The restriction is improvised from the tip of a pen and heatshrink. The other parts are from WebShop.
The solenoid (2pin injector-type connector) is wired as:
- INJE, EC36pin9. The flyback for this channel is populated
- injector common (switched +12V)
GenBoard/Manual/DigitalOut/Table says that h mask is 10 (decimal 16) for EC36pin9.
- h=04 01 02 08 10 20 40 80
- according to above h table-line, boost_channel=40 must be used (see GenBoard/Manual/DigitalOut/Table)
After the electronic and pneumatic connections are made, the basic test reveals if the setup can work.
- with engine off, you can hear the solenoid (ear close to it) click when you switch it on/off.
- with PID all 00, with boost_minpressure=00 the valve prells. We confirmed that our valve prells with 4*8msec period (or higher), but not with 3*8msec (or lower)
- boost_channel=FF temporarily disables the solenoid being altered by the boostcontrol function
- mdh40 (or mdh.., replace .. with the boost_channel's value according to your wiring) switches the solenoid off. This must prevent boost from reaching boosttarget. In our case MAP reached max 140 kPa. If MAP goes beyond boosttarget, either
- the restriction diameter is too small (below dia<0.6mm we had this)
- or the valve leaks even if unpowered
- leak elsewhere
- the wastegate actuator likes to bypass for some reason (stuck?)
- mdhC0 (or mdh.., replace .. with the boost_channel's value + 0x80) switches the solenoid 100% on. This should ensure boost goes beyond boosttarget. In our case the fuelcut_max_kpa=0xA6 (166kPa) was easily reached by the MAP in really short time. If MAP does not reach boosttarget, either
- the restriction diameter is too big (above dia>1.5mm we had this)
- or the valve does not open, even if fully powered (we had this problem first, when the solenoid was powered from a mysterious supply - that measured OK without load - only allowed 170mA instead of the normal 420mA - it must have been connected in series with the iac. When the solenoid was powered from the injector common node, it was all good).
- leak elsewhere
- the wastegate actuator hates to bypass for some reason (stuck?)
Theoretically misc1 (misc1out_minmap, misc1out_maxmap, and all other misc1... varaibles) can be configured so the valve is activated between 110 and 150 kPa, and playing with the ...maxmap=150kPa (doing a boosted run with each setting) it can be reveal if the actually reached maximum MAP changes as expected. Simple systems uses this (but this is not real boostcontrol). For some reason this did not work as expected (we didn't investigate it deeply).
Don't forget to enable boost_channel=.. after the experiment (set back, or just reboot the engine if you haven't saved the "disable" value).
MAP-target config - somewhat OLD text (was written with experimental firmware a long time ago)
- the boost_targetoffs is in 2kPa unit
- the b is also 2kPa unit (with MAP-target).
- This was different at the pre0.0.0 version tried on this car
- note that boost-target table unit is always 2kPa regadless of kpa_conf (that effects most kPa related variables, notably k table and kPa thresholds: 1kPa for normal boost and 2kPa unit for high-boost setup). The b is always in 2kPa unit (unless the installer intentionnally cheats with kpafac to have a fake MAP reading so higher than 512 kPa can be controlled - we can rule this out).
- the config.boost_minpressure is also always 2kPa unit, regardless of kpa_conf
With the small turbo, at RPM=2500..2800 the pressure built up very fast, overshoot was almost impossible to keep <10kPa (with fast pressure buildup at other places). So we slightly decreased the b for that RPM.
# offset is 0 (t=00 scales down target pressure to 0) boost_targetoffs=00 # unit is 2kPa, target b=40 48 4C 50 52 51 50 4F 4E 49 47 45 # appr 0x50=decimal 80= 160kPa # so effectively max MAP-target is # boost_targetoffs + b = 160kPa # we don't activate the solenoid under 110kPa: boost_minpressure=60 # appr 110kPa boost_conf=41 # 4*8=48 msec period, MAPtarget boost_pid_kp=26 boost_pid_ki=40 boost_pid_kd=10 boost_pid_ilimit=52 # the multiple boost_pid_ki * boost_pid_ilimit=5248 sounds good, anything close to 6000 is nice (above 8192 would be stupid). # note that the PID is not fine-tuned using logs, just by feel fuelcut_max_kpa=AC # 172kPa (previously A6=166kPa) fuelcut_min_kpa=0E # this is 1kPa unit !!! t=10 40 60 80 A0 C0 F8 FF
After the tuning
I spenden a lot of time to tune the boost solenoid, it still not absolutely perfect, but it close to that. :)) You can check the results in my config file. And [here] is a log about some driving. You can see how the things work. I suggest, to watch the log with Megalog Viewer.