IAC
This is a 3-pin solenoid. Guess it is stock on motronic 1.3 systems for M20 engine.
Measurements I made on it: MembersPage/GergelyLezsak/IdleControl/Measure
Docs found about similar valve: MembersPage/GergelyLezsak/IdleControl/Docs
Pictures I made: MembersPage/GergelyLezsak/IdleControl/Pictures
Driver selection: MembersPage/GergelyLezsak/IdleControl/Driver
Firmware related: MembersPage/GergelyLezsak/IdleControl/Firmware
Actually we've got a hacked firmware to control it. Now the valve is moving smoothly but still has some instability in idle. It might be related to mechanical problems instead of configuration/control. (Not sure)
Configuration and other issues
PID config changes doesn't really alter the rude fluctuation of my idle. It changes in a ~160rpm range without characteristic period. [Here] is a datalog XLS from megatune, I've created a diagram into it also. Note that with fixed iac value the engine's idle is altering appr. the same way. (Maybe compression diff. between cylinders? I've a somewhat modified camshaft maybe with +30 degrees opening.)
- If lower PWM frequency is not enough to get accurate idle then we might think about intentional overshoot (short diracs) when moving the iac.position.
- Marcell played around with PID settings, without success. P setting of 66(hex) and up became divergent so we choosed 33 as a stable value.
- Different alignment of the two signals shouldn't matter, that's almost certain.
Idle fluctuation usually happens (ie. not counting idle control characteristics etc.) if;
- mixture is too lean, specially at lower map values (idle 'hunts' between two or more VE table bins
- smoothed VE and lambda table in the idle area, no change. (I'm going 0.96 lambda this time)
- I've quite high (54-60kPa) map values to get idle. So it seems (std. value is around 35-50 kPa), camshaft duration and overlap definitely bigger than in std cam.
- air leaks inlet manifold, throttle body flanges, in hoses etc. **Tried to spray suspicious areas with starter spray, no reaction.
- manifold pressure that MAP sensor 'sees' is fluctuative ie. not 'constant'. Add a 'little can' (fex. moped fuel filter) to hose between sensor (should install it to hose quite near the MAP sensor) and manifold plenum connector.
- Map sensor pipe has 1.5m length, but that should not cause any problems if it's inside diameter is reasonably small fex. around 2-5 mm, and instrumentation is done properly ie. absolute pressure that sensor 'sees' is constant etc. as per. above.
- Std. M20B25 camshaft timing has relative short duration (around 220 deg). Increasing duration increases overlap (more if lobe centre angle is lesser), which has ill effect on idle stability but, 250 deg. cam should work. Altought - might want to check cam timing. Cam degrees after top dead centre when inlet valve is fully open ? -How to measure that?
- How will I know the engine is at TDC, and the valve is open/closed?
Should check and correct it mechanically and injection parametrically wise and then look the pwmming.
Amount of ignition advance you are using in idle and near to idle ? around 20-23 degrees in this range. Ignition advance at idle might be bit on a 'high side', maybe engine needs over 20 deg. if compression ratio is much less than what it is in std. engine.
- How much degrees do you advice? (where?)