KgS4

that is to say, does the ECU operate the wastegate to achieve the required boost by monitoring the sensor? So in the event of a leak, could it potentially (over)spin the turbos much harder to achieve it's required boost!? That just seems like bad engineering, correct me if I'm wrong.

zapco98

The whole idea of limp mode is to prevent the turbos from over spinning.

DarinS4Turbo

when it comes to N75 duty cycles (wastegate control) which is basically hard-coded in the ECU from what I've learned. So, the ECU will use say an 85% N75 duty cycle and expects 1 bar of boost, if you're hitting 1.2 or 0.8, the N75 duty cycle does not dynamically change (i.e. wastegate does not open or close) to compensate or adjust. So in the event of a leak, it will NOT try to close the wastegate further to compensate.

The boost sensor is mostly used to check the deviation between 'expected' and 'actual', and the car goes into limp mode as Eric said if they deviate too much.

KgS4

produce actual boost?

DarinS4Turbo

boost, based on the N75 duty cycles being used. I suspect the chip tuners start with messing with the N75 duty cycles first to get the boost they want, then put in the associated "expected" boost values for the ECU to check, which will set a DTC and/or go into limp mode if they deviate too much.
But you're right, they have little to do with the actual boost that is produced.

DarinS4Turbo

1) There are K04 cars (zapco98 for example) out there running K03 sofware, and are producing MUCH more boost (1.4 bar sustained through the rpms!) than they did with K03s. If the software was running closed loop, they should make the same boost.

2) I installed a circuit in the boost sensor to lower it's output slightly. As shown in VAG-COM, the output was about 200 mbars less than actual boost. So, if it were running closed loop, it should close the wastegate a bit more to get the reading it wants on the boost sensor. But no, it didn't - the N75 duty cycles stayed the same, the boost sensor just read low, and my ACTUAL boost (per my gauge) was the same. I adjusted and tweaked it even more, and still no change, until finally it went into limp mode as expected.

I have some more data but those are the 2 big ones.

DarinS4Turbo

and it is ambiguous at best in this area - certainly does not specifically say that the boost control system is "closed-loop" around the boost sensor - regulated using the boost sensor, yes, which can simply mean that if the boost sensor reads too high or low, it will go into limp mode and only produce 5psi of boost.

How can you explain the data I presented above? It totally refutes your argument. People are also discovering it is also possible to change the output from the N75 to the wastegate (whether it be a bleed, restrictor, turning the screw, an ECS race valve, lol) and get more or less boost and NOT get defeated by the ECU (via compensating N75 duty cycles), provided you don't go high enough to set a DTC. People have done it.

zapco98

Which are both "post" turbo...

zapco98

"that's why if you have a post turbo intake leak you will blow your turbos"

Thats not true, you will go into limp mode.

DarinS4Turbo

the wastegate and overspinning the turbos in an attempt to reach expected boost. Again, I already tested this.