hi petrik i have flash the stock ecu, do you have any ideas on how to improve the cold start, it sucks when you try to start it on a cold day it cranks over and over before it fires....but when it's warm it works like a champ... 
motor is low compression 8:1 JE turbo pistons, carillo h-beam rods, 65lbs valve spring, stock cams 105/107, high volume oil pump gear, i don't run a power commander..and battery is always fully charge
B/W thank for sharing the info on the stock ecu i'm very impress with the stock ECU'S ability...
A good question. Spent today a good few hours in investigating the coolant temperature effect to the fuel mix. In general the effect of coolant seems to be very small compared to any other sensor. The coolant has much more effect to the ignition timing.
The fuel delivery is basically calculated with the following formula:
(air_temp_map+0x180) * water_temp_map* ambient_pressure_map * TPS_pulsewidth_map * RPM_pulsewidth_map * TPS_opening_change_map * (Fuel_map * injflow_parameter) + (opening_time_map*Voltage_compensation_map)That should help anyone doing ecu tuning in understanding how different parameters affect the fuel flow. Regarding cold starts a couple of parameters to be named:
injflow_parameter is the 1 / injector size. So if you have increased fuel pressure of have a regulator for turbo that adjusts according to the pressure it may deliver less fuel than expected when at vacuum. So changing the inector size bigger, i.e. reducing the inj_flow paramete A and B should increase the overall fuel flow.
If you have a turbo its possible that the intake air pressure calculated fuel delivery may not reflect the true need of your engine. Therefore the best bet would be tuning the fuel maps (IAP maps), particularly the neutral/cluch on maps for cranking area should be noted. My personal first bet would be to check out the idle maps, maybe even with an old yoshbox. That has helped us here locally to smoothen a few turbo bikes which otherwise have suffered from non smooth idling. For other than idle area I can not recommend yoshbox to be used.
Of course also water temp maps play role, but like said I have not been able to generate a significant difference to the fuel amount when testing with different coolant sensor values. It looks like the coolant sensor value is more affecting to the ignition than fuel.
As you propably know, when the ecu is turned on it squirts a good amount of fuel to the engine. You can e.g. turn kill switch on/off for several times to pump fuel into the engine just to test if its lack of the fuel that prevents starting.
Logging the data parameters during cold start should help in resolving what parameter to change. I dont know if you have tried the software I wrote, but that should be easy to use and log the engine data in this kind of situations. Its only two wires to hook into the ECU harness and you have full engine logging capability for trouble finding and tuning... I would start this by using it and comparing the data with results when actually running the engine to find what are the key differences.
Hope this helps...
