pcmhacking.net

Hope this is posted in the right place.
I have an issue with my EVO. What happens is it starts of fine but then after a few laps the fuel mix goes lean 20:1. It happened today on my first run at the track. Poopered the day from there. The strange thing you can see in the log that the BPW and duty cycle both go to zero. Any thoughts?
An Excel log is attached and it shows up at line 2029 thru to 2104.

It could be a faulty ECM. I saw similar happen once, and thats what it turned out to be. Was loosing the outputs without loosing any inputs. Its sus that O2 voltage drops to 0, which makes it looks like a possible wiring fault, but the injector pulse width drops stops first and you do still have an rpm signal. It's be interesting to check for DTCs and check all inputs in a complete log, but thats my guess based on the csv.

o2 would drop to zero if the fuel is dying off though so that makes sense.

This car is doing my head in. There are no logged codes and yesterday when I got home all the live data looked as it should. This morning I tried to start it an no go it just spluttered so I took off the return line and heaps of flow so put on the pressure gauge. Like me it's too old so I bought a new gauge. It's a little high but acceptable and the car is running fine. Gave all the wiring a good tug along with the turbo hoses hoping to find an air leak and it won't falter. I know whats going to happen. Won't do a thing wrong till next race day. Can't drive it on the road but I do have a spare ECU so might keep trying and if it does it again try the other ECU.

Yeah try that and let us know. Its strange for the injectors to stop if the rpm signal is coming in, im sus of the ecm. Do you have an xdl from tunerpro or just logging to csv?

Just logging to csv. Had a look at live temps this morning and it showed both 15 deg for air and coolant so that's good. Tried to start it and was a real pig but once started it ran fine except for bouts of going lean and a few splutters so I hooked up a separate fuel supply to eliminate pump issues. I went to swap the ECU and found a bundle of scotch locks on the ECU wiring so it looks like it had an alarm or something most have been cut but there are about 4 that are still attached to something. Will post again tomorrow after another cold startup.

Not much evented today. After a few splutters on startup the car ran fine. I tried another ECU and looked at the live data and it's the same. I also removed 4 scotchlocks off the ECU wiring and worked out what the other 4 still attached are for. The boost controller so left them in place and reset the controller. Just thinking what I identified in the log was a decell fuel cut and not the problem I am chasing. Plan on pulling the plugs in the morning looking for coolant on the plugs because of the starting cold issue. There are no signs of coolant getting into the combustion chamber but I need to start a process of elimination.

If you cant reproduce it, and cant find any more clues take the spare ecm to the track and see if the scotch locks solves it and if not - if it does do it, swap the ecm out and see if it goes away then. Im fairly confident its not a sensor or power because the RPM doesnt go away. But if we were looking at the wrong part of the log, then we need to find the right part to confirm that. It looked right to me though, decel fuel cut isnt normally 100% unless your rolling down hill and it kicks in after a couple of seconds to provide engine braking.

Car has been starting and running fine so I decided to fix the sender unit and took out the pump. OMG the amount of rust inside over everything and the scale floating around the tank now confirms I was starving for fuel after a few laps. I am really thinking the log is displaying a 100% fuel cut on decel but I can't tell because the logging of the TPS dosn't work. I must have the wrong address to get that MUT data. Anyway after explaining the essentials to the wife here is a pic of the latest upgrade to fix the tank problem.

Nice