**
That sounds like an internal fault in the ECU, Jeffrey - can’t see any external cause for this. But the internal workings of the ECU is shrouded in mystery - it’s more to it than meets the eye.
Add a test; measure voltage to ground on each of the coil’s terminals with ign ‘on’ - voltages should be identical. Different voltage means current is going through the coil - which implies current is passing through the ign amp, which should only happen with the engine turning. A ‘leaking’ amp may mislead the ECU - but should not interfere with the bullet test…
An injector only fires when the injector’s ECU wire is grounded - which the ECU does every third ignition pulse. With separate wires from each injector, the fault is most likely inside the ECU - as Kirbert says…
Yep I think that’s where we’re at. REALLY appreciate your insight and assistance here. Its been very valuable. I have an ECU en route and will plug it in when it arrives. If that doesn’t work I’ll try another but after that I will need your advise as to which sanitarium I will need to check myself into…
An essential clue is the noid on an injector plug; lit with ign ‘on’ is wrong -g - it implies that the injectors do not click because they are always ‘on’ with bullet connected. No flickering or no light in ‘crank’ is also wrong; both again implies some fault in the ECU - even if I cannot recall that particular symptom/fault before…
Frank
xj6 85 Sov Europe (UK/NZ)
**
As an aside; the injectors are not meant to be continuously powered, and can lead to burnt out injector coils. It’s prudent to check resistances on all injectors; 2,4 ohms between pins, and ‘1’ from each pin to ground. And certainly on the injector used with the noid…
Kirbert
(Author of the Book, former owner of an '83 XJ-S H.E.)
24
How is that noid being connected? Between the two terminals on the injector plug, or between one terminal and ground?
**
Actually, the injector test light, noid, is specifically designed to fit into the injector plug, Kirbert. But any test lamp will do - and the noid, with jump wires, can be used anywhere as a test lamp…
Measuring between each injector terminal is just to verify terminal insulation - and an ohmmeter is then the proper tool. But both tests are relevant…
My point was that with ign ‘on’; there should be no ground in the ECU to light the lamp. So if the lamp is lit; it has a ground ‘somewhere’ which is wrong. And fi the connectro is connected to a n injector; the injector will click open - and stay open with ign ‘on’…
Capitol building @ Sacramento, CA… I guess they have a resident shrink…
Kirbert
(Author of the Book, former owner of an '83 XJ-S H.E.)
27
And presumably the injectors would need replacing, as they are all burnt up. Or perhaps the ground is through the resistor pack, so current is limited.
I’m at a loss. I popped in another ECU from a loner- nothing.
If you say that the bullet is supposed to only “click” every third time that’s where my problem is. But I don’t get where the fault would be. Something I’ve overlooked.
Must be something we have all overlooked, Jeffrey…
The weakness is that we know ‘what’ the ECU is supposed to do, but not ‘how’ it does it. The assumption is that the gets voltage from the coil ‘-’ connection - which varies between ground and 12V as the ign amplifier ‘makes’ and ‘breaks’ coil connection to fire the plugs. The ECU registers and fire the injectors once every engine revolution - every 3rd time the plugs fires…
Hence the idea that disconnecting connecting the bullet connector, with ign ‘on’, imitates this voltage variation. And should fire the injectors every 3rd connection, but it may be more to it than that. The injectors fire the injectors once as the ign is turned on - and does respond to bullet manipulation. However, the coil neg also grounds through the ign amp - which is not simulated with the bullet trick…
That said (as an excuse); does the replacement ECU respond at all to bullet dis-/connections? The weakness of a replacement; is it verified working before fitting - did it run in a previous car?
You need to replay tests, but first; verify that the injectors are in working order as described before. The current through the injectors are reduced through the resistor pack, but his is to protect the ECU in case an injector is shorted. It does not protect the injectors themselves which are not constructed to be powered continuously.
Ok Thanks Frank. I will do the OHMs today. Is it really possible that All injectors failed at once? Also- If I here an on off click when I connect and disconnect the bullet. Doesn’t that mean injectors are functional? Feels like my problem would be the pluse source for ALL injectors. I had replaced these injectors with great success months ago as driveability was substantially improved.
Doug- so far they “look good” very clean. I will do tests however after checking ohms and pins I think I need to go back to the beginning and retrace. Just need to feel like I adequately tested from the getgo. Wish there was a diagnosis test list! I could also use a wiring diagram of the fuel injection system but I can’t find a good source. I have a feeling Im going to have this thing towed to a shop unfortunately.
**
That the injectors click does imply that they are working, Jeffrey - but only if you are sure that they all click, not just one or two…
Clicking does mean that that the injectors activate electrically - but not necessarily that they deliver fuel. Which is neither here nor there - it’s the lack of clicking that is the present problem…
It is barely conceivable that both your ECUs have the same fault, and an unusual one at that - so replacing the ECU was a right move. The problem is that all relevant external tests checks out OK - and even with some of them out of spec; the ECU should still trigger injector action. Which internally is just to ground the injectors for some 50 ms - the duration varying slightly with external inputs…
Confirm; with the noid connected to an injector plug, ign ‘on’ - is the noid lit or unlit? And in ‘crank’ - lit, unlit or flickering? With power to the injector, from main relay/resistor pack - only the ECU can light the noid.
At the bullet connector, disconnect, and with ignition ‘on’; briefly connect the noid between each end in turn to ground. The coil end should light the noid - but what about the ECU end? Then connect the the noid between the two ends of the bullet, and crank - theoretically; the noid should flicker? It is just testing ECU logic - sort of…
As said before; apart from the trigger pulses from the coil ‘-’ the ECU varies the injector pulse width based on inputs from the CTS, the AFM and the cranking circuit (and the O2 sensor if fitted). There are no other inputs, and even if one or more input fails; experience seems to show that the injectors will still click, though with faulty pulse widths…
This is really the poser that utterly bewilders, and made/makes the ECU the prime suspect. What have we overlooked - it’s usually something plain and simple…
