Announcement

Collapse
No announcement yet.

Man vs car (long), a tale of chasing a cutout issue

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Man vs car (long), a tale of chasing a cutout issue

    Over the last couple of days I transcribed notes from my race car "diary" into electronic form and posted them on Bf.c this morning in response to a thread about cutouts. On the chance that this might help someone, I'm posting it here.

    What follows is a long tale of chasing and ultimately solving a cutout problem. At the least it may be entertaining and at best it may help someone with a similar problem.

    10/09/2007

    The car is an early production (06/87) 325is. I replaced the 153 DME
    with a 173 well before this started happening and it ran fine for a
    couple of track weekends. When the problem first surfaced it might
    happen once in a while. But with time it has progressed to happening all
    the time.

    Around 5400-5800rpm the engine experiences a sudden and complete loss of
    power. It seems to occur at higher rpm at the beginning of a track
    session, when the engine is relatively cool, and at lower rpm once
    everything is good and hot. From the driver's seat it feels as if the
    ignition was briefly switched off and there's no stumbling, misfires, or
    loss of power prior to the event. If I persist through the cut-off(s)
    the engine will pick back up and run to the redline.

    I have noticed that the CEL will come on just before the event and when
    it happens the tach will drop a lot more would be appropriate for the
    loss of speed from coasting. On occasion I've seen it briefly go to
    zero. The CEL is so consistent with the loss of power that it can be
    used as a shift light. And the CEL will extinguish as soon engine rpm is
    reduced by an upshift or by accelerating through the cut-off region. It
    doesn't appear to matter whether the critical rpm is reached at WOT or
    at partial throttle. Having to short-shift at 5200-5300rpm on the track
    is a major irritant...


    In an attempt to solve this problem (and fix other problems) I've:

    1) Replaced the O2 sensor (new OE part)
    2) Swapped in a known good AFM
    3) Swapped in a known good 173 DME
    4) Replaced the injectors (new OE parts)
    5) Replaced the crank position sensor (new OE part)
    6) Replaced the TPS (new OE part)
    7) Replaced the main & fuel pump relays (new OE parts)
    8) Replaced the coil (new OE part)
    9) Replaced ignition wires, cap, & rotor (new wires & OE parts)
    10) Cleaned all grounds and primary power connections
    11) Replaced intake manifold gaskets & all hoses
    12) Verified proper operation of the ECT sensor (and swapped sensors)
    13) Proved there are no leaks in the brake booster
    14) Replaced the oil filler cap and dipstick o-rings
    15) Eliminated the evaporative control system
    16) Acid cleaned & sealed the tank & replaced both pumps (new OE parts)
    17) Replaced FPR (new OE part)
    18) Checked C101 for corrosion
    19) As of 10/28, replaced the alternator

    With all of that done and other routine maintenance (timing belt,
    filters, valves adjusted, plugs, etc), the engine runs better than it
    probably has in quite some time. And I now have a nifty collection of
    good spares. But nothing I've done has had any affect on the problem.

    I have proved that the used AFM and DME are good by installing them in
    another E30 and they behaved perfectly there.

    The fault codes stored in the DME are 1251 and/or 1261. 1251 is an injector
    fault and 1261 is a fuel pump relay fault. The only thing I can think of to
    do at this point is to ohm out the wiring in the harness.

    10/20/2007

    I've ohmed out all power and ground connections to the DME, coil, and
    injectors. Nothing remarkable there. But in checking power connections I
    did find that there's about 4.5ohms of resistance to the hot side of the
    coil, which means the ignition switch isn't quite healthy. Just for
    grins, I bypassed the ignition switch to feed power directly to the coil
    and DME "Start" input, to no affect.

    Now I'm thinking that the prime problem may be 1251 (injector fault). If that
    fault, and only that fault, happens after a single event I'm inclined to
    figure out a way to bypass that part of the harness (even though it ohms out)
    to see if that fixes it.

    10/21/2007

    Now I know what it is. The car is possessed by demons. To fix it I need
    a young priest, an old priest, and a large jug of Bavarian Holy Water...

    In trying to see if the injector fault code was repeatable I did several
    runs where I cleared the codes, duplicated the problem, and scanned the
    DME. The reported fault codes were completely random and different each
    time. Maybe the alternator is noisy?

