I dont get it man, I posted about 4 or 5 specific suggestions along with an offer to look at the fellows ecu map and see if I could find anything causing the problem

I guess some people just want to be negative

In any event -

I have yet another idea for the gentlemen with the PROBLEM

I bet he may have CAM GEARS ??? If so - try to zero them out

 

I am not sure how mitsu watches the cam and crank sensors in the ecu however I can offer the fact that with Mercedes the crank sensor can and almost always never trips a code when it goes bad. I would suggest takeing a resistance reading of the cam sensor and compare it to a known good sensor. I would also suggest takeing the measurment hot cause a bad sensor can read good at room temp and then go to millions of ohms resistance at operating temp.

 

Its a hall effect sensor, if it was bad it wouldn't trigger an adequate signal and should throw a CEL.

Martin, I can trigger the results any time I want, I suppose that can be considered repeatable, When I set my MAF level back to stock the issue goes away.. Hope that at least can give you a little insight.. Also, take a look at the knock filters, I'm fairly sure at this point that each one represents a different range of frequencies and the threshold for noise in that range.. If you take each one and produce a 3D plot, you'll see what I mean, it looks like a spectrum frequency map..

 

I think he was talking about getting repeatable results from us (the guys offering data)

 

When the CAS goes bad you will know it! I've seen two of them start going bad and the result is misfires.

Al as far as cam gears go we'll try that but the only effect they would have would be the phasing of the CAS relative to crank angle sensor. I've played with cam timing a lot on stock ECU cars and I have not seen this happen before.

MalibuJack,
Are you using ECU flash or ECU+?

 

I agree that the cam gear setting should have no effect as many cars have gears and do not pull timing

I just mentioned it as I ran across a car which had so called "dual" can gear dowel pin locations and the gentleman had installed them in the so called "DSM" setting instead of the evo setting and oddly he experienced the same kind of timing pulling despite my ecu calibration correction

No doubt his case was actual engine knock - however - I did not take enough time to go into additional diagonsitcs as the cam gear was quickly found to be his problem

 

Has anyone considered that the timing number reported by the logger might not be actual engine timing?

 

Both, ECU+ has been almost exclusively used for logging, But I can use it to rough my timing from a flattened base tune so I don't have to repeatedly flash my ECU, once I get the numbers I want, I can put them in the flash and zero out the ECU+... On the screen cap I posted, the rom is essentially stock, I only altered the idle speed, boost/fuel cut, rev limit, and standing rev limit.. To force the MAF reading to go as high as it did, I altered the value on my blowthrough MAF sensor's control box to about 30-50% higher than stock (at idle I'd normally have 30hz, I set it to 58hz which resulted in a completely high reading curve) which let me force the MAF reading up to around 1300-1500hz at 3000-5000rpm and able to get those 240% and higher load sites.. Reducing the value back to 35hz at 1000rpm (forces my car to run lean from the cams, but I have a map for that) and I'm back in the 160-190% load sites which are normal for the car and my sensor's calibration.

 

I can correlate timing being pulled with the dyno graph. You can see the dips and drops as timing varies.

 

Ok,
but the question is when you alter the MAF and increase HZ readings it puts you in the higher load zones, is your timing lower there anways? If you set your highest load zone (260) to 4 deg flat across the rpm range (low and high octane) and crank up your MAF so that it hits that area when you get into boost, do you see that 4 deg timing flat across the rpm curve or does your logger show something lower? Maybe I'm mis-understanding you that's why I'm asking these questions.

 

AMS-

Are you running 4 degrees flat across the upper right hand side of the Timing Map and still getting the "pulling"? I wasn't sure if your Timing Map still has the huge area of negative timing in the upper right hand corner like a stock map.

If you are running 4 degrees up in the high load - low RPM area and still getting the problem = THAT SUCKS! Because you are left with "phantom map" somewhere doing this.

 

By right hand side of the timing map, you mean the highest load cell (260) and no we are not running 4 deg flat across the RPM range, that was just a reference I was trying to make to MalibuJack. But yes in essence we are fixing timing in the low and high octane maps at the highest load level and the timing is dropping lower than that.

 

WOW, that is a tough one. I think Malibu was thinking that your timing in that area was actually lower and you were just tracing a path trough the map that was unusual.

 

Thats actually what I'm saying, although my map does have some load cells with -1's and -4's, 4's and whatnot.. I did go back and flatten the map and put fixed values in that range and it still did the same thing.. When I didn't give it full throttle and the car didn't ramp its MAF signal as quickly, it found its way into the fixed timing that I had set and seemed to end up with a more normal looking timing number (the one that was in the map) I can't test anything else until next weekend though because I took my car apart to put a bigger turbo and injectors in.

 

Thats exactly my original thought.. If I force my MAF reading outside the map, higher than the 260% column, I actually can reproduce what he is describing even if I have fixed timing.. **BUT** if I roll into it and not generate the boost as quickly (slower rise of the MAF signal) it doesn't seem to do it.. It seems like there is an algorithm that will pull degrees of timing in the last column of the map but in my data it seems to depend on how much over you are, and how quickly you are exceeding that 260% values..

Honestly I started rescaling and recalibrating my sensor size and calibration map to accomodate this and not overrun the ECU's mapping since anything over the 260% column is likely going to trigger a failsafe of some sort, at least in that region.

I am in the process of disassembling the ROM and I might have more insight in a few weeks, but obviously not soon enough to make it worthwhile input in this problem.