Misbehaving Motronic Saga Continues - Help Desperately Needed

[SRR2]

I posted about this a few days ago and have done some work since then. Unfortunately, things are not going well.

This is a '89 535i manual trans, 115000 miles. Engine in excellent condition with lifetime Mobil-1. No oil burning at all. Operating temperature range is normal. Passes all emissions tests. When it's working normally, it's smooth as proverbial silk and gets ~27MPG in average usage.

To recap, during the power-down for battery replacement, the Motronic lost its adaptation. Ever since then it has been acting . Idle wanders all over. Sometimes it's at 700 where it's very smooth and steady, and then when it acts up it ranges from 1000 to 2000+. The fuel cutoff at closed throttle operates intermittently. Occasionally closing the throttle over 1500 rpm the fuel cuts off as it's supposed to. Other times, like during shifting, releasing the throttle does nothing. If the engine ran up to 3000 in gear, it 'hangs' at that speed with the throttle closed. The mechanics of the throttle operate freely. No friction or hangup. When this happens the speed gradually returns to around 1000 rpm over a period of 10-15 seconds. As you might expect, driveability is horrendous.

Here's what I've tried so far: 1) inspected fusible links -- nothing found. 2) Removed idle control valve. Cleaned (wasn't very dirty) and electrical test. No problem found. 3) Checked operation of throttle position switch. Closes as required at closed throttle. Resistance of switch is low in closed position, infinite in open position. Voltages from ECU to the switch are correct. 4) Disconnected O2 sensor while engine cold. Erratic behavior was essentially unchanged as the engine warmed up. CEL set after about 15 minutes of driving. Reconnected O2 sensor and CEL cleared in a couple of minutes. Erratic idle/fuel cutoff continues as before. 5) I tried the conditioning technique I found on Shogun's web site. The gist of that is run the engine up to 5K in 1st, and let off the accelerator and let the car slow naturally to idle. Repeat at least two more times. Then let the car idle for at least 5 minutes. This might have made an improvement. 6) On inspecting the engine, some bad rubber has turned up. Minor cracks, not enough to significantly influence operation, which is consistent with the occasional perfect operation of the system.

And it gets worse! In the process of all this I discovered 1) a leaking radiator, and 2) a leaking power steering reservoir/hose.

AARGH.

Anyway, I'm still soliciting advice. I'm wondering if replacing the original ECU chip with one of these EAT chips would provide better adaptation behavior. The $200 for one of these would be a small price to pay if it would help clear this nonsense up. But I'm looking for any other good advice. The simple fact that occasionally it works perfectly is particularly frustrating. That tells me that the passive parts involved in operating the engine (things like hoses) are probably OK, and the problem is most likely in the vicinity of the ECU. Since the throttle mechanics work correctly, the only source of the high idle must be the idle control valve, and that part appears to work as advertised. So it's likely that it's being commanded to do the wrong thing. If I could single out one thing that appears to be the culprit, it's the adaptation algorithm which seems to tolerate ridiculous operating characteristics. (This why I'm thinking the EAT chip might be a viable solution.)

Oh, one more thing. For about 10 minutes this afternoon the engine was idling perfectly at 700rpm. Cycling the A/C compressor on and off made NO difference in engine speed, providing yet another confirmation that the ICV is working right.

I'd sure appreciate some wisdom here.

[emw525E34]

You did not mention checking your AFM. 535 uses a vane type control flap. Pays to check the resistance when you move the vane with your hand. Sometimes these things have "flat spots" that causes similar symtoms you described. Causes " erattic surging too". I have opened up an M20 AFM to check and clean. moving the carbon tracks about 2mm to get better contact. It takes time to do this but a sunday morning doing this fix is rather fun, provided it fixes the problem. It did on my M20 auto some time ago. I have since moved to using 535AFM controlled by Apexi SAFC to handle the voltage conversion. Works wonderfully after days of tweaking the settings.
Looks like your ICV is OK.
For M20 motronic V1.1, it took me 2 days of driving to get the adaptation I wanted. I figured "just guessing" that flatspots on the AFM gets skipped quicky when adapted. Before adaption, ECU tends to get "stuck" on these dreaded flatspots and cause lots of problems. Very difficult to diagnose.

Best of luck. WHile you get the AFM out, give it a good clean with Carb Cleaner, you would not believe the junk in there!.

[infinity5]

have you tried disconnecting both battery leads then connecting them to each other and leaving it that way for several hours. you never know what a reset might fix, it works miracles for some people. also check that all the electronic components are seated properly and clean the contacts as needed.

it might also be of use to note that many times our fuses (and fusible links) will appear fine and unbroken, but won't be any good. try and actually test them with current to be sure. check and replace all your vaccum hoses. check the FPR maybe? i don't know. search the forums, electrical problems are common place.

[SRR2]

You did not mention checking your AFM.

Didn't mention it 'cos I didn't do anything with it! 8-) Weather should be good here today, so I'll have a look at it. Kinda doubt that there's anything going on there, given the onset of the problem was so sudden (and directly associated with the power-down) and when it works, the throttle response is normal, but it won't hurt to test it and clean out whatever crud might be in there. Generally speaking this engine is very clean inside. Even the breather isn't all that dirty, and the ICV had practically nothing in it.

Infinity5: I didn't mention that I reset the ECU before I tried the adaptation procedure I mentioned. I opened the elex box and removed all the connectors for a half-hour. I wanted to do that anyway just to remove/replace all the connectors in there. But that half-hour unplugged guarantees a reset same as your suggestion about the battery leads.

Even though it looks like the O2 sensor isn't at fault, I'm still thinking about replacing it. It's the original so it's now 16 years old. The only thing holding me back is concern about how frozen in there it's going to be. Any work I do under the car will have to be on ramps and I don't know whether the worst-case (drop the exhaust) is a viable job. Maybe someone can comment on that. If dropping the exhaust proves necessary, what new parts should I have on hand to replace it?

Interesting that you mentioned that it took 'days' to accomplish that adaptation. Since the last reset, it prolly doesn't have more than a half-hour of operating time.

[Robin-535im]

No ideas on the Motronic unfortunately, but as for the O2 sensor:

Dropping the exhaust is easy, but that thing can weigh 100 lbs or more so be careful. You can drop the whole thing from the headers back as a single big piece, and support it on something like a cinderblock or a tonka truck while in the process of removing it.

You might want to replace the rubber hangars, and will probably want new crush nuts for the header/exhaust connection. Winfred posted a good idea for getting the 6 copper nuts off at the header/exhaust connection: soak in PB Blaster for a while and Tighten them a wee bit first to break the bond, then you can remove them. Follow the directions when putting the new nuts on (don't crank them down too much).

Soak your O2 sensor in PB Blaster for a while, whack it a bit, soak it, etc. I've found PB Blaster really works, after years of using stuff like Liquid Wrench that didn't seem to do much.

HTH

- Robin

[Kalevera]

PB Blaster is awesome...stuff is in a different league altogether than wd40/liquid wrench (not to say that the latter two don't have their purposes).

[SRR2]

Thanks for the advice. Where does one find PB Blaster? I've never heard of it.