridge circuit, while the auxiliary resistor R3 and sensor resistor RS form the other arm. The difference in voltage between the two arms is tapped off at the bridge diagonal and used as the measurement signal. The evaluation circuit is contained on a second thick-film substrate. Both hybrids are integrated in the plastic housing of the plug-in sensor. The hot-film air-mass meter is a thermal flowmeter. The film resistors on the ceramic substrate are exposed to the air mass under measurement. For reasons associated with flow, this sensor is far less sensitive to contamination than, for example, a hot-wire air-mass meter, and there is no need for the ECU to incorporate a self-cleaning burn-off function.

[bmwdirtracer]

Hi guys.

If you'll humor me for just a moment, I'd like to ask that you consider a possible culprit besides the maf.
Certainly, I see, Snow, that you have been extremely thorough and pretty scientific in your hunt. And, in those explorations, you have replaced the MAF, and nothing changed


But let's consider the wiring harness .... and .... the injectors. Both are huge variables.


Given the history presented above, especially all the (well-considered) replacements and substitutions, the very first thing I'd do is make sure electrical connections are as perfect as I can make them.

Wurth Electrical Contact Cleaner Oil is actually and truly Magic Spray. I use it whenever I am questioning a poor connection between plug and plugee. (Do you like my new word?)

I really can't tell you how special this stuff is, but be assured it lives in a different dimension than CRC and others.

But of course, pins push backwards sometimes, in a plug. I suggest trying to pull backwards on suspect connector's wires, to see if the pin backs up. Or gently push each female pin with a screwdriver or pick, to see if it moves backwards out of the plug.

Make sure the DME is properly, externally grounded. This is absolutely crucial on earlier E36's. Other major grounds okay?

.......

And then, I'd send the injectors off to RC Engineering. $25 each, best bargain in the automotive world, and the only way to make sure that the other equations are even feasible. Twenty-something year old injectors just are not going to be accurate.


Wait......do you have a scan tool with which you're watching live data? I know that the data isn't easily accessible, and that it's limited -- but, still, there's a bunch of numbers you can watch on a good scanner with old software. (A SnapOn brick with BMW software is great, DIS does just fine, too, Autologic will do it, Launch even, with BMW software.

Compression test? Leak down? Just fishing, you know.....

Does the car have any symptoms which are specifically related to rpm alone? (Like, does it refuse to rev above ~4500, or does it quit bucking and spitting when it hits some rpm?)
The smoke test used a professional machine and ran ~7 minutes plus?

[snow663]

okay i will try to make this clean easy work for us so i will address your list in order.

1: The Harness, upon first inspection i was quite pleased to see that none of the wiring in the car shows any wear or aging, no cracks, hardened insulation or wear spots from physical contact, the connections are still glimering silver no corrosion at the ecu or any of the harness connectors for that matter, and as i went through and backprobed each sensor i checked for corrosion and or loose pins, even spade connectors that were bent or didnt grip the male pin tightly and found absolutely nothing wrong.
2: Plugs And Pugees, i do like that, haha all jokes aside like i said in the harness part i have thoroughly visually inspected all plugs and connectors even splices in the box above the engine and in the injector harness box. and gave each plug a wiggle with the boot pulled back and with the engine running, i could not see anything wrong nor did i cause the engine to stumble.
3: Grounding, i have pulled all the grounds in the engine room and wire brushed, assembled, tested resistance then smeared dielectric grease over the nut and connector. all grounds show 0.4 ohms resistance from chassis even at the sensors.
4: injectors, i have tried swapping in a set of 6 injectors from a company that ultrasonically cleans them and flow matched them to no avail, i ended up putting the original ones back and replacing just one that had a clogged pintle hole. all cylinders look the same on a primary ignition scope test so the problem is affecting the two banks equally. Fuel pressure is 43psi at idle 51 at wot or with vac hose disconnected from FPR also I should add it's got a new filter and FPR
5: Scan tool, no i do not have a scan tool, i have a Digital Storage Oscilloscope
6: Compression & Leakdown, compression is 175-180psi range across all 6 no major differences, leakdown showed that there was some slight leakage past the rings but it was so minimal i could barely tell, i had no air coming out the exhaust or intake manifold.
7: RPM Relation to power, the only thing that IS specific to RPM is that when the car is at a complete stop, in gear, let off the clutch and start rolling, clutch fully released, as i pass 1800 RPM it bucks then is smooth but weak to the top. this is not clutching technique, or my foot on the petal, or drivetrain, it IS the engine.
8: Smoke test was not professional, i blew smoke into the intake and crankcase then pressurized it with a regulated air compressor, at first there were some leaks but they are long gone now. it will hold air pressure almost indefinitely. the VCG was the hardest to get sealed, but i found that putting a thicker washer on the middle bolts by the plug wells to squish the gasket a little tighter has been the best fix for the m50 since well i dont know lol

[dworthy]

Let's see what the second MAF is doing.

It seems everything is starting to point to the DME, for it is the last unknown in the whole thing.
As side question, did you put a chip in the DME?

