There are 2 wires to the solenoid connector, a red which should be live with 12V all the time the ignition is on and a grey/green that the ECU should connect it to ground when the starter button is pressed. Easiest way to check would be with a test lamp. When the fault condition occurs, unplug the solenoid connector and connect a test lamp across it's 2 terminals. The lamp should light when you press the starter button. If it doesn't, the ECU isn't providing the ground connection to activate the solenoid. If it does the solenoid is faulty.
Surely it's not quite that simple. You yourself pointed out that the ECU monitors several parameters before it will enable the starter solenoid and even though the test points to the ecu side, it could well be one of these that is causing the problem, not necessarily the ecu itself. I guess the op could always acquire a s/h ecu for testing purposes and just sell it again for the same money if it turns out the ecu is not at fault. We still don't know if the electrics die when the starter is thumbed do we?
This is correct. The ECU will only allow the starter to operate if: The bike is in neutral The bike is in gear with the stand up and the clutch lever pulled in. Any other combination of neutral/stand/clutch switch then it won't allow the starter operation.
The fault symptoms are that the starter motor does not engage when the engine is hot. This is, in my experience, indicative of an electronic fault with the ECU. The inputs, Clutch, Side Stand and Neutral lines are all just simple switches. These either work or they don't and are exceedingly unlikely to have a heat related fault. As a process of elimination it would be a good exercise to do resistance checks to see if they are working OK. All of these inputs go to the ECU, and even if the switches are good there still could be an issue with the circuit which "sees" if the switch is closed, thus preventing the starter from being engaged. I put my penny on a duff ECU.
It would need two switch faults before the starter wouldn't work with the bike in neutral and the neutral light probably wouldn't work either. The ECUs are very robust but I have heard of one of them in which the starter function failed. If it does turn out to be the ECU he can either replace the ECU or, by joining the grey/black wire of the Body connector (Grey) pin1 to the grey wire of Body connector pin 28 restore starter operation by operating directly from the starter button but without the benefits of the stand/clutch safety interlocks.
I might do the wiring frig if I was desperate, but not as a long term fix. Best course of action is to try and prove the ECU either by testing or, as you rightly suggest, substitution.