    As a next step I'm going to rig something up to allow in-car disconnect
    of the alternator and see if that changes anything.

    10/28/2007

    Since I had an alternator that just needed new brushes, I swapped
    alternators after fixing the spare. That didn't eliminate the problem,
    but it may have moved the event higher in the rpm band. It was happening
    at about 6k and the CEL would come on a couple of hundred rpm below the
    event. More testing is going to be needed to see if it really has moved
    higher.

    Now I only getting a 1262, which is an idle speed control fault and may be
    related to the CEL that occasionally comes on while the engine is idling or
    during a long down hill closed throttle run. But that code can also be set by
    a stall above 600rpm. So it might be related to this problem.

    I'm running out of the possibilities. The only thing I can think of to
    do now is to ohm out all of the wires in the engine harness. But I don't
    understand why a problem there would be rpm sensitive and not also cause
    cause problems at other engine speeds.

    11/16/2007

    I found the spec for the CPS air gap (1mm or.040") and discovered the CPS was
    a lot closer to the crank wheel than the spec calls for. I adjusted the gap,
    but that made the problem slightly worse. So it is looking more and more like
    a problem with the wiring harness and specifically the CPS signal
    wires. Perhaps there's some damage to that circuit or the CPS connector that
    is attenuating the signal. That would be consistent with it only occurring at
    high rpm and with the problem being worse when the air gap was opened up. I'll
    have to catch up on work around the house before I can dive into diagnostics
    on the wiring harness.

    I know the head needs freshening, but before I pull it I want to run a
    compression test and see if more engine work is warranted. If I haven't found
    and fixed this problem by then I'm inclined to replace the engine harness
    while the head is off and it is easier to get to the harness where it runs
    under the intake (for grand problems, grand solutions).

    03/13/2008

    I found other information that says that the air gap can be .012-.028. I
    re-adjusted the CPS, which made no difference at all. I am seeing a pair of
    fairly consistent codes after an event (misfire on cyl 7 and injector failure
    on both banks). The first would be a coil or coil interface problem and the
    latter could be a wiring or injector interface issue. But then, both could be
    caused by a CPS or power fault. For grins, I tried disconnecting the cylinder
    ID connector and saw that the DME did register that failure. It had no affect
    on the problem, so it would appear that the signal is valid. Likewise for the
    engine temp input to the DME.

    I've pulled the boots on "interesting connectors" to look for corrosion, but
    haven't found anything suspicious. I noticed that when one of these events
    occur that the MPG gage drives to 40mpg concurrent with the tach drop. That
    may imply that the DME is not supplying injector pulse width info to the
    cluster.

    I think I need to find out whether the CPS signal is going away, or whether
    power to the DME is being interrupted. I think the check needs to be done at
    the DME.

    03/23/2008

    Power to the DME is good. I probed to wire from the code relay (ignition on
    signal) and the wire from the main relay (DME power) and saw normal system
    voltage at each. I monitored each wire with a test light while driving the car
    to repeat the problem. The light never flickered or dimmed during an
    event. Therefore I'd say power to the DME isn't an issue.

    03/24/2008

    I think I saw a DC level appear on the CPS signal during an event. That could
    be as a result of chaffed wires in engine harness. It seems a stretch that
    both DME's would have that fault, so the harness is now my prime suspect.

    03/25/2008

    I found what should be a good harness, still wrapped around a 80k mile engine
    that was running great before the car it is in was rear ended. So I'm
    summoning the young priest, old priest, and a large jug of Bavarian holy water
    and swapping engines and the harness. I'm tired of screwing with this and I
    want to race the car at Roebling. The fresher engine should mean that I can
    defer the head rebuild and such and I suspect it'll make a bit more power. It
    will almost certainly burn less oil (a track weekend is 2-3 quarts).

    I'll pick up the engine sometime next week...

    04/12/2008

    Work has gotten in the way of picking up the engine and harness... But I know
    one more item that isn't the cause. I pulled the harmonic balancer to see if
    its rubber mount had failed. That could cause the problem, but the harmonic
    balancer looks fine. No signs of cracking or separation.

