Before I go on, I do like working on cars but I am not a professional, just a frustrated IT person, so some terminology may be incorrect and I am still learning some fundamentals so it’s quite possible I am missing something simple here.
Symptoms: Car will not rev beyond ~3800 RPM. The sensation is like a rev limiter in low gear, when in second or third the sensation felt is very jerky, the car coughs and splutters and will not rev beyond 3800 RPM. Up until 3800 the car performs well, smooth power delivery through all gears with no hesitation or complaint at a variety of throttle openings.
Précis from data log analysis: Engine timing is pulled like crazy above 3800 rpm.
A little about the car and engine. E30 BMW. Motor is a 1995 3 litre M3 motor (S50B30). Engine is completely rebuilt with new everything except block, crank & head. Dry-sumped, running 11.6:1 CR, cams are 292/279 intake and exhaust with 12.45 and 12mm of lift and 1.6 and 3mm of overlap. Aftermarket, high performance lightweight everything internally and every engine sensor brand new OEM. Engine is single VANOS with variable timing on intake cam. This old style VANOS is pretty simple, off/on I believe. The travel range of the adjustment has been reduced (more than halved) to mitigate piston/valve issues. VANOS is currently disabled (far as I can tell).
ECU is Motec M800. Motor built by Pip Barker, M800 wired by GR Motorsport Electrics. I would like to think both are done ‘correctly’.
Cam specs can be found here - http://www.catcams.com/DesktopModules/SaveAsPDF/DownloadPdf.aspx?Name=1303413&Url=http://www.catcams.com/portals/0/SaveAsPdf/pn1303413.html
This is a new engine with 400km on it now. For the first 350km engine never went beyond 3500 RPM so this problem was never detected until the first time it went on the dyno to tidy up the supplied base map and tune it a bit further for the next extended run-in period before some serious dyno time is scheduled.
Little bit more info:
a) Initial REF/SYNC configurations have been confirmed by MarkMc at Motec and idle timing and CRIP all been set with a timing light (by someone who knows what they're doing).
b) Dual wideband lambda is running and initial tuning done with quick lambda to date
c) Digital Input 1 is set to 19 Cam Control
d) Aux 5 & 7 are set to Cam Control 2
Troubleshooting to date: There have been a couple of sporadic attempts to troubleshoot the issue myself and some short time from tuner. Ran out of time on the initial dyno tune day to do serious troubleshooting but suggestions to date have not helped the situation.
e) Magnetic sensor levels, voltage filters. Have captured REF/SYNC at 3500RPM to assess these main signals and their cleanliness. Each is ~20volt peak to peak and very clean. SYNC is spotlessly clean and REF exhibits minor sinusoidal interference of about 1.5 volt peak to peak. Voltage filters have been set above this value and even up to 6 volts to really rule out this as a possibility. Filter levels currently set 0.7volt@ 1000RPM and then a flat 2 volt for the rest. This is true of both REF and SYNC.
f) Magnetic sensor levels, duration filters. The supplied map (for an S54B32) had quite extended values on the duration filters. This has been brought back to almost default values for both REF and SYNC without noticeable effect. SYNC filter currently 92 Usec @ 500RPM, 44@1000, 20@3000, 16@6000, 8@10000. REF is effectively the same.
g) SYNC Cam Position. This is a setting I am not sure about. Currently set to Offset 100. I am not really sure what this is used for and need some input here please. Wondering whether if this setting was wrong it would almost manifest itself at all RPM or only as engine speed increases.
h) Aux outputs for cam control have been disabled to see whether that had any impact on issue. None at all. Problem still persists.
You can find the current configuration file here if interested -[url] https://dl.dropboxusercontent.com/u/122 ... asline.e35
[/url]
Data Logging Output
I was able to log some data today to get a better understanding of what is going on when the problem occurs. Looking at the data through i2 is quite revealing. Looking at a second gear roll as revs increase from idle up to about 3760 the ignition advance increases along tracking pretty much to what is set in the ignition map. This is about 28 degrees at 3700 RPM then as it gets to 3770 it starts to pull timing rapidly so that at 3786 RPM timing is negative 2.6. This goes on as the revs bounce up and down around 3700-4000 RPM the peak of ignition retard is about -47.7. All of this is happening at a constant ~37% TPS. Injector duty cycle is probably understandably all over the place when this is happening. Lambda wise, when its running well up to ~3700 lambda is ~.93-.95 across both sensors. When this fault occurs it drops to ~.75-.8. Soon as you let off the throttle everything returns to normal again.
It is probably easier to link to the i2 data log file but that is my interpretation for those who won’t or can’t review it. The file can be found here https://dl.dropboxusercontent.com/u/12244271/20130820-1981301.ld and the period to look at is between 12:08 and 12:17.
Happy to solicit opinions here on what to look at next.
I must admit I did not own a timing light until the weekend, and it is still in the mail from some chap in Adelaide, so I can’t check physically check the advance on the motor as its rev range increases. I hope to do this Thursday when it arrives.
Thanks all.
Burkey
