Mike Billings

I've heard this mentioned in passing and felt it sounded like it could be used to weed out bad MAF sensors on a car with an unknown MAF history (my 2007).

Couldn't find any details on the internet or in the PIWIS TSI documents (factory manual).

As I recall the basic MAF checks are look for 1.0 V with key on, maybe 1.1 V at idle and a smooth increase as RPM's go up.

But does anyone have any details on the MAF balance test? Conditions? Pass/fail criteria?

On my car, after warmup, I found a 20% difference in the MAF voltages, bank 1 to bank 2. I would guess more than 5% difference would be a fail.

Also I think it would be important to swap the sensors from bank 1 to bank 2 and re-test. That way you would find out whether the imbalance was due to the sensor or due to the engine.

I backprobed the connectors because the Durametric couldn't log both MAF voltages. Curious that the Durametric has several channels that it can log with the exact same name. "MAF sensor voltage" as I recall. I would expect MAF Volts bank 1 and MAF Volts bank 2 but can't find any reference like that.

Looking at replacing with 997.606.125.00 from autohausaz.com, $162 each. Taking a risk with Bosch, not Porsche original replacement.

Ideas?

tguaturbo

Are you buying only the sensors with out the housings? I've been told that's a no no though there are various opinions. Did you try cleaning with a MAF cleaner before testing? As for me anytime I have had a MAF related code come up I just clean them and clear the code and they are good for around 6 months.

nzskater

The housing is static, how would buying new housing come in to play at all, other than being clean?

Brainz

Subscribed for feedback on how this plays out. Could be useful.

I agree with OP that bare Bosch sensor should be fine. The housing is nothing more than a plastic tube.

Mike Billings

First I measured the 20% difference in MAF voltage, bank 1 to bank 2, just using one DVM, trying to hold RPM at 3000, but the RPM was not steady.

Next day I used two DVMs at once for the two banks. Rev'ed to 3000 rpm and captured the the two DVMs with a photo. Bank 1 was only 7-8% low. Swapped the two MAF sensors. Repeated test.

Still bank 1 is 7-8% low at 3000 RPM. So I think the MAFs are working identically.

Curious at idle, it was the other bank (2) that was low, by 3-7%, before and after swapping MAFs.

No idea how old the MAFs are or if they are degraded, but therefore no functional difference in the two MAFs.

Why is the MAF voltage imbalanced by 7-8%? Both MAFs feed the same throttle plate, but maybe a difference in the two conical air cleaners, turbos or engine internal.

I'm doing this work to investigate a P1373 code, bank 2 valve lift below limit. MAF is listed as one of the possible root causes in the workshop manual.

Trying a few tricks and tests with the valve lift solenoid as well. will let you know.

Kevin

Intake leaks, intercooler leaks, failed diverter valve. You have a lot of ducting before your throttle plate. Intake leaks will screw your sensor reading. I'd be looking at your lambda and fuel trim. The ECU will trip a out of range code for the MAF if you have a issue. You can also check your camshaft deviation and variocam movement between banks. Lastly a sticky VNT linkage will also contribute to the above issues indirectly. I'd first pressure test and then check your DV.

Mike Billings

Thanks everyone.

I pressure test about once a month! No leaks there. Pressurizing the lower hoses to intercoolers.

Plan to build a tester that pressurizes the MAFs, down through the turbos, to cover the equipment that Kevin mentioned.

Would like to suggest that when your intake is pressurized, pull one of the little hoses off the diverter valve. Pressurized air should leak out. That proves the changeover valve isn't stuck or connected incorrectly. I believe it's possible to have a problem there that puts vacuum to the diverter valves when you are trying to make boost!

If the intake checks out I'll start a very long thread for P1373. Did a lot of checks in the last 6 months and wrote it all down. I have a lot to say/ask about fuel trim, VNT, datalogs, solenoids, etc.

Test drives only until it's fixed.

Kevin

FYI to test the DV you will need to pull vacuum on the hose/port..
Second, with your P1373>>while it is showing your Bank 2 valve lift, it can also be a lifter from Bank 1.. What will dial this in for you is a misfire CEL on that cylinder as the problem gets worse. I would also change your oil >sooner (less miles) vs stretching it out.
What are your Type 1 thru 6 overrev report and engine hours?

Mike Billings

Kevin,

Will get that data and make those checks. Yes it has significant over-revs and a couple extra oil changes. Thanks

I think I have it cornered on bank 2. I wired the valve lift solenoids to switched +12V.

When bank 1 solenoid is energized at 2000 RPM, 4 things happen:
1. Bank 1 fuel trim goes from 0.95 to 1.05.
2. RPM drops to 1800.
3. Engine note changes.
4. All of sudden, you can hear valves working!

When Bank 2 solenoid is energized, none of those 4 things happen.

I've been watching for a misfire. I'll focus on that more since you said it's one of the right things to look for.

Funny sometimes when the Durametric captures a code, you can click on the code and get more info, but sometimes you can't.

Was thinking take it to someone with a PIWIS, maybe it wants to tell us which cylinder #.

Mike Billings

So, no misfires in the log ever. Cams dev. is always 1-2%. 2300 hours total. Significant over-revs:
range 4 4430
range 5 438
range 6 93

Odd thing with the fuel trim. It will go from +20% left -20% right, and then flip flop the other way, with nothing correlating to it in the log, such as RPM or other sensor (photo). I wonder if my valve lift is intermittent, and causing this.

BTW I think we have a better MAF test than mentioned above. In the log below, we can see two things that show the MAFs are working correctly.

1.On the left, the MAF (orange) follows the throttle angle (gray) about as smooth as I would expect.
2.On the right, the MAF (orange) follows the RPM (blue) when throttle is open.



I think the MAFs are good!

I'll expand the leak testing to include pre-turbo pipes (DO-88) and DV.

I have these very light but stiff wires that hang down from the VNT rods. They almost touch the ground. You can slip a board with markings on it under the car and measure the travel on the VNT actuators. I suppose it's not necessary because any problem there would trip a code. I pulled the rods off to make sure the VNT vanes move (very) freely inside the turbo.