Resolved: 5 Years of Intermittent Poor Running [Long]

[bmwman91]

I figured I would post about this since I have made a few threads over the years asking for advice. The issue is really unlikely to affect anyone else since, well, I caused it lol.

When I first got the car I was maybe half-way through my fourth year of college. A few months prior I had crashed my first 318iS and this was my long-sought replacement. It needed work though. The engine was poorly cared for, and the interior was coated in sawdust (PO was a woodworker & parked the car in the wood shop for years). I was visiting my parents on a weekend & cleaning the interior when I made an oops. While tightening the rear passenger side seat belt bolt under the seat, I pinched that big 4mm black wire that comes off of the battery's positive terminal in the trunk (the 4mm one that parallels the huge 20mm diameter one). FUNT, a shower of sparks! I checked in the trunk & all seemed OK. I finished cleaning & put the interior back together.

By this time it is 11:30PM and I need to make the 2 hour drive back to college for class the next day. Crap. The car would turn-over, but it would not start. I didn't smell any gasoline either, and no fuses in the box were blown. What DID that wire do?! Well, I used a multimeter to check the terminal block up front & I wasn't getting 12V on the portion of the block where the 4mm wire was. I also wasn't getting 12V where I pinched it (I taped it up after this). WTF?! It was getting 12V at the battery...it's just a wire, right?!

Nope. After some digging, I found that there was a big permanent fuse installed on it about 18" from the battery. Sure enough, I had 12V on the battery side & 0V on the other. Luckily for me, I still had the first 318iS' wreckage that I was parting out (much to my parents' chagrin), so I chopped the good part (fusible link is the name) out of there & used some heavy butt splices to replace it. The car started & ran fine, so I encased it all in electrical tape to keep it together & went back to school.

Over the following years, it would give me intermittent problems: stumbling idle, loss of power in the 2000-3500RPM range, bad gas mileage...but all intermittently. I could never nail down what was causing it because it would usually stop misbehaving when I stopped & restarted the car, or let it warm up completely. I replaced ignition coils, oxygen sensors, the AFM, fuel injectors, fuel pump & rebuilt the fuel system, replaced all vacuum hoses...and none of it ever really solved the main issues. The issues improved a little since much of that stuff was pretty worn, but the issues were still there.

Within the last few months, I found that the car would refuse to start sometimes, usually only on hot-starts. It would crank & crank, but no raw gas could be smelled, and initially I thought that the TRE high-flow fuel pump was bad. It seemed like pulling the pump cover & wiggling the power connector fixed it, so I never really got stranded anywhere. The bogging in the 2000-3500RPM range was also getting worse, although really only for the first 5 minutes after a warm-start. Overall though, the car was feeling sort of sluggish. My wide-band O2 sensor didn't show any issues with the AFR either. I was also finding that flooring the car for more than 7-20 seconds continuously led to a CEL (and code 1221 / 1222). Weird.

Anyway, last night I met my dad at his house before a big family dinner thing at a restaurant. I told him I'd drive. Nope, car wouldn't start lol. We took his car & I figured I would solve it once & for all when we got back to his place (he has every tool imaginable). Must be the fuel pump, I thought. Nope. The power connector was not even getting 12V, and the DC motor on the pump showed a proper ~0.5 Ohm static resistance. Hmmm. Up to the fuse box. Nope, nothing blown (except fuse 14...high beams, and I am sure it had been blown for months lol). I pulled the fuel pump relay & tested it...it was fine. Could it be the main relay? It tested fine. BUT, oddly, the 2 hot-at-all-times input terminals to it were only showing 2.3V. Huh? Those are connected directly to the big +12V block on the firewall, right? How the F am I getting 2.3V on here?

The main portion of the block showed 12V...yet 24" away I was getting 2.3V. ?!?! We pulled the harness apart a little & everything was intact. Tracing it showed that the main relay was powered from that 4mm +12V wire that follows the big one. Eventually we figured out that it sits on its own isolated portion of the main +12V block up there, and its little section was also at 2.3V. OK, that explained why the rest of the car was getting 12V since it was on a separate circuit. All the injection electronics are powered exclusively by that 4mm wire. Well, the only thing it could have been was that fusible link or a bad crimp job.