As for the timing, you would feel the lose in power once you started out, but it wouldn't hurt to double check to ensure it is on.

[snow663]

Let's see what the second MAF is doing.

It seems everything is starting to point to the DME, for it is the last unknown in the whole thing.
As side question, did you put a chip in the DME?

As for the timing, you would feel the lose in power once you started out, but it wouldn't hurt to double check to ensure it is on.

I have another DME. They both run exactly the same the other MAF I have is not known to be working with any car so it's kind of an invalid test it could be defective as well.

Sent from my Z799VL using Tapatalk

[bmwdirtracer]

I just reread the thread, and a couple of things stand out.

To me it seems the most crucial bit of info is the part about the symptoms disappearing for a short while, after you "Reset the DME". I'm going to assume this means you're doing a battery reset?

Hell, I don't even recall that early of a car having adaptations , limp mode, or anything beyond a code clearing at reset. Hopefully Abel will check in on this.

If you don't have a scan tool, how do you know there are no codes? (Don't trust "stomp test")

I'm assuming that you don't have ASC+T. (?)

The 1800 rpm is likely to be a significant clue, also. Vanos issue? Just because you can control it with air doesn't mean it's working correctly. Unfortunately, I do not recall being able to watch vanos angles on live data on this early a car.

You're using premium fuel? Is there any knocking from the engine which might cause the knock sensor to retard things? (Maybe unbolt knock sensor(s?), wrap them in foam for testing? Do the battery reset, with this, of course.

I'd personally want to see all the available live data. With your analytical acumen, it seems to me you'll see the offending parameter.

[snow663]

I just reread the thread, and a couple of things stand out.

To me it seems the most crucial bit of info is the part about the symptoms disappearing for a short while, after you "Reset the DME". I'm going to assume this means you're doing a battery reset?

Hell, I don't even recall that early of a car having adaptations , limp mode, or anything beyond a code clearing at reset. Hopefully Abel will check in on this.

If you don't have a scan tool, how do you know there are no codes? (Don't trust "stomp test")

I'm assuming that you don't have ASC+T. (?)

The 1800 rpm is likely to be a significant clue, also. Vanos issue? Just because you can control it with air doesn't mean it's working correctly. Unfortunately, I do not recall being able to watch vanos angles on live data on this early a car.

You're using premium fuel? Is there any knocking from the engine which might cause the knock sensor to retard things? (Maybe unbolt knock sensor(s?), wrap them in foam for testing? Do the battery reset, with this, of course.

I'd personally want to see all the available live data. With your analytical acumen, it seems to me you'll see the offending parameter.

Yes I mean disconnecting the battery and letting it sit disconnected for a while.

I have been relying on the stomp codes do I actually need a GT1 or equivalent scan tool?

There is noise from the vanos cup gear could this be interfering with the knock sensors? Also would it be worth my time to replace the vanos as a whole, and while I'm in there do the chain and tensioners too. I know it's expensive but I'm at about 206k miles

Always use 91 no ethanol no exceptions.

I have INPA installed on an old desktop but I don't have the ADS interface cable. Should I just buy one of those for the hell of it?

Sent from my Z799VL using Tapatalk

[bmwdirtracer]

Get the cable for INPA. That's the cheapest, easiest live data you'll be able to access.

Hopefully, Abel (328Power04) will visit your thread; he knows the particular attributes of each computer generation, as well as INPA, very well. I personally do not deal with early E36's much anymore, particularly on elusive engine diag. Later cars have much greater monitoring abilities....but, that said, I really feel like watching live data is what you need right now. While the scope tells you a ton about actual electron flow, the live data tells you what the computer thinks about those signals.

Unbolting the knock sensor(s) and wrapping them in foam is free, except for your time. (Or, cable tie them to a hose, etc....it's just a test)

200k mile vanos is certainly highly suspect. Call Chris Gant at Dr.Vanos.com, and ask him whether these symptoms sound like an early vanos item. He's a very forthcoming guy, and shares his wisdom freely.

In view of all the work you've done, I'd call guide rails, chains, tensioners and vanos seals pretty much de riguer.

Just my 2 cents.

[dworthy]

Chris, you cannot watch live data with the early OBD-I systems, and that really sucks.

Yeah, normally you will hear empty paint rattle can noises typically start at 2k and goes until 2.5k if the seals are bad. But yes the VANOS does start at 1.5k IIRC in on that car.

In reference to INPA, you will need a cable, most of the time a USB to the 20-pin connector works just fine. However for a better connection, I like using the old school serial port style. I own a Dell D-830 with Win-XP specifically for this task and it works great.

[bmwdirtracer]

Darin, when I was on Maui, I used a SnapOn brick with BMW software to see temperatures and airflow, and some other parameters on cars as old as my '89 E34 535i. I'm pretty sure this was also all available on my GT-1 clone, in "expert mode".

[dworthy]

Interesting Chris, as I have looked all over INPA and cannot find the live data stuff, but never tried the GT-1 in that fashion. I will see about giving that a try on the E-36 to see if that works.