    Continued in the next post...
    The car makes it possible, but the driver makes it happen.
    Jim Levie, Huntsville, AL

    #2
    5/21/2008

    I finally picked up the new(er) engine and started prepping it for
    installation. While looking for an oil leak in the vicinity of the cooler (the
    leak was because the fitting was separating from the cooler) I saw that the
    harmonic balancer was wobbling around. For grins I swapped in the balancer
    from the new engine just to see what would happen. The wobble was less and
    that raised the cutout to around 5800rpm and made the problem less
    frequent. I'm onto something now...

    Placing a web-cam and a headlight in the engine bay and running up to the
    problem area clearly showed the harmonic balancer wobbling around. More
    importantly I could see that the crank bolt was wobbling (worn bearings, most
    likely). At the point that the cut-out occurred the harmonic balancer was
    wobbling like crazy. I'm thinking that the wobble is disturbing the CPS signal
    to the point that the DME just gives up. We'll see what happens when the new
    engine is in.

    6/6/2008

    With a prodigious expenditure of effort I got the new engine & transmission
    installed in time to go to Barber for the NASA event. Setting the engine on
    the mounts at 9am Friday morning. Only to have the shifter break too late to
    be able to effect a repair. The good news is that the harmonic balancer and
    crank bolt run dead true. Ah well, I did have a good weekend
    instructing... Note to self, don't ever let an early 911 driver trail
    brake. Exciting things are gonna happen!

    6/14/2008

    The new(er) engine seems to run great. I've flushed the cooling system as the
    car the engine came out of was an auto with a leaking tranny cooler and there
    was a good bit of oil in the cooling system. I've fixed the shifter and as
    soon as I check out the front driver side strut (the shock collar may be
    loose) I'll see if it will run to the redline. I elected not to change the
    harness along with the engine on the premise that one should make one change at
    a time.

    6/15/2008

    Cautious optimism abounds! I've done a good dozen runs to the redline in 1st,
    2nd, & 3rd without once having an "incident". Sticking a pipe on the crank
    bolt with the old engine mounted on my engine stand, I could detect motion in
    the front main bearings! That could easily explain everything.

    7/7/2008

    Okay the problem isn't completely solved, though I can live with it if it
    doesn't get worse. When trying to start on a cold engine, I have to force an
    immediate rev to above 2500rpm or it will die with the same sort of cut out
    and tach going to zero and it may experience one or more events before I see
    +2500rpm. It pretty much requires full throttle to accomplish
    that. Once I've satisfied that condition the engine idles perfectly.

    On the track I see some events (at about 5500-5800rpm) until the engine is
    fully up to temperature, when they just about vanish. Occasionally, once the
    engine is hot, I'll see a balk and tach drop on a downshift just before the
    clutch is engaged.

    I can't think of any fault that would explain this, so I'm at a loss as to
    what to fiddle with. I suspect both have a related cause. I did try a third
    DME at the track, courtesy of another E30 owner, with no affect on the
    problem.

    7/14/2008

    Raced at CMP. While the problem is still there, it hasn't gotten any
    worse. The first half-dozen laps, even if I warm up the engine, are a bit
    dicey. One of these events in a race environment in the wrong place can be
    "exciting" (hint; think massive trailing throttle oversteer).

    10/20/2008

    I replaced the engine harness with the newer one (out of a 90 car). That
    required a bit of modification to match my car's configuration. I'll see how
    it behaves at Barber and RA.

    12/15/2008

    Maybe the harness was the fix. The car ran fine for both events. But with the
    history on this I won't be surprised if the problem comes back.

    3/30/2009

    Well, its back! The car ran great a CMP for three hours in the enduro and all
    through the weekends sessions. Then at RA it pitched a fit on Saturday in the
    sprint race and ran great for the enduro and on Sunday!

    7/6/2009

    This is weird. All through this season the cutout problem has been more or
    less present. The engine may act up on one day, on all three days, nor not
    all. Sometimes the engine just cuts out and sometimes it acts like an early
    rev limiter. I'm out of ideas on how to fix it. Until I get an inspiration I
    guess I'll have to live with it. Depending on where it happens, racing the car
    when it is doing this can be challenging.

    Since changing the harness I've swapped in different relays, CPS, AFM, DME,
    ECT, etc to no affect.

