It helps to understand the sequence of events that occur for starting. One big thing to note is that the ignition system and fuel injection system are very separate with the one caveat that the EFI system is dependent on some signals including the ignition system in order to run.
When the key turns on, the EFI Switching Relay turns on (if the inertia switch is ok). (A relay not present on some other years). This in turns switches on the Main Relay (in the trunk) providing power to the EFI ECU and the Injectors.
When the ECU wakes up from key on, the first thing it does is flip the fuel pump relay on (grounds it) to pressurize the fuel rail. (You should be able to hear this with the trunk open and key first-on for about 5 seconds). Assuming there’s no leak-down issue (failed pump check valve, or problem outgoing fuel pressure regulator) the fuel rail is up to PSI and ready for injection.
The ECU will not power the fuel pump back on until it gets a +12V signal from the starter relay (indicating cranking) and then for sustained operation, the RPM signal from the Lucas AB14 Amp on the white shielded wire going back to pin 18 of the EFI ECU (one of the 3 way splits off of the Amp’s control of the negative side of the coil). This can be a common failure point.
With RPM signal, it will begin looking at the sensors and performing injection. Without the RPM signal, it will do nothing. The injection system is a peak and hold type, meaning that the injectors are OPENED by a high current pathway in the ECU, and then within milliseconds switched through the power-resistor pack on another pathway to sustain holding them open. (Sometimes the connector there needs a little cleaning). The amount of injection is LARGELY determined by the RPM and Coolant Temperature Sensor (2 prong unit) behind the B bank thermostat housing). A problem connection here will often result in non-starting, poor starting, or massive massive over-fueling to the point of non-starting. Most of the other sensors are simply trim inputs and unless the signal is plain out wrong from them, they can be unplugged and a car will usually still start. (Including the IAT, the TPS, O2 sensors, Load Switches, and even the MAP sensor vac tube (to a limited extent on that last one).
The injectors are wired in 4 sets of 3, but actual firing is done bank to bank as the internal circuits in the ECU are fused into an A bank and a B bank control system so that each side can be independently trimmed by O2 input (when they warm up).
IF the ECU turns the fuel pump on with key-on, that circuit is functioning and isn’t a problem during starting. It’s also worth noting that the control side of the Fuel Pump Relay gets its power from the Main Relay.
One way to check the ECU’s ability to fire the injectors is to turn the key on, and listen for them to all click ONE time with full depression of the gas pedal (what’s really happening is the turnstyle hits the full load switch and, assuming it is working, it will trigger a single burst of the injectors).
The ECU controlls the injector’s negative side. The injectors all get power from the Main Relay on their pink/black wires, but will show +12V power on the other side, unless at least all the 3 injectors on that control circuit, and often the entire 6 on that bank are unplugged because the power will just flow through the pathway back over to the control side of the injector (until it is grounded to operate).
The most similar problem I’ve seen where a car like this starts but doesn’t was a guy w/ a problem connection at the CTS connector, resulting in the car starting IF he pulled up the inertia switch (cutting off the fuel system) but NOT when it was depressed. The reason was the car could start once the fuel had leaned out (with injection off and fuel vaporizing in the manifolds) but NOT with excessive fueling from the ECU because it had a bad input from the CTS.
The ignition system:
The AB14 Lucas Amp box is really nothing but an enhanced GM HEI 4 pin module. The input trigger side is given by a 2 wire VR sensor in the distributer that creates a rising AC voltage of 12 voltage spikes per rotation of the distributer shaft (because it has a 12 point metal star wheel attached).
On the other side of the amp module inside the box (the other 2 pins), one side gets a +12V input with key-on (as does the positive side of the coil) along with a points-style suppression condenser / cap hanging off of it (can cause failure if it internally leaks to ground), and the negative side is split 3 ways —well 4. One is direct to the coil to cut charging and trigger the spark, and the other two are through a white silicone blob with resistors inside. One leads to the Tachometer, and the other connects to the EFI ECU on pin 18 through that white shielded wire (RPM signal source.) The 4th connection is to a zener diode bonded to the metal case for voltage spike suppression.
As noted that RPM signal wire to pin 18 of the EFI ECU can be problematic… usually in the engine bay.
If it ends up you need to go through the process of detailed continuity testing, I’ll post some diagrams to help with that, but hopefully this will sort out as something simple.
This is mostly right, but the connector is upside down here. The long row would really be on the bottom, so you kind of have to flip everything but the incoming connector mentally:
Sometimes this helps if it becomes necessary to make a wire-loop in the engine bay (like between the pins of the CTS) and check that continuity on the proper pins at the ECU. PITA work you want to avoid if you can solve it any other way.
~Paul K.