Unsolvable mystery: Are you the detective for the job? (with video)

[ghatid]

I changed the title to get a little more interest:

My 1991 325ix hesitates/shudders when I release the gas suddenly, but it runs smoothly if I release it slowly. It also shudders sometimes if I just tap on the gas when it's sitting in neutral. When I do the stomp test, I'm getting 1215 (AFM), but the idle sounds good and stays at a constant ~700 rpm.

Does anyone have any ideas what's going on? It basically feels like the car has to take a while to realize that I've released the gas pedal. Here's a video of it with me tapping on the gas. http://www.youtube.com/watch?v=JRoCkDxpToU

What I've tried:

1. Cleaned/tested ICV. Using a 12V power supply, I checked to see if the flap moves. I also tested the resistance between the pins and get 25, 24, and 45 ohm. Bentley says it should be 20, 20, 40, so I think it's close enough. It also buzzes when I have the key on 2. The problem is, it doesn't seem to affect the idle if I unplug it while it's running.
2. Cleaned/tested AFM. I opened up the AFM and changed where the tracks go as well as checking to see if the voltage changes constantly while plugged in and it does. I followed the tests in http://www.the944.com/afm.htm. Is it worth buying a used one to test it out?
3. Cleaned fuel injectors - done by Mepeh.
4. Tested the TPS. There is continuity right when the flap is just about to be closed and when the flap is about 90% open.
5. Checked for vacuum leaks. The breather hose was really messed up, but I replaced it along with the fuel pressure regulator hose. The vacuum is strong enough that when I put my hand on the air intake into the air box, it feels like it's going to suck my hand into the car and leaves a suction mark, but the engine doesn't die. Is it supposed to? There is a whistling sound that I can't find the source of (it's not the vent hole on the bottom of the air box).
6. Tested the temperature sensor (blue one)
7. Cleaned the throttle body.

[Gregs///M]

I would start with ruling out vacuum leaks. I believe your symptoms are related to a possible vacuum leak. Perhaps the oil vapor hose under the intake manifold, the oil tube under the IM to block, or perhaps the gasket from the block to the IM. There are many possibilities for an old engine and a smoke test is the only definitive way to find them.

Also, related to #2. That article about AFM's has incorrect information. For AFM's for our cars, and most other AFM's after 1986, you must use the correct regulated input voltage to properly test calibration. A 9V battery is a poor substitute.

Also, you cannot test an AFM with a digital multimeter. DMM acquire voltage based on a average over time. So the range testing you are doing is actually an average display of voltage. Only an oscilloscope is capable of acquiring realistic voltage feedback of an AFM. Poor AFM's exhibit brief voltage drops in the scale of microseconds, which will show up on an o-scope but never on a DMM. Using the original 20+ year old AFM is crazy! See my AFM thread for more tips and info.

GL

[jlevie]

I would swap in a known good AFM, reset the DME by disconnecting the battery to clear the fault code, and see if the code returns. If it does the fault will be in the engine harness or in the DME.

[ghatid]

I haven't gotten a chance to check the vacuum pressure yet, but in regards to the AFM, it seems like there isn't a problem with mine, unless it takes a while for the computer to notice that there's a problem.

I did some reading about the codes clearing (because of jlevie's hint) and read that the computer doesn't clear the code until you do a reset even if they AFM is fine. I had unplugged the AFM while testing to see what was wrong with the car, so I think that's why it lit up. I ran the car for a few minutes but the code didn't pop up again. Hopefully that solves the code.

Now I just need to do a leak test. Will update.

[ghatid]

Okay, so I bought a cheapo vacuum pressure gauge and it looks like my car idles at around 14-15. I shot another video of the issue (I don't really know what to call it) with the gauge showing as well as the AFM open. I don't know if that's suggested, but I'm not driving the car around and I had it open because I was moving the contacts to a new track.

Does this video help diagnostic any? I was measuring the vacuum from between the fuel pressure regulator and the intake manifold. https://www.youtube.com/watch?v=qlwkYnL2EKA

Does it seem like I might have worn rings? From this forum, it looks like it is a possibility:

http://forums.pelicanparts.com/101-projects-discussion-forum-bmw-3-series/616151-89-325i-low-intake-vacuum.html

"Open & close throttle quickly Drops to 0, jumps to about 20 May confirm worn rings (especially if idle shows only about 15-20) Verify with a compression or leakdown test. "

Would having worn rings make it have trouble getting back to idle as well as suddenly lose power when I stop accelerating? Or valve seals?

[ghatid]

Crap, it looks like I have a cylinder with low pressure. I did a compression test and got:

170, 155, 155, 115, 150, 160.

Then I dropped in some oil into the 4th cylinder and the pressure went up to 170.

All signs seem to point to worn rings. What could have caused this? Is there anything else that could be wrong?

I'm guessing my next step is a leak test.