Bad Chassis Ground

Last October the starter on my daily driver E46 325i locked up. While waiting for the parts to arrive I drove the 323i for a bit. At Home Depot the engine cranked very slowly and would not catch. Note: this is different than my long unresolved no-start issue.

Slow cranking is a sign of a weak battery or bad starter. But I have a OPTIMA RedTop that is maintained on a charger and the starter was new at the time of the engine swap (just a few thousand miles). I used all of the original grounds and the car ran great for the prior four years (aside from the other mystery no-crank issue).

Clearly, the starter wasn’t getting all of its 12.54 volts, but it was getting some. Since I was at Home Depot, I bought a $30 multimeter and started probing. The battery, jumper post, starter lug, alternator and diagnostic plug all showed 12.54 volts. These are all “hard points” that can be traced back to the battery itself. But when cranking, the starter dropped to 9.3.

Next obvious step was to check the grounds. They were tight. I’m using the same grounds as when the original 2.3 engine came out – from the alternator support bracket to the subframe. And from the oil pan to the same point on the subframe. It’s a M8 self-tapping screw into the boss for the motor mount. It’s not original but it worked. Early E30s also had a similar ground setup. With my new multimeter I checked the resistance of each cable and they passed.

For three nights I went back to Home Depot to try different things. I left a note on the windshield that I was having car trouble and I was working to resolve it. I’m so lucky that they didn’t have it towed away. I think they accepted that I was trying and not giving up. Such a relief that they weren’t jerks about it! But I caved after three nights and had it towed home.

At home I changed out the starter. I dug into the steering column and ignition switch. I disassembled the C101 connector and diagnostic connector, trying to trace the circuit back to the switch. I assumed I had a voltage leak but I admit I don’t really know what that is. I was just looking for a smoking gun. Nothing added up.

After taking a break while Winter made a brief appearance I returned with a back-to-basics approach. This must be a ground issue. I have a proper Fluke multimeter but both meters showed the ground cables had 0.0-0.1 for resistance. Knowing that your grounds can never be good enough, I added a third ground strap from the valve cover to the firewall. The engine cranked like normal!!! Now I’m on the right track.

Here’s an alternate method to test your engine or starter ground. Clamp a jumper cable on your engine (the M20 lift point is good). Clamp the other end to a point on the chassis – hood hinge, strut mount studs, etc. The jumper cables are long enough that you could probably clamp to anything else in the vicinity, such as a toolbox or another car. This is a super ground – large capacity, easy to secure, and a quick confirmation there is a ground issue.

I removed the third strap to cause the issue again. I replaced one, then the other, main ground straps. No luck. I then moved each strap to a different ground origin. I moved each to a different bolt on the engine block. I also tried brand new generic cables from AutoZone. The one constant was the M8 screw on the subframe….

I had scrubbed and cleaned the subframe and M8 screw. I even used the harsh brake cleaner. But it was the one remaining variable. I couldn’t easily try another ground point because there isn’t one. My E21 fenders and frame don’t have another existing bolt or hole of the right size. But there was an existing small hole left over from the original engine install (maybe a wiring bracket).

I bought a new generic 15” ground cable at AutoZone that uses 3/8” lugs. This is larger than the usual M8 (5/16”) size for BMW bolts. So that meant drilling out the existing hole on the fender to accept a M10 bolt. I resisted doing this because I really don’t want a bunch of useless holes in my sheetmetal. But I had to try. A quick drill and bolted up at each end and…. full cranking!!!! Success!

This whole time it was a bad ground point on the body. The simplest of things. But still a mystery on why it “failed” in the first place. The subframe is hard-mounted to the frame rail. And it worked fine for four years. I’ve never heard of the ground failing and couldn’t find it in a Google search. There is an oil leak from the oil return tube (“bitch tube”) but this is common on the M20 and I cleaned the ground area well. Maybe oil has worked in between the subframe and the frame rail? It’s interesting that later E30s have the ground point moved to the frame rail. Maybe this is an old issue that BMW was aware of and I learned of it the hard way.

Bring on the 2020 driving season!