First of all, you know way more about electrical theory than almost all other visitors to this site. That makes my job more fun. Second, it sounds like you know more about this model than most mechanics, so probably the best thing I can do is make a few observations that might have some value.
My background is Chrysler, but Mitsubishi and Chrysler do things very similarly. Mitsubishi makes a lot of parts for Chrysler, in fact, my favorite radio to repair is a Chrysler cd / cassette combo built by Mitsubishi.
As for the crankshaft position sensor, if you're measuring voltages on it, I know it will have three wires. One is the feed for it and that's usually 5.0 volts, but it could be 8.0 or 10.0 volts. That will be present anytime the ignition switch is in the "run" position, and leaving the ignition switch on won't harm it. It's on regardless if the engine is running or not; it doesn't care. Another wire is the ground, but it goes to ground through the Engine Computer so that circuit can be monitored. It will typically have 0.2 volts. I never tried to measure the signal voltage, but if you do, I suspect you will see it vary from near 0.5 volts to 4.5 volts if you slowly turn the engine by hand. Crank and cam sensors always have a magnet in them, and something with notches on the engine disturbs that magnetic field as they pass it. That induces a voltage which is amplified by the circuitry inside the sensor, and that's why they can have a varying output voltage while the engine is standing almost still but turning very slowly.
The sensors with two wires rely on the movement of whatever it is with the notches to disturb the magnetic field fast enough to generate a pulse of voltage in the coil of wire around the magnet. Speed is a factor in how much voltage is developed, just like in a generator, so there will be 0 volts coming from that type of sensor until the engine is turning at cranking speed. Since no voltage is applied to them, there is no difference between ignition switch off or ignition switch on / engine not running.
The next comment has to do with what Chrysler calls the automatic shutdown (ASD) relay. Mitsubishi might call it an "ignition" relay or "fuel pump" relay but I believe they do the same thing. The Engine Computer turns that relay on when it sees engine rotation, and it knows that by the pulses it receives from the cam and crank sensors. Once it turns that relay on, that sends current to the ignition coil, injectors, and in Chrysler's case, the oxygen sensor heaters, alternator field, and fuel pump or separate fuel pump relay. That's a safety design and is real effective and reliable. In the case of a crash that ruptures a fuel line, the engine can't run with no fuel pressure so it stalls. When it stalls there's no cam or crank sensor pulses so the computer turns off the ASD relay. That turns off the fuel pump so it doesn't continue dumping raw fuel onto the ground creating a fire hazard.
The clue to a problem with the sensors is there will be no voltage to the ignition coil or injectors if the ignition switch is on but the engine is not running. My reason for sharing this is everything in those circuits should be dead when the ignition switch was left on so nothing should be overheated. On much older cars it WAS possible to have current flowing through the ignition coil with just the ignition switch on but that rarely damaged them. I suspect in your case there may have been something that was planning on failing soon anyway and the least little extra nudge sent it over the edge. Leaving the ignition switch on wouldn't cause a part to fail but it might hurry one up that is about to.
Don't have an answer on your fuse issue. I usually check for voltage on both sides. You can't go by the resistance checks on the crank sensor because there's a lot of circuitry inside it that you'll be measuring through. That won't work so the 14 meg ohms isn't valid. You have to test for an output signal. If I were to do that, I'd use an oscilloscope that I use for tv repair, but it's just as correct to use a digital voltmeter, but there are some stipulations. First of all, they are designed to measure 60 hertz AC voltage. Crank sensors usually develop three or four sets of pulses spaced far apart during each crankshaft revolution. You'll get some pulses much higher than 60 hertz, then a relatively long pause before the next set of pulses arrives. The display on the meter will be bouncing around a lot. It may show some voltage but the actual value means nothing. Some meters read hertz directly without regard to the voltage level of the signal but even those don't work on crank sensors. They'll pick up a series of three or four pulses, then a long stretch of silence, then pulses, then silence. The meter will get dizzy trying to display that.
I'm not real familiar with the ignition power transistor. I know what it does but as far as diagnosing it, that was always a part of Chrysler's Engine Computer and rarely caused a problem. The only thing I could suggest that might apply is to check anything that can be replaced or unbolted to see that there is heat-sink compound there to help transfer heat away. If you understand Ohm's Law, you know that voltage times current equals power which means heat is generated. Those switching transistors are designed to be turned fully on or fully off, never partway in between. At fully off there is no current flow so 0 amps times any voltage equals 0 watts and no heat buildup. That one is easy to understand. But, when it's fully on, there can be lots of current flow but since it's practical a piece of wire with no resistance, no voltage is dropped across it. 0 volts times any current again equals 0 watts and no heat. The problem comes in when there's a little voltage and a little current at the same time. THEN you have power dissipated and heat generated, and you gotta get rid of that heat because that is the deadly enemy of semiconductor electronics. Circuitry problems can reduce how effectively that transistor turns fully on or off. The voltage spike from the ignition coil's primary winding has an effect too but I can't remember the reason. Other circuitry in the driver module generates heat too.
Anyway, there's my thoughts on the subject. Take anything you might find useful.
Friday, November 2nd, 2012 AT 5:03 AM