The Solution
We pull the trunk apart & cut all the electrical tape off of the replacement link & splices. Sure enough, one end of the wire just falls out of the butt splice. Knowing myself 5+ years ago, I was tired, in a hurry & I probably crimped the wire & didn't bother to give it the 5lb tug test. So for years, my fuel injectors, pump & O2 sensor have probably been getting 5-10V instead of 12-14V & intermittently getting no power at all because that wire had slipped out of the splice & was barely making contact. This explains why the car ran the worst at RPMs that cause the most vibration lol. It was once again around 11:30 and I didn't have any large enough butt splices, so I opened the used one back up & carefully crimped it onto the wire. It passed the 5lb tug test, so I figured I could at least get home on it. Today I will be replacing it properly with new splices & some adhesive-impregnated shrink tubing.

I started the car, and immediately I could tell something was different. It started up in half the time, the idle was rock-steady, and when I started driving it pulled harder & more evenly than just about any time I can remember. The car runs fantastically now. My mileage has improved, I get at least a stock amount of power & the idle doesn't choke every time I come off-throttle lol. I still get a CEL after flooring it for 5+ seconds, but I suspect there's a vacuum leak somewhere & it get angry with me when in open-loop mode. I will be overhauling my coolant system next weekend, so I'll address it then. So, let this be a lesson. When doing ANYTHING with butt splices, make sure that the crimp can hold at least a few lbs of tension.

Chances are that no other E30 owner will ever have this particular issue, but I figured I'd share since I have been trying to remedy all this for years. Thanks for reading & giving me advice on stuff to check over the years. At least I now have a car with a mostly new fuel system, ignition parts & fresh hoses everywhere!

EDIT
Maybe this could apply to others based on responses elsewhere, and here. It also sounds like corrosion can occur at the fusible link and cause similar issues. So, if your car runs poorly and you have tried everything else, cut the covering off of the main/aux 12V wires in the trunk & look for damage. The fusible link itself is covered in shrink tubing as well. If you have to cut that off, you should find some high quality electrical tape (not the local Home Depot junk) and cover it up when done. Any exposed metal on it is live 12V, and it sits against 20+ year old painted chassis metal. Obviously, the first thing you should do before any of this is DISCONNECT THE +12V CLAMP FROM THE BATTERY! You can find a clean fusible link in a wrecking yard, or as others have done, carefully install an aftermarket fuse setup. I opted for the factory part since it is nice & compact and fits under the original plastic cover.

I'll post pictures tonight/tomorrow of how I fixed this "properly" over the weekend.

[DesktopDave]

Thanks for sharing...if only I had a nickle for every problem I'd caused myself...

I fixed a 325is with the exact same problem.  PO had jumped another car & burned that fusible link out.  After I patched it up with an 8mm double-nutted bolt, the car ran great.  Alas, the rust was too far gone (at least I still have the leather sport seats).  Happens pretty frequently in that situation.  But I digress...

The odd thing is that you still had a CEL - I'd never have guessed you had an intermittent failure on the smaller feed wire.  I always troubleshoot that problem when the CEL won't light.

[bmwman91]

Interesting. So far, responses on bimmerforums, and yours, make it sound like this fusible link gets blown (or corroded) a lot more often than I would have though.

The CEL still comes on if I really flog the throttle. It has to be either a vacuum leak, or the wide-band O2 sensor system irritating the Motronic. I suppose it could be something with the TPS or maybe a cam position sensor, but the car runs perfectly smoothly with the light on. The light usually turns off once I settle at part throttle for 5-15 seconds. It has been doing that to me for years, though. I am fairly certain that it is due to the wide-band O2 system. The code is 1221, which is more specific to the O2 sensor control electronics rather than the output.

[DesktopDave]

Like a lot of things on our cars, I suspect it's a bit under-designed.  I have seen two burned out myself, out of maybe ten cars or so.  Of course, I tend toward e30's that need serious help...

I have the same trouble going down long hills.  I get an intermittent CEL and a 1221 as well.  It goes away as soon as I get on the gas again.  I was figuring a vac leak, vac purge valve or something like that.

[bmwman91]

Well, much to my dismay the car has slowly been starting to behave badly again. It is not as severe as before, but the symptoms remain.

