2015 WRX DAM Drop

I have slowly been ramping up the boost as ive been tuning my car using SubaruEdit after figuring out as much other little hiccups here and there and learning the tuning strategy behind these cars. I know the knock sensitivity is insanely sensitive to knock in general. i have decreased the knock sensitivity by 3% in the cruising areas because turning the AC on would read a knock of -2.8 when i first started tuning. 3% difference got it down to reading -1.41. I know i can still go a bit further but was trying to tune it the best i can without desensitizing them too much to where it doesn't record an actual knock event. The issue that rose up is around the 6,000 RPM area. All of a sudden it started pulling timing and then drops the DAM down to 0.688. I have tried reducing timing and lowering the boost pressure back down again and it does the exact same thing in the same area. I also replaced the spark plugs and changed the oil and still have the same results and it happens at that same exact area every time. I also walnut blasted the intake valves ~17,000 miles ago. I was pro tuned for all of the same parts before tuning it myself and i never had any issue in this range and was also running roughly 5 degrees more of ignition timing and slightly more boost as well. So I am currently out of ideas and would appreciate some advice on what else i can try. attachment is from the 1st pull when it started occuring.

Subaru's are common for a -1.41 of knock. Wverytime i log my subaru im seeing a -1.41 of feedback knock. I thorught the same thing but apparently its common in subaru.

The 1.41 knock i know is normal to see, but the DAM drop at 6,000 rpm is what i am confused about. im not seeing anything that would be causing it to just drop the DAM like that. I only do 1 WOT pull each time i flash my car as well so i know it has not ever saw any sort of knock in that load/rpm range where the knock learn is populating on the log.

By default -1.4 is the step size so 1 count results in -1.4, 2 counts -2.8, and so on. A 2 count is serious enough I'd resolve it.

DAM steps also depend on knock severity.

DAM drop is an alternative method of knock reduction the ECU uses, which also only occurs during active knock. Knock occurs, DAM drops instead of FBK or FLK response, but the cause is still knock, so if you want to avoid DAM dropping, you have to resolve the cause of the knock.

Yes the false knock from AC clutch engagement is common. That's why I had the RE add the monitor for AC clutch to the Accessport so we could at least show people it lines up perfectly with the knock response. Under load I've not seen the AC clutch cause false knock. Some mild adjustment to knock sensitivity is safely possible on that engine, but the ECU also misses some knock as well, even with stock sensitivity, so be careful.

Hello, just checking in to see how your tune is going. I hope all is well.

Sorry for bringing this back from the dead but it seems like we are having the same issues and I have been super stumped as I'm on ethanol at about a 40% blend and I'm getting DAM drops at high RPM and load. I'm also monitoring fine knock learning correction which sits at -9.5 and I've noticed it will drop down to 11 in some instances then the DAM drops and it will go up to 9 or so when adding timing back in and I'll see the DAM rise on that occasion. AFR is around 11.5 to 11. I'm super confused as I'm not running much more timing than stock AND I'm on E40. I also have EGR/TGV deleted as well as a front mount IC and catted J pipe. Really annoying since I can't ever finish a ramp run. Any help would be greatly appreciated. Thanks!

OP, might be a silly question (probably is), could it be a lean knock?

Keith, what method are you using for the AFR calculation - is it actual lambda corrected to gasoline, corrected to a 40% blend, or a calculated value from fuel usage?

Keith,

You are requesting an amount of fuel that cannot be delivered within the direct injection window. This causes misfiring and the severe engine knock in your log. The over advanced ignition timing is also compounding the concern. Direct injection requires some additional understanding and HPA has some great materials on it to help you out.

Before more damage is done, I suggest going back to pump gas ignition timing if it was working well without knock, using stock cam phasing tables if not currently, setting WGDC tables to zero, avoiding boost, and watching all of the HPA materials on what direct injection is, how it works, and why there's a window of time where it has to be done, as well as information on tuning ignition timing.

You can always modify cam phasing down the road, but leaving it stock will simplify things as you sort the rest out.

You can always advance ignition timing later once everything else is sorted, but for now advancing it is just causing more harm.

After you make those updates and use the HPA courses to gain the knowledge on what you're running into, I suggest you focus on lambda target and start of injection tuning. When you can safely do a pull like that, then slowly start incrementally adding some wastegate duty to increase airflow, datalog and confirm you never exceed a safe injection window again and you'll have a much better outcome.

Thanks for the replies everyone. As far as what Gord said I am still new to all this so I don't understand what you mean by EFR. I did add about 25% to the injector open time tables as I had read on the cobb tuning guide. I am definitely going through a bunch of the courses at the moment and am willing to take the time to learn. The big reason why I am even doing this is so I have on hand experience which helps me learn. Thankfully I have not sent a rod to the moon yet.

As far as what Mike said I have set everything back to stock as far as Timing and AVCS and I am also going to go back to 91 octane and just start over as soon as I go through this tank of fuel. Wastegate Duty at 0 for now. Once again thanks for your advice I will try to report back with results as soon as I'm able.

Sorry, Keith, it was a typo since corrected - should have been AFR.

Ah ok yeah using gasoline scale because it's what I know and understand but shooting for identical lambda. 11.6 AFR so around .78 - .80 lambda. I've also checked and replaced my plugs recently and they didn't show signs of detonation.