    3/15/2010

    The 96 car was destroyed at RA in August. I've built a new car from an
    88. Only the transmission, drive shaft, and engine harness were transferred
    over from the 96. I have a fresh race built engine in the car. I'm seeing the
    same problem as before. I installed an IQ3 in the 96 car just before it was
    wrecked and can see in the logs the tach signal go almost to zero for 0.2-0.3
    seconds. Depending on which DME I have in the car I consistently see either a
    1261 or a 1251. The problem is far worse now and can be duplicated on the
    street. Where before it was only happening above 5k it now occurs around 3k.

    I have a spare analog input channel available and I plan to sample & log some
    of the DME inputs and outputs.

    4/19/2010

    I logged data for the DME power & ground connections, coil/injector power,
    and fuel pump relay control/ground/output. I saw a decline in the voltage
    supply to the DME with increasing rpm. Cutting the shrink wrap off of the
    fusible link revealed a mass of white powder and a completely corroded link. I
    spliced that out and installed a 60A lug type fuse in the engine bay.

    Coil/injector power and power to the fuel pump relay never varied. But
    coincident with the cutout and rpm drop the ground to the fuel pump relay
    would go away. That combination strongly suggests a loss of timing data to the
    DME. Now to come oup with a way of logging the CPS signal.

    4/22/2010

    I've got it! With a rented digital storage scope in the car, powered by a
    1200w UPS, I can see the CPS waveform presented to the DME go wonky when an
    event occurs. Changing the CPS or the DME makes no difference. The wave form is
    generally smeared somewhat all the time. Only when it gets pretty nasty looking
    does the DME drop fuel relay ground and tach output. I think this means the
    harness is bad.

    I had to repair the harness that came with this car. Some idiot "fixed" a C191
    problem by splicing the wires. That meant having to cut them to get the
    harness off the engine. The loss from cutting out C191 and the additional loss
    from cutting out the splices left the harness too short. A classic case of
    what results from not fixing it right the first time.

    I got a harness from a wrecked 525i for free (there's little call for
    those). I got the engine side of C191 and enough wire to repair the
    harness. So now it is back to original configuration. I installed it in the
    car and could see a crisp waveform presented to the DME. Driving it on the
    street reveals no problems.

    I surmise that the coax in the harness that routes the CPS data from the CPS
    connector to the DME has just degraded with time and heat. It is sort of
    interesting that two harness appear to have had the same problem. If I see
    this sort of thing again I'll cut the harness open and replace the coax.

    11/21/2010

    Since changing the harness the car the engine has run perfectly through dyno
    sessions and a bunch of race & track weekends. I declare this problem solved!

    The lesson here is that nothing beats proper diagnostics. If I'd used a
    digital storage scope to log data in the beginning I'd have prevented over a
    year's frustration and avoided lot of work and parts costs.
    The car makes it possible, but the driver makes it happen.
    Jim Levie, Huntsville, AL

    Comment


      #3
      Phenomenal detective work, good sir!

      So if I understand you correctly, wiring in a good C191 fixed the problem? If so, how does the C191 affect the CPS signal? The C191 is just for the injectors, right?

      Comment


        #4
        Talk about frustrating.
        But thank you for sharing.

        *bookmarked

        Comment


          #5
          No shit!

          I went through the EXACT same problem, with a very similar cause. (Wobbly balancer was only problem)

          That IS frustrating, and you are right a scope is the only thing that would catch it, my meter refresh rate didn't catch it measuring AC voltage, and the muddy signal was enough to hide it.
          -Dave
          2003 Lincoln Towncar | 1992 BMW 325iC | 1968 Cadillac Deville

          Need some help figuring out the ETM?

          Comment


            #6
            Originally posted by Sagaris View Post
            Phenomenal detective work, good sir!

            So if I understand you correctly, wiring in a good C191 fixed the problem? If so, how does the C191 affect the CPS signal? The C191 is just for the injectors, right?
            In order to swap in the third harness I had to splice in the engine side connector of C191 to make that harness usable. I probably would have used that harness when I built the new car except for the botched repair for a bad C191. Because I didn't have the parts to fix the harness at that time, and didn't know that the root of the problem lay in the harness, I used the harness that had been in the previous car.
            The car makes it possible, but the driver makes it happen.
            Jim Levie, Huntsville, AL

            Comment


              #7
              Wow. Sounds like my weekend. Ended up chasing every problem down only to find a gummed up CPS. Sigh.

              Interesting read though! Sounds like hell!

              Comment

              Working...
              X