Hello and my question is about the odds of salvaging via cleaning MAF sensor, when I suspect it to be the culprit of the problem.
Brief history: my MAF (Siemens) was changed about 1 year ago, by part-changer mechanic (who I no longer use) - and I started to diagnose, smoke test and do deep dives into real cause. SES light kept coming back with P01717/P0174 vacuum leaks. I eliminated those and for a while the car drove and acted beautifully.
Then, on a hot >90 F day, car going uphill, with A/C on - I got misfires. The lost power, struggled with acceleration, max speed 40 mph.
Reported codes: P1351 and intermitent P1346 or P1347, and P0300.
I replaced old with new coils, spark plugs and injectors - still misfire P1351.
Another mechanic suggested that it must be an electrical problem. After checking the wires, and seeing nothing suspicious - I unplugged the MAF sensor, drove - and all the power, acceleration came back, and misfires GONE.
The new driving (with no MAF plugged in) gives an abrupt jolt when I slow down with a break, before fully stopping. Thus, before I spend another $400 on a new MAF sensor - I decided to clean it. I cleaned it thoroughly - but it only helped for 2 miles, SES came back and misfire P1351 is set again. What should I do? Try to clean it again, empty the whole can? Does cleaning MAF sensor ever work after the contamination? Is there other way to test the MAF electronically - I am not sure if it hasn't overheated and fried in the guts. No MAF specific code is there - when I plug it in.
By chance are you running the aftermarket k&n filter?
Sent from my SM-G531H using Tapatalk
If you have a K&N type filter. The one where you oil it, so that it traps the dust. If you over oil it, the extra oil contaminates the MAF. Cleaning the MAF sometimes works.
Normally with multiple misfires, I would expect to see lean codes and you would have a rather large vacuum leak to find. Unplugging the MAF makes the DME run the fuel mixture on tables inside the DME. The mixture is run a little rich. This tends to marks vacuum leaks, so the engine runs better and the owner thinks the MAF is faulty. Smoke testing being the best way of finding vacuum leaks. As you don't have lean codes and you have a scanner. Do the following test to eliminate vacuum leaks as a possible cause.
With a hot engine at idle. Check your fuel trims, 2 short term and 2 long term. If any are >8% then you likely have a vacuum leak. Rev the engine to 3,000 rpm for 30 seconds and watch the short term fuel trims. If they reduce significantly, might even go negative, then its confirmed that you have a vacuum leak to find.
Another thing to check is the connections on the DME. Sometimes oil or coolant can wick up the wires from a leaking sensor and contaminate the pins on the DME. This has been known to cause misfires.
Last edited by Muzz258; 03-24-2021 at 11:24 PM.
1999 2.8L Z3 Roadster,
2000 3.0L Z3 Roadster,
There is only one thing more pleasurable than working on a Z3, that's driving it top down on a fine day.
Sounds like some good info above
Sent from my SM-J727T using Tapatalk
Kn filters are junk
Thank you all for responding!
Firstly: no K&N filters - everything is stock, i.e. a square air filter box from 2002.
Lean condition: I eliminated them all, changed many rubber hoses, gaskets, O rings, CCV and fuel purge/breather valves.
That saga was filmed into 6min video here ---> https://youtu.be/O6YTTiGFhZI
I smoked the intake and the exhaust – 1) intake: no leaks in intake smoke pressure builds up quickly and spits out the smoke feed hose; and 2) fuel exhaust line: smoke goes to the tank, passing fuel expansion tank, to dust filter and out the air vent pipe (in the back).
Fuel pressure: no problem, ~50psi at the fuel rail, and I changed fuel pump a few months back.
Fuel trims: unfortunately when MAF is plugged-in, I get “open loop fault” on my obd2 reader and I don’t know my trims: O2 sensor is deactivated – STFT and LTFT are fixed at 0. Once I unplug MAF – I get 0…+6 on SFTF and -3…+2 on LTFT, thus confirming no vacuum leaks, and that is at idle – when I drive – trims looking even close to 0. Obviously, those are vary a bit when I rev without MAF– but it all simmers down to 0. Since “open loop fault” happens when MAF is plugged in – I suspect the problem with MAF unit guts.
Circuit continuity from PCM to MAF: I did check the voltage with key-on-engine-off (KOEO) and then car running. I don’t have the oscillograph, or other time graphing device. Used a simple voltmeter and here’s my readings for the 3 sensor pins:
KOEO: 10.62 V at power; 0.00 V at signal; 0.00 V at ground.
Car is running (parked): 13.6 V at power; ~0.05 V at signal; 0.00 V at ground.
I expected signal voltage at ~4V, but I could not confirm accuracy –perhaps voltmeter is averaging a changing reading over some time. I would like to conclude that circuit continuity is OK with such readings. However, the oil wicking into the computer is a possibility – and I will check, especially on that signal wire – so thanks for the tip!
Also – do you have an opinion about where to buy MAF sensor? ECS is quoting $376; Amazon only $156. This is for model 5wk96050z, from Siemens / VDO. Big price difference, not sure why.
It will not work in 95% cases. If you guys look inside and do not see any thick layer of dirt (or oil ) on wires I strongly recommend to leave it like it is. Have been through this many times.
Last edited by besembo; 08-08-2020 at 03:01 AM.
You could squeeze maf pins and trace them to dme. Test them for continuity, its very easy.
Thanks, besembo.
So far, that is my unsuccessful experience with MAF cleaning too. Cleaned it another time (insanity) - no change, rough idling, misfire, code P1351 (cyc 5 misfire with fuel cutoff).
Do you know what are the MAF pins in computer connections? Could I find it on the internet Or do I need to buy a wiring diagram somewhere? A new investigation territory for me.
Wire harness photo.png
I did voltage experiment on the middle MAF pin (signal wire) in the unplugged wire connector. With unplugged MAF, parked car - I revved it loud to about 3000 RPM, but the signal wire voltage was only about 0.04-0.05 Volts - which is very low for revving scenario. So now, will need to check voltage on MAF pins of computer. But which pins?
Ok - googled and found this gem on newtis.info:
https://www.newtis.info/tisv2/a/en/e...-meter/nLVflKd
Bookmarks