Two Cylinders Firing, 1985 325E

[Leirad1985]

I've been trying to revive a 1985 325 E that I think has great potential.

Here are the symptoms:
Cranks very well, but only the # 1 and # 6 cylinders are sparking. The sparks on the remaining cylinders are MIA. I've tested the wires and the cables, I even tested everything on a separate car. I've verified that spark plugs, wires, coil, and even computer can operate an M20B27 engine just fine. I've noticed that the injectors are firing, and that I have gas.
I'd appreciate any insight into what might cause only two cylinders to spark.




1985 BMW 325 E Manual Two door
1987 BMW 325 E Automatic Four Door
2000 Subaru Legacy Outback Automatic

[Manolito]

You did not mention the distributor cap. That is the first place I would look if I had spark at some cylinders and not others. Remember that a crack in a distributor cap looks like someone took a sharp pencil and drew a line from the terminal to ground or to another terminal. The spark follows the edge of the crack, burning it to a flat gray color. Usually visible only on the inside.

[Leirad1985]

I've switched the rotor and distributor cap from a working engine and tested mine as well. The cap itself doesn't seem to be the problem. The plugs are new and tested.
I've been wondering if the DME can some how isolate cylinders, or if the coil could have some kind of a grounding issue. The coil is also brand new.
Feel free to speculate. I appreciate it.

[Manolito]

I am not familiar with this engine, but the way that the coil-on-plug ignitions can track individual cylinders is via the crank position sensor. Yours is at the rear of the engine and reads the flywheel. The DME detects misfires by noting transient variations in RPM through the two rotations it takes to fire all six cylinders. I would pull the sensor and examine whatever it reads (teeth? magnets?) in the flywheel through its hole. If the flywheel looks OK, try another sensor.

[Leirad1985]

I managed to get the engine to start, but only two cylinders seem to be firing. I have determined this by touching the exhaust pipes before the engine gets terribly hot. The cylinders are the #1 and # 6. I'm going to triple check the rotor and distributor cap. Hopefully I find more evidence there.

[Manolito]

Cylinders 1 and 6 run together at the same crank position and 180 degrees different in cam position.

One thing they have in common is that they both fire off the same crank position sensor trigger point, which focuses my attention on that.

I am trying to noodle through the rest of the ignition and valve train and fuel injection to figure out something else that could be wrong that would fire only those two cylinders, and nothing comes to mind. .....need more coffee

[markdas]

sounds like you have verified that only 2 cylinders are combusting (based on the heat out of the 2 cylinders at the exhaust port) just wanted to confirm that you have physically checked each plug removed and grounded to the engine as you crank it over?

Agree with Manolito about the link to the crank position sensor.....check the sensor is good (resistance should be 500 ohms I think I remember)? Also check the toothed wheel on the crank pulley...look for damaged teeth etc. (or is your taken off the flywheel)?

Have you checked the coil from another car?

Engine earths? ECU?

[Leirad1985]

Thanks, Fellows.
I've pulled out plugs 2-5 and checked them against a ground--no spark out of them, or out of a working spark plug from a working car. I've also managed to shock myself with the number 6 plug while doing something similar.

I'm very tempted to test the sensors from a working car.
The Bentley manual recommends a resistance of 960 +/_96 Ohms. Both of my sensors produce about 1000 Ohms directly to the pins at the computer.
The sensors in this model come from the flywheel where you'll find two magnetic sensors. One works on the wheel teeth and the other works based on a metal pin on the flywheel--which I've verified in place.

I decided to push it on two cylinders around the block and see what blew up. Nothing did,
but I did notice that the speedometer was not working. I haven't checked the manuals yet to see if this is in any way related to the computer.

I'm going to brew a large cup of coffee.
Thanks!

[Manolito]

Here goes a theory based on no facts or understanding whatsoever:

I have never seen a system with two crank position sensors, but what if the pin reports top dead center and the teeth fire the spark plugs? Now assume that the leads are crossed to the two sensors, or the two sensors are reversed in their holders.

That would make the car fire only at TDC, which is just about where cylinders #1 and #6 need a spark.

This theory may not be correct, but I think I am getting warm.

[Leirad1985]

It's worthy a try. One sensor gives out the engine speed, and the other gives out TDC. I'll try it anyways.
Coincidentally, the previous owner of the vehicle sent me an email today, asking about the car. When I wrote back about the status, they wished that they hadn't asked.
I'm still enjoying this puzzle.

Thanks

[JayStockE30]

I have the exact same issue, any luck?

tried:
3 different ECU's
different distributer
different wire set
checked resistence on the flywheel sensors, and swapped them around
different spark plugs

didn't check coil, ecu grounding, or rotor because sparkplugs #1 and #6 have great spark, nothing out of the others.

[David_Hussey]

Informative post here. Thanks for the great information.

[Leirad1985]

I did manage to fix it finally. The crankcase position sensor was faulty, but I only noticed that once I swapped it with a working one out of another BMW 325E. The sensor now hangs suspended in a corner of my garage. The fact that the car would start on two cylinders is still unexplainable. I have speculated that the sensor was defective, but able to let the computer figure out some kind of TDC info. I don't know if the ECU has any other way to determine a fall back position if the sensor fails. From most of the posts I've read, the engine simply won't start if that sensor fails.
All that said, my lesson has been to not trust test results when all the symptoms point to the part. I had been hesitant to buy a new sensor, and in the process spent far more time and money on other parts that I suspected would've been the cause, but therein lies the fun of working on one of these things. The car is actually now a daily driver, since my other ride got totalled. Now I have to be extra sharp on how to keep it running.
Other things that have failed since:
Temp sensor, which flooded the engine,
Slave clutch cylinder.