finally, today is the day, where i say, my project is finished and the car is fully operating as it should be, again.
the hardware side was easy, after i got ripped off with the wrong engine as described above, i bought a 2020 GTS wreck with very low mileage and used the engine from there. the 'bad' 2.5S engine i keep stored for 'bad' times
the software side took me really a lot of work, for sure about 200 hours of try & error till it got perfect.
the main problem was, that the accessport doesn't fully support all tables nor that all 'implemented' tables are fully understood in terms of logic flow and dependencies.
so i trapped from the beginning into the tricky DME dtc P061B which always occurred together with a PDK dtc's P1990 & P199A.
When this happened, the car immediately went into limp mode and was almost undriveable, even after restarting it.
Luckily this happened only on some specific 'race' conditions, so the car was usable, but not in every condition:
The car had to be in Sport-Plus Mode, you have had to rev above 5500 rpm with more than 25% of throttle, resulting in turbo boost, which is defined in sport-plus-mode not via the map, but with the compressor ratio, even when there is currently no overboost condition for the manifold. than releasing the throttle full to 0% Pedal value. After 0,5 to 1,5 seconds the car tried to remove the overboost in the manifold via the wastegate to make a fuel cut off. this always went wrong, without any significant reason. and exactly in this condition it shifted itself into 7th gear, rev-limited to 1500rpm and no more shift possible = undriveable.
in normal mode (=eco mode) and in sport-mode (NOT the sport-plus-mode) this problem never occurred.
i have the cobb pdk tune too, but it was not possible to change anything in the cobb torque maps on this platform, which seemed to be the reason of the error - because of different live data between engine and pdk.
by filtering the DME dtc, i could make the error 'on-the-fly' resetable, which made testing much easier without clearing all dtc's. but the PDK error still occurred, but able to reset it with a turnkey.
after about 50 different software versions and hundreds of logs, comparisons of different ecu's, analysis of thousands lines of logs etc i could solve it. it was a very complicated combination of parameters of airflow mg/stk, negative torque for the pdk, fuel-flow quantity (which is not controllable via cobb anymore, thanks to EPA regulations), intake-airflow and exhaust-airflow and some spikes in the stock 2.5 tables on the wrong place.
i managed to have a lambda 1.0 for the normal mode with a flat torque of 330nm from about 2000rpm to 7500rpm. for the sport, sport-plus and boost mode, i have up to 571nm, but also pretty flat curve, which makes the driveability great. in the stronger modes, i dropped lambda regular down to 0.84 for keeping combustion fat, cool and knock free. for the boost button, i have a torque reserve of 20nm for the sport and the normal mode. in sport-plus mode the boost button has no effect, because the plus mode is like a permanent 'boost mode'.
in sport-plus-mode i made the throttle pedal also linear, and not exponential like in the other modes. so it's really a pure form of quick and controlled driving in sport-plus-mode.
in normal mode, the car doesn't feel agile, but is very nice to cruise relaxed and needs very little fuel of about 8L per 100 km.
when you want to show off, put it in sport-mode and it immediately changes to pops & bangs and an aggressive throttle map as well as additional up to 240Nm.
when you want to go really quick, put it in sport-plus-mode and you can keep up with the big toys easily
key features:
- based on stage 1 2.5gts map
- timinig & knock tuned for 91 octane / 95 ron fuel
- vmax 310km/h (277 stock)
- eco torque limit 330nm
- sport/sport+ torque limit 571.59nm from 3500-4000 rpm
- exhaust pops only in sport (from stage 0 stock 2.5gts), none in eco & sport+
- throttle in sport+ linear, other throttles like stage 1 2.5gts
- turbo/wastegate limits: wastegate max 1.35 bar, map setpoint 1.41 bar, overpressure 1.71 bar (results in 1.45 to 1.51 bar max boost)
- knock detection from stage 0 stock 2.5gts
to reflect this also on the instrument cluster, i adapted it to my current torque-curve, too.