One for the WTF files.

Since I acquired the E28 from Ransom and all during his ownership it has had a rather rough idle. He chased vacuum leaks, I checked engine and ECU grounds, leakdown and compression, injector leaks and spray pattern. I also redid his work with the ignition system and installed new cap, rotor, plugs and wires. All of this to no avail. It idled like a V8 with a big cam. It also always seemed a bit down on power and try as I might it never did better than 12.8 MPG and for most fillups it was spot on 12.1. I drive a lot of short distances in a day. Average trip is about a mile so I accept miserable economy as a fact of life.

Last week I replaced the electrical portion of the ignition switch because of an intermittent no start on hot days. I was carrying around a screwdriver so when it failed and I was at the grocery store or something I could bridge the terminals on the starter and fire it up. Luckily this time it failed at my shop and I just so happened to have the free time. I popped the column shroud off, replaced the switch and it fired right up. Sweet!

The idle bumped up about 50 RPM and was super smooth. Running around town it seems crisper, perhaps not more power, but different. Today I filled up the tank and did the math. 17.8 MPG and I replaced the switch at the 3/4 tank mark!

So yeah. Not exactly sure how that happened but it certainly isn't a placebo effect considering I thought I was fixing the starter.

Maybe a huge amount of resistance in the switch for the "ON" circuit, causing lower voltage at the ECU and sensors, and more load on the alternator which would drag your idle down? That's my best guess.

resistance, or it was cutting out subtly and quickly enough that you only noticed it was a slight miss?

I've been favoring the mad_machine theory. I don't know enough about the ECU internals to know how much of a "boot process" there is with a momentary disconnect.

A bunch of resistance in the ON position seems like it might wreak all sorts of havoc with what the ECU thinks it's seeing (and doing), but I don't see it causing a load on the alternator? Unless it were either leaking to ground or doing something that caused the voltage regulator to read incorrectly, but it neither cooked nor drained batteries (nor melted anything else).

If its electrical and possessed or behaves like magic, its a ground issue.

in my e28 the ground by the battery was a little sketchy - more exposed to the elements - a replacement strap really helped it run better. Also had a weird issue - heavy rain would cause massive stumbling/misfires - water shorting out an 02 sensor wire and sending funky signals to the dme. Had to unhook the battery to reset.

Amazing how poorly a computer will work, but still work, when it is only getting 9v or so. And that voltage probably fluctuated a bunch depending on what was getting power cycled.

Ransom wrote: A bunch of resistance in the ON position seems like it might wreak all sorts of havoc with what the ECU thinks it's seeing (and doing), but I don't see it causing a load on the alternator?

Lower voltage in the ON circuit means lower voltage feedback to alternator, alternator works harder to compensate, generating more heat where there's resistance. I had a similar problem once from an alternator output wire burning up.

It's always nice to solve more problems than you expected, I'll keep that in mind if I ever have strange issues as a possible cause.

oldtin said: Also had a weird issue - heavy rain would cause massive stumbling/misfires - water shorting out an 02 sensor wire and sending funky signals to the dme. Had to unhook the battery to reset.

I've had this issue as well on my 86 535i in slush/snow which had the O2 sensor connector under the car by the sensor, for 87 and newer they moved that connection up to the firewall to prevent that problem. If you ever experience it again you can just disconnect the O2 sensor wire at the firewall where the wiring harness goes through it, there will be two green wires plugged into each other there that can just be unplugged until things dry out or are sealed better under the car.

Adam

NOHOME wrote: If its electrical and possessed or behaves like magic, its a ground issue.

That sir is an incredibly true statement.

appliance_racer wrote:

NOHOME wrote: If its electrical and possessed or behaves like magic, its a ground issue.

That sir is an incredibly true statement.

Reminds me of the time my Samurai's radiator fan would only work if the driver's side door was closed. It was a weak ground for the fan in the front grille panel, when the door was closed it would straighten up the body enough to get good contact.

Pocket oscilloscopes are quite possibly the greatest debugging tool for electrical issues in regards to vehicles. It makes figuring out of your ECU is actually getting a steady 12V really easy.

My first E28 (535i,) I chased rough idle. Turns out they just idle like crap, according to my BMW-meister mechanic buddy. Ran GREAT everywhere but idle. I just reset my expectations compared to modern BMWs and Japanese cars.

Less resistance in the lock cylinder probably affected a ground issue somewhere else. I now have a ritual for addressing grounds, tightening fuse holders, cleaning relays and window switches, etc when acquiring old BMWs.

I love a case study that STARTS with the solution. Just knowing that it's not going to be left hanging for all eternity makes me happy to read it.

Also, congrats on fixing the car!

GameboyRMH wrote:

Ransom wrote: A bunch of resistance in the ON position seems like it might wreak all sorts of havoc with what the ECU thinks it's seeing (and doing), but I don't see it causing a load on the alternator?

Lower voltage in the ON circuit means lower voltage feedback to alternator, alternator works harder to compensate, generating more heat where there's resistance. I had a similar problem once from an alternator output wire burning up.

Thanks, I'm one notch better-educated on alternators! But doesn't that still suggest that stuff would get melted or the battery cooked, given that the alternator would be trying to charge a system that didn't need it? Or maybe we just didn't reach that point yet...

Ransom wrote:

GameboyRMH wrote:

Ransom wrote: A bunch of resistance in the ON position seems like it might wreak all sorts of havoc with what the ECU thinks it's seeing (and doing), but I don't see it causing a load on the alternator?

Lower voltage in the ON circuit means lower voltage feedback to alternator, alternator works harder to compensate, generating more heat where there's resistance. I had a similar problem once from an alternator output wire burning up.

Thanks, I'm one notch better-educated on alternators! But doesn't that still suggest that stuff would get melted or the battery cooked, given that the alternator would be trying to charge a system that didn't need it? Or maybe we just didn't reach that point yet...

It would cause a buildup of heat at the point of resistance, and could cause battery overcharging too depending on where that is.