[toykilla]

looks like we still need a good ECU solution!

problem discovered by skidazzle:

QUOTE

Both Tim (t32gzz) and I have had problems with the car running into limp mode when we are boosting above 8-8.5lbs.
Basically the car will go into limp mode, but there will be no MIL (check engine light) commanded on. There also are no OBD2 codes stored... but, if you clear the engine codes with an OBD2 scanner, the limp mode goes away.

The experience is very similar to what happens when a MAF sensor dies on you... there is no revving past 2500, car idles rich enough to make your eyes tear!

We both noticed this happening when we are boosting on the highway... Tim's happens to do it much much more often than me, though.

So, we both go to Jotech to get some testing of his car done on the dyno. We get a couple runs in (5.6lbs, 7lbs, 7.5lbs), and then on 8.5lbs, the same limp mode is achieved.

I start to mention it sounds airflow related... so we do some more passes while looking at the Super-AFC2 for Airflow %..

This is what we noticed:

When the car sees 100% airflow before 4500-4700, the car will go into limp mode. We did this many times it can be duplicated every time.

So, we went home, keeping our boost set to 7.5lbs, and all was well. For the last week, we have been trying to figure out how to get better numbers on what is happening...

Well, today we decided to get more details and facts on what is actually happening...

Here is what we did:
We did some runs at 5.6lbs, 7.5lbs, 7.7lbs and 8lbs.
What we looked at was the following for each boost setting:
1) Max Airflow %
2) RPM @ Max Airflow %
3) MAF Sensor voltage
4) OBD2 Input of AirFlow

Here is what we got:

MAF Sensor output to the ECU is a 0-5v range.
The Voltage of the sensor when the AFC2 says 100% is 5.115v
This corresponded to a MAF reading of 33.872 lbs/min on the ECU.
This is all the air the stock MAF/ECU combo can meter.

At all boost settings, 100% airflow was reached.

Here is a breakdown of the RPM where 100% airflow was reached:
5700 @ 5.6psi
5100 @ 7.5psi
4900 @ 7.7psi
4700 @ 8.0psi (Limp mode achieved)


Now, here are some thoughts:

Seems like the ECU does not like to see Airflow hit 100% before 4500rpm (maybe still in closed loop?!?!). Also, ECU does not seem to care if the number is 100% over 4500rpm... This is why it is critical to have a boost dependant fuel delivery system (like emanage, or DFMU)... The stock ECU just cannot deal with boost!


No one besides Tim and I have seen this, because I think we are the only TT guys running well over 7.5lbs of boost...
You might ask "What about the ATI S/C guys...? They got up into 8 and 9 lbs..."
Well that answer is easy... those centrifugal S/C's do not reach that boost level until they are well over 4500RPM. In fact, those kind of S/C reach max boost at redline.
See, a turbo'd car can reach max boost at 3500rpm...


What's next?

The next thing we need is to see what an N/A car with a Super-AFC2 reaches as MAX Airflow... I wonder if its even close to 100%.
The other thing to get is an OBD2 logged N/A car, and see what the MAX Airflow (computated by the ECU) is...

Tim should be making a post in the general engine forum, asking for help on that.

Only option I see is to clamp the MAF sensor voltage to something like 4.7v... This way the ECU will never see 100% flow. Since we have a boost dependant fuel delivery system, I do not necessarily care if the ECU sees true airflow above what an N/A car would see.

Here are links to clamping MAPs (which is the same idea):
http://www.thedodgegarage.com/turbo_map.html
http://minimopar.knizefamily.net/perf/mapzener.html

As a side note... it would be better to put the diode before the Super-Afc2, as opposed to after it, as then you can still increase the voltage to the ECU by raising the Air correction % in the fuel maps. If you put the diode just before the ECU, it will never go above the diode voltage...




Questions/Concerns?



PS: Props to GaryK, in first trying to convince us all the MAF sensor was ill-equipped for F/I on the car.