- Idle drops low when coming off of the throttle, bounces 1-3 times & then mostly steadies out.
- Occasional sluggishness from ~2400-3200RPM.

I am thinking it HAS to be a vacuum leak. I did some investigating earlier to try to figure it out. The only leak I have been able to find that would pull in un-metered air seems to be in one of the front timing case gaskets. I can't figure out where exactly. Basically, I duct taped a bike pump to the valve cover breather hose & pumped it a little. I could hear air hissing out somewhere down there. It is a small leak though, and it takes a fair bit of time to bleed off even 1PSI.

It might be that, but I am not 100% sure about how the PCV hose connects to the throttle body. It seems to run to the non-vacuum side of the throttle plates, AND the vacuum side. There is very little resistance between the hose and the non-vacuum side, and a moderate amount between the hose & the vacuum side. It would seem that this is basically a fixed bypass across the throttle plates, with the idle control valve as a second, adjustable bypass. It seems odd.

How much vacuum is the crankcase under during normal operation? I can't imagine that enough air is getting through a gap in a gasket to cause this sort of nuisance, especially if the crankcase is not under full vacuum! This is really really frustrating! It ran so damn nicely after fixing the electrical problem! The issues aren't as severe now, but they are still there!

Can anyone suggest an inexpensive way to do a "smoke test" so I can see where air is leaking out from?

[bmwman91]

I did some more poking around. It seems as though the car remains in open-loop operation mode. The O2 sensor input to the ECU sits at a voltage indicating a constant RICH condition, and the AFR is ~11.5:1 (I have a wide-band system). The ECU should most certainly be correcting for this.

The engine temperature also does not seem to get all the way up like is used to (a problem commonly found with age lol). The thermostat must be on its way out. The temp needle gets just a hair past the 1/4 mark. From what I have read, the ECU will remain in open-loop (and warm-up enrichment mode) until coolant is at or above 70C. So, that could explain the super rich smelling exhaust at idle, and possibly the unevenness. So, the thermostat is crapping out. It may also be that the coolant temp sensor or its connector are old & damaged.

Maybe the intermittent nature of some of these issues (that are becoming less intermittent) is because the coolant temp was slowly dropping as the thermostat died. Perhaps I have been getting the most problems now because the ECU is fluctuating right at the warm-up / warmed-up operation mode boundary. It certainly jives with the problems being most apparent at warm-starts & warm operation.

[bmwman91]

Some updates:

I found that the LC1 wide-band O2 controller's analog output seemed to have a 50-100mV DC offset. So, I looked closer & realized that I used the same wire/lug for the O2 heater power ground and the analog signal ground when I installed this many years ago. I measured 30-40mV offset at the ground wires' junction. So, I separated them. The analog signal ground is now grounded on the same lug as the Motronic on the front battery tray, and the sensor heater is grounded with the HIDs over on the passenger's side hood latch bracket. I got my laptop hooked up & did some runs. The AFR is more or less dead-on at 14.7 (well, it fluctuates above/below this, but stays well centered) at idle & part-throttle. When floored, the AFR ranges from 13.2 to 14 pretty steadily (changes with RPM).

So that helped things. I also cleaned the carbon track in the throttle potentiometer & used a little silicone spray. It was getting a little jumpy when I measured the resistance. That stuff seems to work wonders on old potentiometers, and it seems worthwhile to try it before dropping $80 on a new one.

I reset the ECU as well. I am thinking that doing that is what made the car run so well around the time of my OP. The car suddenly runs really well again. I'll give it a week this time before I go declaring everything to be OK.

I also seem to have resolved an issue where my CEL would come on after flooring the car for extended periods of time (usually 5th gear on the highway). The wide-band controller's narrow-band emulation mode had the max output voltage set at 1.1V. I think that the Motronic expects like 0.8-1V as the max or something. So I adjusted the rich-condition to output a max of 0.9V & have the usual 0.45V at 14.7AFR. I can't get the CEL to come on anymore, which is nice.

There are times when I wonder why I want to own a 20+ year old German car...

[Balkanac]

There are times when I wonder why I want to own a 20+ year old German car...

i wonder the same thing sometimes...