2003 BMW 325xi 2.5L Straight 6 About 120K Miles; Looks Well Kept Multi Strike COP Ignition. History. Customer bought used about 3 months ago

Transcription

1 2003 BMW 325xi 2.5L Straight 6 About 120K Miles; Looks Well Kept Multi Strike COP Ignition History Customer bought used about 3 months ago Claims since day one runs rough on cold start Sometimes MIL would come on soon other times after a couple days DTCs for Misfire 3 and 5 then Injector shutdown then lean code MIL stays one once set until a shop clears codes Used car dealer tried repair (plugs, coolant thermostat, sensor) not better To a BMW specialist (misc. Expensive things) not better To a good general shop (2 new coils and injectors 3 and 5 as DTCs indicated) not better Before customer brought to us she had about $2k of repairs into it She might not have paid for all but i saw the bills lots done to no avail Problem from day #1 is same problem she brought to us Runs rough on cold start Sometimes takes couple days before MIL comes on DTCs for Misfire 3 and 5 then Injector shutdown then lean code seems pattern NOTE: throughout and presently car runs strong and smooth once somewhat warm

2 2003 BMW 325xi 2.5L Straight 6, About 120K Miles; Multi Strike COP Ignition Our Initial Findings Runs rough on cold start,etc. as described in History BUT not immediately DTCs for Misfire 3 and 5 then Injector shutdown then lean code Freeze frame data for all codes consistent with cold start (almost immediate) problem Ambiguous as to which is first (3 or 5) but we think 5 first (why I'll show later) Test drives confirm car runs strong and smooth once somewhat warm Smooth running data indicates #5 seems less smooth but acceptable once warm Coolant level down but owner stated never added coolant and we did not previously for her And previous shops messed with cooling so it might mean nothing more than burping Still this is a BMW with symptoms suspiciously like head leakage so we proceeded thusly Our Initial Tests Overnight pressure checked coolant system lost just 1 psi (possible equipment leakage) Bore scoped all cylinders no visible water detected Also insides looked appropriately dirty (no steam cleaning detected) Leak down tested 1, 3, 5,4, looking for indications in coolant system none detected All cylinders had crank case hiss and leak down above 8% bmw suggested limit Attributed to cold and oil dry cylinders again no coolant air detected Nor from exhaust or intake best we could tell by ear (normally sound scope but not then) Neither leak down nor relative compression suggests mechanically bad cylinder(s) Customer took car back and a few days later left it overnight with us. Following is second round of testing starting with relative compression test done (below) Note things we did not do yet: absolute compression, compression waves; We did swap 3 and 5 coils to 1 and 4 but did not have chance to see results via scan too Will see result of that swap when customer returns

3 1 st Run Cold First Cranking In Yard 2 nd Run Cold Cranking Seems Fine

4 Next test(second round of testing) was scoping running Fuel Injection amperage on a cold start. Below you see a slice indicating a cylinder missing Injection (#5) NOTE: ALL cylinders had Injection on this initial start; engine seemed smooth for minutes; then after a bit (as engine warmed a bit) roughness noticed and then the Injection #5 dropped Unfortunately I fumbled and lost being able to save the continuous scoping that showed that phenomena in action. Below is a slice of the 2 nd run of the Injection scoping after the roughness started. Note: this roughness eventually went away during my next set of scoping which concentrated on ignition. I mention this because it bears on my hypothesis I'll present next and give the data that drives me to it. I wish I had precognition of what I'd see and what may have struck me as important I would have organized my scoping accordingly. I was on a fishing expedition yesterday. When we get the car back in couple days I'll be more organized. For now I am hoping those more expert in BMWese can suggest things and comment on my hypothesis

5 2003 BMW 325xi 2.5L Straight 6, About 120K Miles; Multi Strike COP Ignition My Hypothesis Arrived After Examining Data; Offer Front Up For Your Consideration First a Fact: According to Mitchell This Multi Strike COP Once Warm Has This Is Wave Low Battery Less Multiple Strikes Low Engine Temperature Less Multiple Strikes Higher RPM (> 1390 RPM) Less Multiple Strikes Note: as understand it there should be from one to five voltage peaks depending on engine temperature and the last peak is the most decisive. As you will see later in my scoping there are more than 5 peaks?!?!? Is this right or wrong bad or OK??? As I understand it these are the rules the computer follows to add in or not these extra peaks: Given these rules I assume on cold start there is one peak and then as engine warms multiple peaks (strikes) transition in. I have no reason to think the battery is low on this vehicle and all DTCs occur at idle. Now again I did not know what I was looking for exactly so I did not pattern/setup the testing specifically (unfortunately) but I did start to wonder if the missing is occurring at the transition from single to multiple strikes. This rough running seems to us to be absent initially then occurs perceptively after running a bit then diminishes and goes away when finally warm. Indeed when I revisited the wave forms I do see that transitioning and indeed best I can coordinate in my mind full multiple happened about the time roughness went away. Hypothesis: Ignition transition from single to multiple not right causing this missing. If this is so What causes it? And How can it be fixed?

6 Coil 3 / 5 Single Strike Then goes to Multiple BTW somewhere in this transition the roughness diminishing

7 #1 #5 Coils for 5 and 4 show Higher AMPS #3 #4 5 and 3 trip DTC but 5 is real culprit we think Coils with Higher AMPS Show less oscillations

8 This psdata file will be attached

9 In Summary: Could the transition from single strike to multi strike cause the up front missing? Could the fact that there are 7 multi-peaks (not less as per manual) be a problem? If any of this is possible what is the cause? The fix? Any other things come to mind?