The tester is very simple because the ECU is very simple. Most input signals are switched to ground, most output signals are pulled to ground. Most of these signals are tied to LED's and resistors to +12V to monitor them. The analog signals are referenced to the regulated 5V signal put out by the ECU. The crank angle sensor (CAS) is a magnetic reluctance sensor, like a magneto, which is hard to simulate, so that is the only actual engine part that I used with the simulator. A hand drill can be used to drive it; I put a tachometer on the leading ignition signal to monitor "engine speed".

Two slightly tricky items: first, the IgF (ignition feedback) signal must be supplied to the ECU or it will throw a code 12 (series 4 ECU). The IgF is normally generated by the trailing coil pack, however, I was not interested in having running coils generating 50 KV in my lab! So I built an IgF simulator using an LM555 monostable pulse generator, triggered by the trailing coil select signal (falling edge, I think). I based this on the information in this great article at TeamFC3S. The other tricky item is that the voltage levels of the ignition signals is 0V to 5V, like digital TTL levels. It is the logic in the coil packs that convert the ignition signals to higher voltage & current levels. I recommend using an oscilliscope to watch these signals. I put NPN buffers on these to be able to drive LEDs. At very low RPMs (like 5 - 60), you can see individual firing and fuel injector events driven by the ECU.

I have taken a hiatus from the project due to my short attention span and other projects. I hope to pick it up again some day: my goal has been to modify the ECU firmware to enable the UART built into the ECU to communicate with the outside world. From this point, any internal state can be monitored or modified! Unforntunately, it appears that this will involve modifying the ECU to move signals around.