This page is about Hi-Z injector tuning, at a BMW M30B34 turbo engine.
These are new 48lbs Hi-Z (12 Ohm) Accel injectors
Did some measurements on the injectors: MembersPage/GergelyLezsak/InjectorMeasurement
Still couldn't get reasonable control over injectors. With normal values of inj opening I can't get normal idle. At a bit higher RPM, VE table needs to fall below idle area. It started to show somewhat good idle when I changed injopen to 400-800us, rampup to 2000-2400us. No need to say this made low loadsites(60-80kPa) to be pig rich because it can't give narrow pulses at all then.
Interestingly higher loadsites (100-120kPa) were also disrespecting VE table at all, I could change VE from 30% to 255% without significant change in lambda value. In fact the only thing really changed behavior (and lambda) was the injector opening settings and req_fuel.
I even wonder how the car's operating so smoothly with an untunable VE table.
New data here:
I simply can't get the car normally idle/drive. It wobbles between 500-1500 rpm rapidly or stumbling around at 800 rpm -no matter how much air is allowed (via MDI commands). It changes between these behaviors without any change in config.
- This behavior is cured by now. It needed ignition based idle to be turned ON, and everything was changed immediately.
All the time I see high (1.3++) lambda values and feel heavy smoke (lot of unburned fuel). Fuel is still burning in the exhaust slightly (definitely missfires).
I'm not suspecting mechanical problems because this car war running (almost) perfectly before we pulled the old injectors and Genboard (with problematic map values) out.
Both ECU is 1.0.38 firmware with powerflyback. I saved and reloaded the old config to the new ecu, changed MAP range according to the High-boost sensor in the new one, disabled pwm-ing for the new injectors.
Later I added the 30V transient suppression diode also (made no change at all). I've the feeling that I totally can't control these injectors. I can change req_fuel from 2 to 8ms without noticable difference in idle. Situation is similar with injector characteristics value.
Ignition advance is also interesting here, I've flatten the whole idle-and-nearly area on ign and VE tables, still my advance gets retard somehow. (Hope it's not ALS, I did my best to disable it.)
- I suspect it was ignition based idle control again.
Here are a bunch of configs, tables, msq, and xls logs we had tried: