Topic: Cobb APv3 strangeness: DAM dropping to 0.000 during steady speed highway driving

[Durandal]

My car (2007 WRX with mods and Cobb AP with airboy tune) was acting a bit strange last night.

Symptom: Cobb AP DAM dropped from 1.000 to 0.000 during steady state highway driving.

Car: 2007 WRX, VF39, STI intercooler, catted turbo back exhaust, K&N intake, Airboy tune for Husky 94, running on Husky 94 at the time. build thread with more info here

Conditions:
- Happened during highway trip back to Calgary from Lethbridge last night (Monday, October 12th).
- Steady speeds of around 100-110km/h with cruise control on, in fifth gear, for the previous hour or so.
- Fuel level around ¼ tank when it happened.
- I pulled over soon after noticing, checked oil and engine noise. Couldn’t hear anything abnormal in cabin or listening to the engine bay. Oil level was within normal limits.
- DAM stayed at 0.000 for the rest of the 60km or so left to get to Calgary, still steady speed highway driving.
- When I got back into the city and had to accelerate after traffic lights, I noticed that the car would hesitate noticeably while accelerating slowly through low rpms (2000-3500 etc) and would only build boost up to 9.75psi or so.
- No engine codes were thrown at any time during all this.
- DAM still at 0.000 upon parking the car at home.

Possibly relevant info: earlier in the same trip, I used 4th gear wide open throttle at least a couple times to pass semi trucks. Saved max boost level on the cobb showed 19.75 psi for that portion of the journey. At some point around that time the DAM dropped to 0.500 according to the saved min level, but It went back to 1.000 before I noticed it, so I am unsure of exactly what conditions caused it or how long it was below 1.000. Fuel level was roughly ½ tank at start of trip home.

I'm planning to poke around the engine bay a bit more to see if there are any obvious vacuum lines detached, or noticeable oil leaks or spots where there might be a boost leak. But aside from that, has anyone seen this sort of issue before, or recognize the symptoms?

[RockThePylon]

Could be a failing knock sensor, or something rattling like a sunofabitch making noise that the ECU thinks is knock.

[Durandal]

Could be a failing knock sensor, or something rattling like a sunofabitch making noise that the ECU thinks is knock.

How do knock sensors typically fail? Do they read like knock is always happening, or read that none is happening, or some other wonky sort of feedback?

I did a short data log of driving at normal cruising conditions and feedback knock stayed zero the entire time. Described it a bit more in my build thread.

[greatnorthgnome]

First thing I would do is figure out if it is actually knocking. Build a det can and see if its actually knocking where it says it is while data logging. While you do that I would stay out of boost and drive pretty gingerly until you figure out whats up.

[mudferret]

I went through this, and ultimately Sunny adjusted the thresholds to "knock" events as a result.

As mentioned, firstly you want to confirm that it's not knock, and ruling out your knock sensor as well. Sunny used Knock Ears while logging to rule out real knock. My phantom knock always showed up at a certain rpm (~2-2.5k), (0.6 ish?) load, and vacuum range (~-3-4 "psi").

Something was resonating at that point that was fooling the knock sensor. This was all done ~50k km ago, and my car still produces the same power it did when I got dyno tuned. I'm still at a loss as to what is creating the perception of knock however.

[greatnorthgnome]

It seems really common, mine used to do it all the time on the highway too I would get long strings of flkc, around that same load range.
After confirming it was false knock I eventually just raised my knock sensor threshold to disable below .75 and enable above .8 loads. It would never affect my IAM, but it was still annoying that it was pulling some timing for no reason.