850i mystery start problem

[mchen]

My '91 850i, Manual 6 speed, doesn't start normally for the last few days. I can get it started, but only if I disconnect both batteries, reconnect them (disconnect both, key in run position, wait 30 sec., connect left side first, then right), then it will start. Batteries are charged (one of them brand new). door/dash all seems to light up and work. When I try to start, I hear a click (assume relay/solenoid) is engaged, but it doesn't crank. 6 months ago, due to bad DME, couldn't get a reading, so had it replaced. Then we had trouble with EML, so I had EML replace, but with used "Automatic Transmission" version (couldn't find any "Manual Transmission" version, and didn't reprogram the EML). With battery disconnect, EML/DME is reset, means something to do with one of the computers (EML, DME)... or relay? Is there a difference in EML for the manual transmission expecting the clutch or brake to be depressed to start? I don't recall having to do that. Any idea?

Extra info:
Recently, there has been heavy rain, and flooding, so I had to drive through some deep water.
Clutch master cylinder or brake leaking, had to add fluid recently (just noticed, so not fixed yet).

[BMSman]

Have you checked the engine ground strap near the passenger side engine mount?

[mchen]

Will check, but if it was lose, why would it start when batteries are reconnected?

[affa]

Try running with just one battery (the new one, assuming it had the required CCA) battery. Mixing batteries is a recipe for head scratching.

Should really be replaced in pairs.

Does your disconnect battery-restart trick have 100% success rate? Might point us to something.

I had similar problem with another car, a merc. Yes it was the ECU. The battery disconnect cleared the non persistent codes and allowed the engine to start, then it started to collect codes again and will eventually not start again after a few days until batteries are pulled. Pretty common problem with that ecu where the pcb will overheat and curl, causing cracks in joints and tracks.

I don't have a manual so I don't know if it has an electronic shift interlock. I do think both the EML and DME are coded (flagged) for manual and auto. May be revtor can chime in.

Any other symptoms? Clues? Weird behaviors?

Let us know.

[Timm]

I reckon the reason it starts after disconnecting the batteries is because the fuseible link(s) gets moved slightly - they are notorious for this non-start behaviour.

[mchen]

The Engine Ground Strap was fine (but just to be sure, I removed it and wire brushed to get best contact).

Battery remove and start worked three times. I was not able to make it work today (as I was trying multiple combination of connection/start). One battery won't start the car. I am going to get another battery just to change them in pair.

One other odd information is that when I measure the voltage across the front terminals, I get 12.6V. If I measure positive (+) terminal to different parts of engine compartment, they read 12.6V... EXCEPT the intake manifold of left bank, it reads 11.8V. Probably nothing, but odd.

Timm, what is "fuseible link(s)"? Where are they? Relays in fuse box? Why would they move with no battery?

- - - Updated - - -

Got new battery, so now both side is new. Per Timm's & Wuffer's recommendation, I checked the fusible links. There are 2 of them in the passenger side, positive (+) terminal cable. One of them is encased in small black plastic box mounted on the wheel-well, off a small gauge red wire. Opened the plastic casing, and I see the "fuse" and checked conductivity, and it seems fine. The other fuse is inline with thick cable, under double black shrink tube. Instead of cutting out the shrink wrap, I just checked conductivity on the small wire, and it also seems fine. BTW, when I disconnected the batteries, I don't always remove all 4 terminals. Got tired of removing all 4, so I was taking only the negative ground cables from both batteries... so the "fusible links" were never "physically moved" (in case there was a small fracture that could get connected), but could have gotten a jolt.

When I tried to start the car, it clicked once and didn't start. Tried again, and on second try, it did start. EML/DME seems to need to learn to idle/run, as it's a little rough. Not sure I solved it with 2 new batteries, but for now it's starting.

Will have to see if the car will restart tomorrow after letting it sit for a bit.

[cartoonz]

... Per Timm's & Wuffer's recommendation, I checked the fusible links. There are 2 of them in the passenger side, positive (+) terminal cable. One of them is encased in small black plastic box mounted on the wheel-well, off a small gauge red wire. Opened the plastic casing, and I see the "fuse" and checked conductivity, and it seems fine.

That's not it...
There is a 50amp one off the right battery and an 80amp one... neither of these is "off a small gauge wire" Lots of time they look ok, might even meter ok, but they are cracked.
Your problem is either one/both of those are cracked OR you need to replace the "other" battery you didn't do last time...
Sounds like you are conditioned to expect the worst (ECU, spendy computer, etc.)... That's not the cause on this one...

[affa]

Correct, you can only really test batteries and fuses under load. Metering it for continuity is just a quick and dirty way to test fuses. Often a powdery corrosion on fuses is a giveaway.

You can check the voltage drop across the fuse during start to see if it's faulty.

[mchen]

You were all correct, it was the "80A Fusible Link" (under 3 layers of rubber shrink wrap)... there were corrosion. Not visibly broken, but when I jumped the fuse with "jumper cable", the car started. The old batteries may have been OK, but will never know now (they were traded as cores for the new batteries).

[Timm]

Great news - I reckon if I say "It's the fuseible link" to all of the threads on here I would still be right 50% of the time!

[mchen]

It's amazing how much "fusible links" caused problem. I have not seen it until you gave me the words to search for. Even then, I didn't believe you until I cut open the wraps. I own enough BMWs (525i, 740iL, 740i, 850i, X5, 525i) that I should have seen this before, but I've been lucky. I managed to get to work, return, and get gas without having to get into the trunk :-) It was a good day. Thank you all again.

[shogun]

That is always the first thing to check, two fusible links, see also my website, just recently I included the pics under http://twrite.org/shogunnew/data/electrical.html
Whereever the battery is, the fusible link is nearby, some E34 in engine bay http://www.bmwe34.net/E34main/Mainte...sible_link.htm
E32 under rear seat http://www.bimmerboard.com/forums/posts/377728/
More pics http://www.bimmerboard.com/forums/posts/377948

[danby]

Thank you so much -- saved me buggering around with all the relays and possible wiring issues- small fracture on the 80amp fuse butted against the rivet and voila drilled our replaced with a small bolt ans back in action... is there any reason these fuseable links are in place other than an the potential from a short from a massive collision ?just wondering if they really need to be there in the first place?? best regards Danby

[Timm]

You'd find out if you had a loom fault why they are there!