latest news,
2005-04-05
I just noticed i have to clean this page up, i'll do that later
Been installing some cars with V3's and got into the same problem on 2 cars. We have
- PWM-able IAC solenoids on both of them, and idle is getting extremely high
- the solenoid is wired to a FET
- there is low voltage GenBoard/Manual/DDFlyback (installed and tested) on the channel driven by given FET
Tried a config like this: What can be wrong?\n
iac_step_seq=D8 #// config.iac_conf bits: #define power_off_iac 0 // power off stepper 1 step cycle after each move #define low_power_halt 1 // don't move iac if batt < 8V #define stepper_iac 2 // enable stepper based iac #define precise_idle 3 // 0: on/off 1: pwm/stepper based #define advance_control 4 // 0: off 1: enabled # so for PWM solenoid: # advance_control + precise_idle = 0x10 + 0x08 iac_conf=18 iac_max_steps=BF iac_tps_thres=0F iac_cold_idle_temp=BF iac_warm_idle_temp=C7 iac_cold_rpm=8C iac_warm_rpm=6E iac_cold_start_pos=99 iac_warm_start_pos=73 iac_afterstart_rpm=0A iac_afterstart_duration=30 iac_afterstart_steps=04 iac_kp=0A iac_ki=05 iac_kd=13 iac_integral_speed=36 iac_integral_limit_dec=30 # set smaller iac_integral_limit_inc=50 # set it higher iac_integral_deadband=0A iac_deadband=05 iac_pid_conf=01 iac_overclose_interval=0A iac_ref_pos[0]=99 iac_ref_pos[1]=99 iac_ref_pos[2]=99 iac_ref_pos[3]=8E iac_ref_pos[4]=87 iac_ref_pos[5]=80 iac_ref_pos[6]=80 iac_ref_pos[7]=7A iac_ref_pos[8]=77 iac_ref_pos[9]=73 iac_sol_channel=e0 # do you intentionally invert it? 60 would not invert it. In any case these use the injector channel(s) from the element6 (index is 0 based) of h[0] iac_ign_advance_change=50 iac_ign_retard_change=50 iac_ign_advance_limit=01 iac_ign_retard_limit=01 iac_ign_threshold=01 # from tables.txt: h[0]=.. .. .. .. .. .. ?? .. # the given h[0] element is very important ! It can switch any number of injector channels, likely 1 or 2 in this case # 1 injector channel enough for 4.5A, although normal 2A continuous ES2J SMB diode is a bit on the weak side
The way to test the precise idle HW is through mdi.. command (either stepper or solenoid).
Eg.
- mdi01 commands a very small (1/256) duty
- mdi80 commands 50% duty
- mdiC0 commands 75% duty (or iac_max_steps=BF might limit duty to 0xBF ?)
- mdi00 lets the idle controller control act (according to the configuration)
This way you can get an idea if it's a hardware/wiring issue.
- measure voltage on the configured channel useing +VBATT supply as base
- measure current consumption through the solenoid
- check MAP
- listen to engine
- etc...
mdi01 stalls the engine (take away air; I stop my engine with mdi01 if I want to keep ECU power to save data to MegaTunix without overwriting EEPROM config by saving with mcs) unless there is a leak or something.
Also, with mdi.. you can test what reference positions are good in function of coolant temperature. You will configure somewhat higher reference positions than those, and let the iac PID integral rest below zero, close to iac_integral_limit_dec value. (So the integral rails at negative iac_integral_limit_dec before the valve could close too much and recover slowly).
2004-12-20
I can report that the volvo 960 is running really nice, Only some idle problems, could be caused by a bad solenoid, as well as config. I took it for a 300km drive yesterday, works perfect.
