Is my EGR goosed?

[vexedben]

Evening everyone.

For context, 2019 T6 Highline 150ps CXHA with Pendle stage 1 remap.

So in a previous post not too long after I got my T6 I was asking about what felt like turbo lag or extreme version of it. A battery change and OBDeleven later, I did some digging and it seems to be a lagging egr valve. For the most part the egr tracks lovely, within 1-2% of commanded but occasionally between 1200-2000rpm I’ll tap in throttle and be met with what feels like the engine resisting the increased power request, then all of a sudden the feeling goes as the boost takes over and away we go.

I’ve actually managed to get some data out on a run home and run it through Claude to process into a useful graph for me which is hopefully on the post somewhere. The areas where the egr valve appears to struggle is on quicker changes in requested valve positions and the error is huge, but it does then catch up. This seems to tie in with the above description of hitting a resistance, then woosh, off we go as the valve catches up and stops diluting the clean air in the engine.

My question really is, could this be linked to something else or is my egr cooked?

Cheers, Ben.

 

[mmi]

The areas where the egr valve appears to struggle is on quicker changes in requested valve positions and the error is huge, but it does then catch up.

On your graphs the data samples seem to be even more than 10 seconds apart so can't expect to catch any glitches as the actual data samples are actually not recorded at exactly same moment. In reality they are captured serially - but Claude seems to have adjusted them to appear at exactly same moment.

Please post the raw data of the logs.

I would start by recording requested boost vs. actual boost as well as pedal position and turbo actuator commanded vs. actual movement.

---

Below a sample of a VCDS log to show how lively the EGR data actually is (e.g. below in 410-420 seconds at steady run) - each dot represents a data sample (recording approx. 5 samples/second).

 

[vexedben]

Righto. Thanks for replying. Here’s the csv of the run from Wednesday. I’ll see if I can get some data this weekend on the boost pressures. I did find what looked to be a leak from the elbow coming out the charge air cooler, which I’ve nipped up but the feeling still remains although it’s definitely smoother on throttle on.

 

[vexedben]

On your graphs the data samples seem to be even more than 10 seconds apart so can't expect to catch any glitches as the actual data samples are actually not recorded at exactly same moment. In reality they are captured serially - but Claude seems to have adjusted them to appear at exactly same moment.

Please post the raw data of the logs.

I would start by recording requested boost vs. actual boost as well as pedal position and turbo actuator commanded vs. actual movement.

---

Below a sample of a VCDS log to show how lively the EGR data actually is (e.g. below in 410-420 seconds at steady run) - each dot represents a data sample (recording approx. 5 samples/second).

View attachment 325624

Didn’t manage to get any data this weekend. However would you be able to just confirm these are the correct channels? The only one I’m not sure on is the actual boost pressure. Is intake pressure correct in this case?

Thank you, Ben.

 

[vexedben]

So here we are. Managed to get OBDeleven to log. I wish it was easier to navigate but alas. Here’s the full run with egr, turbo actuator, boost spec and actual and pedal position.