^ so after all this discussion IS THERE actually a fix for an "off the map" maf readin? besides going piggyback? within the ecu... can you scale the maf readings?

 

Yes.. The MAF Size and Calibration table is where you would look..

 

Swapping ECU's, and trying the 'bad' ECU in a good car will at least confirm if it is totally an ECU issue or a sensor/wireing etc problem

 

Now this is interesting. Please elborate MJ.

You VB guy you.

 

Markus's car ran two ECU's but I don't think we have enough data to determine, and the things is we run the same based calibration in each ecu so it would not be a good test. We would have to take another calibration, copy over the fuel, ignition, ect settings over and then upload into the other ecu to see if there is something else in the code on that particular calibration that is doing this. If this is a 'load overrun' problem then I think we'll find it's only on particular calibration since I've run a bunch to the max load scale and have not had this problem until now.

Malibujack, sound like it's pretty clear in your case that it's doing the same thing as Markus's car. Now I need to get his car back in here, either rescale the MAF or even turn the boost down so we're not maxing out the load scale and see if the ecu stops pulling timing.

 

The service manual says the ecu will advance timing at higher altitudes, the MAF has an atmospheric pressure sensor, would this have any chance of effect? Like if the barometer was doing something

 

I have another ecu to test with now, an 04 that I can try the same thing next week.. Can you throw on an S-AFC and reduce the MAF signal about 25%? That should let you determine for sure its a MAF overrun without doing a forced rescale at the moment..

 

I'll have more info next week, the values aren't well documented yet.

 

Sounds good, first I'll try to turn the boost all the way down, that should bring the load down enough.

 

I doubt its a bad ECU, it sounds like his ROM is the same rom as in my car.. which does the same thing when I intentionally run my MAF signal high in those low RPM regions. It sounds like the group of roms used in the 03/04's have slightly different tolerances for overrunning the ECU maps and how they react to it. Its likely that the US Market roms have a failsafe to prevent damage if the car wildly overboosts, since that is the RPM range where it would likely happen.. Its not that the boost cut wouldn't catch it on a stock rom, but if it ramped so quickly, it takes 1000ms (a second) for the boost cut to kick in.. The easy way to do that is to pull timing like that on an aggressive "attack" or increase in the MAF signal.

Since I'm now able to experiment with two different ECU's I should have more data when my car is back together (I took it apart to upgrade the turbo to a GT35r and 1000cc injectors, which with my blowthrough setup people are eager to hear data on)

 

I missed your post, but higher altitude would result in a lower air density, even though the boost might be the same or higher, on the stock maps, if you look at the path the map takes, lowering the load value (similarly to running it through an S-AFC) by having a higher altitude would result in more timing, which picks up a bit of lost horsepower due to the lower air density..

Since the baro sensor reads anywhere from a vacuum to sea level (1 bar) its unlikely it would throw the numbers higher than necessary.. So you then move on to the air temp sensor, and the MAF sensor.. On average the normal air temp you will see is somewhere around 100 degrees, probably 110 in a dyno room, on a cold day and the car not completely warmed up, it might see lower.

But the Air temp sensor has a normal range it operates in, anything outside that range would likely trigger a CEL code, and go into a safe mode (which could be anywhere from throwing a CEL and simulating a higher air temp to the ECU's internal workings so things can run "Normally", or running limp mode)

That pretty much leaves you with the MAF and its signal with relation to boost levels and the ECU's internal calibration.

 

I think once you go to the slower spooling turbo GT35R, you will not experience any of these symptoms.

I would agree with the RPM range in question. In fact just got done speaking with a client about this exact situation where the AFR's were in check, the boost was reasonable, yet the max timing was 2-3* lower than both the high and low ign maps in the 4000-4500rpm range on a 2005 GSR.

I finally just input the timing numbers the car ECU was putting there anyway and called it a day.

 

Has anyone solved this problem by going back to the stock intake?

I know this has been observed on stock cars. But all the same, if its a MAF "problem" it seems worth a try. Especially if it seems to be a corrupted MAF signal.

 

I have encountered this scenario at the track and I am on the stock intake.

 

I know the source of the problem!
I've looked at the disaassembled dsm ecu code-and i belive evo workes the same.

The "load" is calculates as MAF(V) pulses between CAS signals -so if U have high boost at low RPM -u'll be out of the table-and the ignition is pulled back.

U can reach this condition even on a stock turbo with wastegate shut in cold weather.