The 928 starts fine and will run good for about 10 minutes. After it warms up, the engine starts to burble. At first it's intermittent, then it becomes more regular, then the engine will die. It'll restart, but burble and die. If I wait until the engine cools down again (3 hours), it'll run fine for another 10 minutes.
It seems the engine is running too rich when warm. I smell gas fumes in the exhaust when the burbling starts. I thought it could also be that it's not getting enough spark to burn the fuel. As I've replaced many items in the ignition system and vacuum systems, I suspected them first.
|I checked all the new plug wires for damage and to see if they were going to the correct cylinder. I checked the new distributor cap for cracks and proper fit. I checked the new coil and coil wire. I noticed the coil wire was loose and didn't fit snuggle into the coil. I swapped the coil wire with the old wire, that didn't solve the problem, so I took the new coil and coil wire back to Susan (DEVEK) to get checked out. Jim (DEVEK) found that the new coil wire had 0 ohms resistance, so he lent me another one, but it didn't make any difference.|
I checked the manuals for trouble shooting clues and I talked to Jim and got some suggestions.
Here's what I tried:
Temperature Sensor II
I pulled the plug off the temperature sensor II and checked the resistances. It's located on top of the engine behind the fuel pressure damper. There's a wire clip around the base of the plug that needs to be removed before you can unplug it wire. On my car, someone had already tried to remove the plug with out removing the clip first and had broken off part of the plug. The Manual said it should have 12 k-ohms when cool (14-67 F), 3 k-ohms when warm (68-175 F) and 400 ohms when hot (176+ F). I checked the first two readings, which were right on. As the engine was already having problems when warm I figured I didn't need to get the hot reading to reject this as the problem. Temperature Time Switch (for the Cold Start Valve)
I also checked the resistances on the Temperature Time Switch. It's located on the front of the engine under the fuel pressure damper. It also has a wire clip around the bottom, and is even harder to get to then the Temperature Sensor II. I had to pull the passenger side plug wires off the distributer, and move them out of the way, as well as the vacuum line to the fuel pressure damper. The manual gives three different resistances checks and 2 different temperature ranges. For each of the plug ends to ground there should be no resistance (0 ohms) when cool (below 86 F) and 100-160 ohms when hot (104+ F). It also said the resistance should be 25-40 ohms when cool between the two plug ends and 50-80 ohms when hot. All my measurements were in range except one lead had a 30 ohm reading when cool. Since the engine was running ok when cool, I figured this wasn't the cause of my current problem. Ballast Resistors
|The ballast resistors are located on the drivers side (US) of the engine compartment behind the power steering resevoir. There are two of them, one is suppose to be 4 ohms and the other 6. Both of mine checked out.|
I checked the ground wire from the battery to the chassis held down by a wingnut. It was a little frayed , so I'll order a new one. While I had the ground strap disconnected I put the battery on a charger. The car had been sitting a lot lately, so I figured it'd get the battery up to full charge. After charging I cleaned the contacts and reconnected it. I also checked the ground on the passenger side of the engine, right below the smog pump line. It's attached with a 10 mm bolt, and has been know to come loose. Fuses
I opened up the fuse panel (under the passenger side foot rest) and checked all the fuses with a ohm meter. I also checked the amp rating on all the fuses and replaced 3 that were of the wrong rating. I also spray contact cleaner at the contact points. Spark Plugs
I used a timing light to see if the spark plugs were getting current. When the engine was hot, and burrbling, I checked each plug wire with the induction pick up to see if the light would work. They all seemed to get current. I then pulled a few plugs (3) to check they're condition. They looked good on one side and a little dark on the other. But they didn't look fouled (wet/oily), so they weren't flooded but they weren't burning clean. Fuel Pump
|Even though I don't think fuel is the problem, Jim suggested I check the fuel pump. I did this by jumping the relay (XVII). By connecting the left and right sockets and listening for the whine of the fuel pump I could tell if the fuel pump was working. The ignition doesn't need to be on to do this test, but the battery needs to be connected. Usually you would also check the fuel filter at this time. I had replaced it about 500 miles ago, so I assumed it was ok.|
Jim also said to check the cold start valve. If the valve is stuck then the mixture would be too rich when the engine get warm. The cold start valve is located on the passenger side (US) below the air box. Once the car was warm and burbling, I disconnected the plug wire (blue plug with a black cover) and started the car again. This time the car ran worst. I reattached the wire and the car was back to it burbling state. I had to raise the air box to get to the plug wire, which made it a real hassle.
Bruce Cumming (friend from softball and a BMW mechanic) came by with an Ignition Control Unit (ICU) from a 78 BMW 733? that had the same Bosch part number (0227100008). The Ignition CU is located on the passenger side (US) engine compartment wall, behind the remote jump terminal. We unplugged the old unit and plugged in the "new" one. I started the car, and let it idle for a while. The car ran fine, the idle was high (~1100 rpm), but that's normal when cold. As the car warmed up the idle dropped a little, but stayed at around 1000 rpm, but the burbling didn't come back. We strapped the replacement CU to the engine cross brace, and I took the car for a test drive. It ran fine, power was good, throttle response was good, and best of all the engine was warm and it still had power.
I drove the car back to my garage and listen to the engine. Bruce suggested we adjust the idle. I told him that I still had one disconnected vacuum line, the one to the transmission. I told him I was unable to connect the short hose between the metal vaccuum line from the transmission and the one from the engine. He offered to help. I put the car on the ramps and jacked up the back. He was able to connect the hose after removing a small heat shield. Adjusting the Idle
Now it was time to adjust the idle, first we checked to see if the vacuum was good, then proceeded with the "idle" procedure. This involves a screw in the front of the throttle body and a screw in the air flow meter. The one in the air flow meter requires a special tool (9187) , but Bruce was able to fabricate one out of a coat hanger using wire cutters and a hammer. I connected a tach/dwell meter to the coil (negative lead of the coil and ground) and disconnected the oxygen sensor (green wire at the fuse panel). Bruces turned the screws, clockwise = richer and counter clockwise = leaner. He set the idle to 750 rpm.
The Ignition Control Unit is held on by 3 phillips head screws. I took of the air intake tube and held the upper radiator hose and smaller engine water hose out of the way while I removed the screws. It would have been a lot easier if the water hoses weren't there, but it's not worth removing them. The old ICU then just slide out. I made sure the plug was securely attached then positioned the "new" unit and reinstalled the screws.
I took the car out for a drive, and it ran great. The engine seemed to rev freer, and maintained power up to 6000 rpm. The transmission shifted more smoothly, the clunk (1st - 2nd, and 2nd - 3rd) was gone. I drove for about 45 minutes and it was a joy. I'm so glad that the car is finely run well. Thanks Bruce, and the folks at DEVEK.
A new ICU would cost about $350 and a used one is $175. The actual time to remove and install a new one was less then 30 minutes, but it took several weeks to find the problem.
Copyright 2000, George A Suennen