What governs the air flow uphill the MAF?

[jolecanard]

Hi everybody!

My problem is simple:
The air flow measured by the MAF suddenly drops for a second or two and eventually makes the engine die.

The MAF is not faulty (replaced with a new one and readings are consistent with the engine behavior).
The throttle body is not faulty (replaced with another one and position readings are consistent with engine behavior).

Thus, what is before the MAF and could make the airflow suddenly drop? Filter is clean.

[kpn3nc]

why does it have to be air? why can't it be a timing issue? check the output of your cam position sensor (assuming it's coilpack of course) also check for vacuum leaks...everywhere

[jolecanard]

I assumed it was air flow because of the MAF readings:
when the engine is going to stall, the mass airflow measured by the MAF sensor is lower than it should be (car's idling). As a result the engine "load" increases (basically the throttle opens more), but if the airflow "drop" shown by the MAF is too significant, the throttle opening is not sufficient (fast enough?) to compensate for it.

I don't have the readings saved on this computer, but I will try to post that tomorrow.

I have not paid attention to the timings, so I'll look into that. Overall I keep ending with a supposed vacuum leak... what's the best way to look for it? Soap and pressured lines?

[GTIVRon]

I assumed it was air flow because of the MAF readings:
when the engine is going to stall, the mass airflow measured by the MAF sensor is lower than it should be (car's idling). As a result the engine "load" increases (basically the throttle opens more), but if the airflow "drop" shown by the MAF is too significant, the throttle opening is not sufficient (fast enough?) to compensate for it.

I don't have the readings saved on this computer, but I will try to post that tomorrow.

I have not paid attention to the timings, so I'll look into that. Overall I keep ending with a supposed vacuum leak... what's the best way to look for it? Soap and pressured lines?

When the engine is stalling, air flow readings will be lower....

As far as finding vacuum leaks... old school trick is spraying electrical component cleaner or something pretty flammable and spraying it on the hoses. When RPM comes up a good amount, you found the leak.

[jolecanard]

The thing is that the air flow gets lower BEFORE the engine RPM starts dropping. See the readings below:

First event, car is parked, I am not touching/doing anything but recording

Time [s]	RPM	Engine load [%]	Injector Period [ms]	Air flow [g/s]
7.88	640	18	2.87	3.75    --> idle
8.28	640	18	3.28	3.64    --> idle
8.68	680	15.8	2.46	1.31    --> idle, air flow drops badly
9.08	440	20.3	4.51	9.33    --> RPM drops, car responds by opening the throttle...
9.48	600	28.6	4.92	3.58    --> which saves the car from stalling
9.88	720	23.3	4.1	3.64    --> RPM keeps going up for about a second (lag)
10.28	800	18	3.28	2.81    --> RPM keeps going up for about a second (lag)

Same recording, a few seconds later, still not interacting with the car:

Time [s]	RPM	Engine load [%]	Injector Period [ms]	Air flow [g/s]
21.87	680	15.8	3.28	3.28    --> idle
22.28	680	14.3	2.46	2.22    --> idle, low air flow
22.78	560	14.3	3.28	3.31    --> RPM drops
23.18	440	20.3	4.1	5.25    --> RPM keep dropping a bit, but throttle opened and air flow increased
23.69	560	22.6	4.51	2.86    --> RPM going back up, but air flow is still too low despite wider open throttle
24.09	600	21.1	4.1	3.08
24.57	600	21.1	4.1	4.22
24.99	440	23.3	4.92	4.72    --> Why does it drop? Is that due to the high injector period?
25.48	400	30.1	8.61	3.5     -->  same question?
25.88	920	36.8	6.97	1.44    --> pretty bad airflow despite throttle significantly opened, RPM is going up as a result of the larger air flow 1 sec earlier
26.28	1120	17.3	3.28	1.78    --> same remark
26.78	680	13.5	2.87	2.61    --> the airflow has been too low for too long, engine is dying
27.18	320	17.3	4.1	5.22    --> there we go
27.58	80	34.6	0.41	0       --> :facepalm:
27.98	0	34.6	0.41	0       --> :facepalm: :banghead: :mad:

Basically this recording shows that the air flow drops first, and in some cases, even opening the throttle wider is not sufficient to save the engine from stalling (yes, I am sure 100% the throttle is working fine).
I guess I'll go look for the vacuum leak. Anybody has a scheme of where the vacuum lines are running?