Shitty cold idle and WOT performance. TPS? Fuel? Help!

[CorvallisBMW]

I'm trying to figure out why my TPS is acting up (or at least I think it's acting up). Here's the situation:

- If I run the TPS as normal, and quickly open the throttle, the engine dies. Instantly. I can see my AFRs go to max, which leads me to believe that the fuel is somehow getting cut when it sees the WOT signal

- If I disconnect the black/brown wire on the TPS, everything is fine and I can rev it to my hearts content. But it appears as though the ECU is not getting the WOT signal then, and so my AFRs at WOT are dangerousely lean (mid 14s to mid 15s, eeek!)

Could this be a bad TPS? Or am I missing something else? Here are a few vids I made showing the results of some multimeter testing and some engine running:

multimeter testing:

https://vimeo.com/134806617

engine running/dying:

https://vimeo.com/134806651

I'm stumped on this, I need to get the TPS working properly before I can tune properly. I'm tempted to just replace it, but at 2 years old it's hardly old, and they're about $80.

My other theory is that the TPS isn't bad at all, and that this is due to a lack of fuel flow or pressure. The car still has the original 325i fuel pump, so it's possible that a sudden spike in fuel demand is simply beyond it's capabilities, so the fuel supply is inadequate (same idea would apply to a faulty FPR). But in this case I would expect the car do stumble and die more slowly, not just come to a screeching halt.

[jlevie]

The first things I'd do would be to test the switches in the TPS and to check rail fuel pressure while driving. With that information in hand further diagnostics may be needed.

[CorvallisBMW]

I've been searching for some information on how to properly test the TPS, but haven't what I'm looking for yet. Do you know the proper measurements to take at which locations/pins?
edit: nevermind, found an excellent youtube video on an old e23 7er, which uses the same exact TPS.

I'm going to rent a fuel pressure tester later today and give that a shot, though I'l only be able to use it while stationary.

[AndrewBird]

The TPS isn't a potentiometer, it's a multi-position switch. You should see continuity on one pin and ground with the throttle at idle and continuity on the other pin and ground at WOT. Mid way should have no continuity.

Use the continuity tester on your meter to check this. It's the one on the far right bottom of your meter.

[AndrewBird]

Check for continuity between the brown/black and brown/blue wires as well. I wonder if they are shorting somewhere and the the ECU thinks the throttle is at idle when actually at WOT.

[CorvallisBMW]

Yes Andrew, I know how to test for continuity :)

I tested the pins on the TPS itself, and everything was OK.

Then I decided I should check the plug as well, so I plugged it in to the TPS, pulled back the rubber boot, and tested the back side of the pins where the wires come in. I tested between pins 18 and 2 (idle) and got the correct readings: 0 ohms at idle, open line off idle.

But on pin 18 and 3, my meter was flashing between 0 ohms and open line, like it couldn't figure out wheich it was. This doesn't happen if I test the TPS directly, only if I test the plug. But once I pen the throttle to 80% or so, it goes to 0 ohms.

If I check continuity in the plug directly without it being atached to the TPS, I get the same strange cycling between pins 18 and 3.

I noticed something odd though, I'm getting 12V on pin 3, which is the black/brown wire that is used to reference WOT. Shouldn't the reference signal be coming in through pin 18, the brown wire?

This led me to check the Voltage on the plug and see what's going on. With the plug attached to the TPS and the key in position 2, I get 12V between WOT pin 3 (black-brown) and ground. I also have 12V between pin 3 and pin 18. This is while the throttle is closed... so the TPS is saying there's an open contact between these pins when I test them for continuity, yet there's voltage between them on the plug! And when I test for voltage across pins 18 and 2, I get 0V when at idle (which is supposed to be 0 ohms, so shouldn't I see voltage?) and then it jumps to 12V after I open the throttle.

This all seems totally backwards to me. I assumed that the ECU sends a reference signal, and when a contact is closed the voltage flows through and back to the ECU, telling what's up. But I am seeing the opposite. Does the ECU treat a loss of signal/0V as the "trigger" for idle or WOT conditions?

Andrew, I get around 18Kohms between pins 2 and 3 on the plug (black-brown and brown-blue)

[AndrewBird]

I'm not exactly sure how it works, but black/brown is power, brown is ground and black/blue is signal. I'm guessing your DMM is reading the signal wire weird and showing a "fake" voltage reading.

I'd check for continuity/shorts between the TPS connector and the ECU connector.

Also, checking stuff with the plug connected to the switch will likely give you weird readings as you are measuring things parallel to the circuit instead of in series. Meaning you will also be measuring the ECU side of things (combined) unless you unplugged the connector at the ECU. Unplug the TPS connector and check things there.

[CorvallisBMW]

I started tracing the plug wires back to where they connect in to the harness, and found some janky connections from when I lengthened the wires in 2010 during the original m30 swap (back when I was young and dumb...). I removed all the butt connectors and electrical tape, ran new wires to the plug and connected it all properly with solder and heatshrink tube. Maybe there was something goofy going on there, I'll test it when I get home tonight.

I'll pull up the wiring diagram and start checking for continuity between the connector and the ECU pins too.

[CorvallisBMW]

Did some testing last night.

After rewiring the TPS harness, the idle is now fixed! It started and idled very nice, even at a coolant temp of 80F, which the day before was total shit. I can't test anything colder since it's like 100F every day when i get home, but I'll get to it eventually.

It will still die though when I go to WOT, so something else is still amiss. I have a brand new Walbro 255 liter/hour fuel pump on order from Amazon that will arrive today, and I'm going to hit up the yard and pull an FPR too, just in case. I'm still leaning towards the WOT thing being fuel related since I see the AFRs go to max when it dies.

[CorvallisBMW]

OK, update time.

I fired it up this morning and it was idling shitty again. Zero change from yesterday, but suddenly I'm back to square 1.

Swapped out coil, no change.

Swapped out FPR (for a 2.5 bar, not a 3.0 bar, because it was all I had), no change.

Rented fuel pressure tester, saw 2.5 bar on the 2.5 bar FPR. Saw 3.0 bar on the 3.0 bar FPR.

What I did notice though was that when I pulled the fuel pump fuse while the engine was running, it suddenly began to smooth out and run super nice, until it ran out of fuel entirely.

All the FPR and fuel pump tests in the Bently check out perfectly. BUT, when I go to WOT and the engine dies, my AFRs go to max (indicating extremely lean). This confuses the piss out of me.

How can I have good fuel pressure but be super lean? The TPS tests perfect for WOT condition, so the ECU should be seeing the signal and increasing the injector duty cycle accordingly.

What about a massive vacuum leak? I don't see one anywhere, but I know if I create a vacuum leak by pulling off the brake vacuum booster hose from the manifold, the idle smooths out. It's a high idle, but it's smooth.

If anyone has ideas, I'm all ears.

[CorvallisBMW]

Possible hit on this. I tested for continuity between my TPS plug and the ECU. According to this diagram The TPS is block #6), I should see continuity between pin 2 and the brown/blue wire, and pin 3 and the brown/black wire:


Instead I'm seeing Open Line, on both. It's like the TPS isn't even being seen by the ECU. Yet if I unplug the TPS, it will hardly run and everything is shit. So the ECU is seeing something. Ideas?

[jlevie]

Remove the backshell from the DME connector and see if the idle & WOT signals reach the DME (with the Cable connected to the DME).