Ecumaster forum

Hi, I encountered a problem with my turbocharged engine (SR20DET) with a PWM Idle solenoid. I calibrated the Charge temp. estimation Table to 0 everywhere, as I did not know which numbers I should estimate... I have tuned very excessively for an acceptable VE Map (advanced lambda fueling model) for my engine for many days with approximately the same outside temperature. At the end, engine response was nice, Lambda not too far away from target lambda, and Idle was stable at lambda 1 for about 1 minute after driving. Then engine started to oscillate around the Idle speed, with lambda oscillating with the engine from rich to lean, but I could not do a thing against that.
One nice day with about 10 deg Celsius warmer outside temperature the whole VE map in low partial load was not correct anymore (did not test higher loads). I fought this engine with oscillations in Revs from almost stall to 1800 and AFR from approx 12 to 16 for over 120 minutes idling until the lambda values became stable at around 0,96 to 1,04 and revs calmed down to 200Rpm oscillation. (to be mentioned here, I tried Ignition control, PID control and both DC and Ign error correction table to fight oscillating engine speed, but nothing seemed to work really good...)

Please take a look at the low rpm and map part of the VE Maps. In the heat soked map, the low numbers after the high plateau are correctly tuned to lambda 1!!!!! is that possible to be the true VE or would a good Charge temp estimation solve that problem?

How do you guys get rid of lambda drift in different environment conditions and heat soke (and oscillating idle)?! Btw I have to pass an emission test in one week

What I thought what may solve that problem of drifting lambdas: a maf sensor. EMU could switch to the maf-table in low load conditions, by map value...
BIG improvement or just replacement of one evil with another?

Hi!

Where is your IAT sensor mounted?
What ignition timing are you running in the idle area?

/A

Hi, timing is at 18 deg in the idle area. My IAT sensor is mounted in front of the engine bay: under the headlight in the bend of the coldpipe. Sorry, I dont have a better/newer photo. What I just found is, that I had an IAT correction of 97% when Idling was that bad. I adjusted it now to 100% and will see if that makes it idle better.
Also, what I may have to mention is, my intake cam has 260deg and 12mm lift, ex cam is stock (246@8mm I think?) and the engine has very low inertia due to lightweight twinplate clutch and aluminum crank & accessory pulleys.

Hi!

Try to lower your timing at idle to 5 deg, low timing at idle usually makes it run mush smother.
What injectors are you running and what fuel?

/A

Ok now as you say it, lowering the timing would make life easier! I use Nismo 740 Injectors with 95 Oktan fuel.
On thursday I will be at the car again and report. Any suggestions how these huge VE values in the heat soked, long time idling engine could exist so close to the idle?

there is a nice baseline example of the charge temp table in the help files.

Please attach here the log from your idle.

Well, this is the log, when I came from the smooth map (start of log) to the rollercoaster map (end of log).
It consists of two parts due to its size.

Thanks for your efforts!

Well, retarding the idle ignition advance in combination with proper IAT Correction setup did the trick! The steady-state Idle is pretty good now,
other than the Idle when driving! What can I do to "catch" the Idle RPM safely?
Fuel cut/resume are set to 2000/1950RPM with spark enabled during overrun cut.
Idle Ignition control has 5 deg. of range in advance and retarding. PID Control is activated too. Also I set the idle speed raise to 250RPM when faster than 2 km/h but this does not seem to work very well, as engine almost every time dies when disengaging the clutch and downrevving from only 2000rpm.

Cheers Clem

HI
Play with Idle RPM ref table this is table for dc of idle valve when IDLE IS NOT ACTIVE so on the area about 2000 rpm rise dc to value when you idle or perhaps more if needed.. should help..