Idle oscillation/hunting

Since I have started paying attention and getting to know my car my car's idle oscillates +-150rpm. Long term fuel trims at idle learn to around -5% but then Short term fuel trims fluctuate +-5% as the idle oscillates. It feels almost like a misfire as it dips down to 600ish rpm and logging confirms the ecu registers a "roughness" event every now and then. Not as often as the Rpm oscillates though.

The thing that really has me scratching my head is that coasting in neutral. It idles fine. Short term fuel trim adjusts the long terms to almost exactly 0%. There is no oscillation.**

First thought - full leak down tests, as fuel trims were all over the show. Leaks solved. Idle issue remains.

Retune on dyno to compensate for an engine that was tuned with leaks. Idle issue remains.

Now i have been throwing myself at learning all tuning myself. Adjusted injector scaling and fuelling to compensate.

Used Injector dynamics' plug and play data to scale injectors, adjust latency, minimum pulse width etc etc.

All fuel trims improved dramatically. Idle issue still remains however. I thought for sure this would fix it.

Did some Maf scaling adjustment to compensate for the new injector data. Idle issue remains.

I've now found a table where Idle Timing is based on speed (above 2 kmph it becomes active, and it is scaled vs RPM). "This is the minimum base timing in idle mode when vehicle speed is greater than the 'Base Timing Idle Minimum Vehicle Speed Enable' threshold."

When I compare it to the idle in gear and idle in neutral tables (which are scaled by coolant temp) the values are different at normal engine operating temperatures the values are different. This is also the only time that the idle oscillates. On warm up, when the timing values are the same on both tables, no idle issues. As the car heads towards normal running coolant temps, the timing is reduced to 12% from 17%. Where as above 2kmph it stays at 17%.

Ignition Timing is not something I've read deeply into or understand with any real knowledge. But, it seems to me to be something worth looking into.

A stock map of my car shows unchanged values compared to my current tune. Which i find strange.

I guess i'm looking for thoughts. Could i be putting the cart before the horse here. Is my crank position sensor actually the problem?

Car info: ADM Subaru Liberty GT Spec B. ECU id: A2UG002T

[img="blob:https://www.hpacademy.com/3a64a835-c3d8-4b1b-8679-32ae0f7ce14f" alt="" alt="" ]

Give the engine what it wants. Sounds like you've identified that it wants - 17 degrees of timing at idle

Idle is tricky condition -- you are trying to balance the torque of the engine with the load produced by the internal and external losses.. So a very tiny variation in load (fan kicks on causing an alternator to create more load) can knock the engine into a new operating condition. And if it the response at this new condition is too much torque then it swings the other way. As the engine speeds up with the throttle closed the manifold pressure drops (more losses) and the engine slows down.

Ignition timing can be a pretty instantaneous response to a load variation. So tweak that idle timing to find a solution that doesn't provide too much response as the engine dips, just enough to nudge it back to the correct idle speed.

Good Luck!