Next car to install on is a Volvo PV, a really old car with a high powered engine. Project page at MembersPage/EmilLarsson/VolvoPV
2004-11-30
Installed a V3.2 board in a Volvo 960 16V turbo, MembersPage/EmilLarsson/VolvoNineSixty
*2004-11-01
On this friday i'm going to Lycksele to try to fix some problems on the Volvo and tune it to run as smooth as possible.sensor :)
Tasklist:
- Change the backlight power resistor to 22 or 27 ohm, it's pretty dark in northern sweden now.
- Test BoostController, We have a solenoid from an Audi S4
- Check that megatune really works, changing all of the VE bins, and that things are saved to flash when they should
- Something is strange, idle and low-load is hard/impossible to tune.. need to find the cause and fix this. I'm not sure how big your injectors are (and if you drive them correctly, 1 pulse per 2 crankrotations), but GenBoard/Manual/Config/InjectorOpening should be very useful.
- Check dwell settings, If anyone has dwell values for a stock volvo/bosch coil, please tell me
- What is needed to enable knock? Knocksensor connected (there are 1..2 small RC onboard as well to check).
- If anyone has other tests they want us to do, please list here :)
This weekend I installed a v3.2 board in a Volvo 142, equipped with a newer turbocharged engine (B23ET), with a motronic ECU stock, We cut the stock motronic connector and crimped the econoseal receptibles to those cables, We only connected new wires to WBO2 and trigger (Used a distributor with builtin VR Sensor)And installed a couple of fuses and 2 relays, The motronic transplant was with no fuses at all..
It was hard to get it running in the beginning, did a big mistake when we were to adjust ignition advance, Made the mark ATDC instead of BTDC ;)
Had to set a large injopen time (1.7ms) to get a decent idle, got too much fuel otherwise..
The last mistake was, when we almost had it running fine at 0.9 bar of boost, we set the acc enrichment way too high, and killed the sparkplugs.. The car owner lives 160 km's away from me, hopefully the car works with new plugs now, otherwise i'll go there next weekend or something..
FW problems
my_make at http://kombi.ulkhyvlers.net/vems/files
f -mmcu=atmega128 -o vems.elf
comm.o(.text+0xca6): In function `srl_generic_P_upper':
/home/vems/firmware/comm.c:823: undefined reference to `pgm_read_byte_far'
make: *** [vems.elf] Error 1
rm injconf.o eventqueue.o spi.o phasesignal.o timing.o benchmark.o ringbuf.o ignconf.o multitooth.o menu.o knock.o
The above is related to old avr-libc. Upgrade to newer avr-libc (or winavr). Maybe Mik will commit a small workaround ifdef, but easiest for now seems to just upgrade.
Engine dyno test
I've been dyno-testing a stock volvo carburator engine, first with it's stock points ignition system, and then later with the Genboard 3.. We have still to make fuel plumbing to try injection, I had a gain of almost 10hp with Genboard-ignition (101.7BHP peak ;) I still have a ancient firmware (early may or something), will update that and try some more.
The ign.advance raises with raised rpm! with the same values in the whole n-table.. Could i have done something else wrong in the config? or is this fixed in a recent firmware?
Toyota Supra -88 install
the 7M-GTE Engine in the car is a 3.0L turbocharged, separate coils for each cylinder (Maybe wasted spark, not sure yet...).
New info on the trigger wheel, Does this work right away? or is firmware changes needed?
All of these sensors a
- 1 VR sensor on a 24 teeth even spaced cam-wheel
- 1 VR for TDC Cylinder 1 signal (CAM index, OK)
- 1 VR for TDC Cyl. 1 + 180 camdegree (effectively TDC cyl4 CAM index )
The HOME signal is the secondary trigger.
My project car is a volvo 242 -75.. That will be equipped with a
B23 Turbo engine..
GenBoard/VerThree plans:
I will install the v3 GenBoard into a Volvo 142 with a B230 engine, currently controlled by Motronic.
- Stock IAT and CLT sensors will be used.
- Distributor with hall sensor, 1 trig per ign.event
- Hall sensor with 1 pin for cam position