I have produced a completely new issue and couldn’t find anything in the forums.
After pulling my injection wiring (at the fuel rail) and resoldering a brittle wire, the engine failed to crank more than a few degrees.
Sure enough, cylinder 4 was completely full with fuel. 5&6 are quite wet as well.
So besides hoping that nothing broke (the starter simply stalled, not much energy involved and with plugs removed everything turns nicely):
Is there any factor of three involved?
Should I check for damage? Of course an oil change before running it; there is a lot of fuel in the pan. Bores are already re-oiled of course.
What could have caused this? The battery (I recall, fwiw) was disconnected and all connections are fine; all injectors click, the pressure regulator is ok by all means… I changed the power resistor against a spare but didn’t have time to crank this morning. Grounds are all good.
Did anyone have that issue before and can someone point me to a possible factor of three?
Hoping for any guess. All suggestions most welcome
Thank you,
David
I wish I could remember better but something dimly tells me the injectors only fire in two groups of three each. They do not fire individually in time with the valves. If that’s the case, and if one of the groups of 3 is 4,5 and 6 then maybe there’s something wrong with the system sending power to them when it shouldn’t.
But couldn’t it just be that those 3 are leaky? I would try taking them out on their hoses and sitting each one in a jar of some sort, leave them a day or so and see if there’s fuel in any of the jars.
**
Open injectors are scarce as hens’ teeth, David - whether leaking under power or mechanically stuck open…
The injector are permanently powered from the main relay via the resistor pack - the ECU ground them to open. Ex factory, two groups of resistors run 1-5-6 and 2-3-4 respectively - but your, or POs may have crossed wires. but injector grouping is immaterial to engine operation.
However; one step is to disconnect injector plug and run the pump while observe delivery in the cylinders (no mean feat!) - this will solve whether the problem is mechanical or (most likely) electric. Nominally, injector leaking is tested by fuel pressure leakdown, but there are pitfalls…
Electrically; use a test lamp or noid connected to each plug in turn - with ign ‘on’; there should be no lamp light. Any plug with a lit lamp is grounded but finding that incorrect ground may take some doing. Initial focii on the point where you did the soldering - and the replaced resistor pack…?
That the injectors clicked regularly (all of them?) complicates matters - a permanently grounded injector will not ‘click’. Another factor, though somewhat unlikely, is that the ignition failed and fuel accumulated - cranking engine to verify spark quality should be tried, with fuel relay removed…
But much depends on the injector plug testing - and some verification as required for mechanical injector leaks…
The only permanent damage is that permanently grounded injector coils may have burnt out - they are not really meant for continuous power…
Okay! (but why 4,5,6 then?) - the injectors are fired at every second ignition pulse, so 1,2,3…
Glad that there’s some factor of three involved; we were right, David!
However; one step is to disconnect injector plug and run the pump while observe delivery in the cylinders (no mean feat!) - this will solve whether the problem is mechanical or (most likely) electric. Nominally, injector leaking is tested by fuel pressure leakdown, but there are pitfalls…
It is not mechanical. Fuel pressure remains stable, regulator is fine. Frank, I have jumped the pump relay so the pump runs with ignition because someone has miswired the relay somehow. Next thing I did was change the resistor pack so troubleshooting results might remain unclear; but I haven’t cranked yet. What I did was remove all the connectors and test each injector at a time, with the replacement pack. all clicked.
As every injector has ist own pair of wires I guess it’s not at fault - but maybe one wire is grounded and the three are connected somewhere. just as Dick suggested. I will investigate.
That the injectors clicked regularly (all of them?) complicates matters - a permanently grounded injector will not ‘click’. Another factor, though somewhat unlikely, is that the ignition failed and fuel accumulated - cranking engine to verify spark quality should be tried, with fuel relay removed…
See above, but it doesn’t flood normally and even if I left the pump running (which I don’t) it wouldn’t flood. Maybe It has to do with either cranking,m or it was the resistor pack that suddenly failed, but how… the wiring loom being suspect has all the connecti0ons correct, I checked thoroughly and rewrapped everything to be sure. Did find one blank patch I overlooked!
But much depends on the injector plug testing - and some verification as required for mechanical injector leaks…
If it is not rectified tonight, the whole contraption is coming out and will be tested separately. With the jars, and Then I can check the flow rates also.
The only permanent damage is that permanently grounded injector coils may have burnt out - they are not really meant for continuous power…
That, and the hydrolock. But it seems fine, fingers crossed.
She runs. Bit rough as of now but this will hopefully clear up when the oil and fuel residue in the intake has burned off and the plugs get clean. Then rough as usual, hehe.
I drained a litre of petrol from the intake and suspect the amplifier because everything else seems reasonably sound. Will have a look inside.
This, or the intake flooded for some reason (cold start injector maybe) and fuel got sucked into the intake.
Thank you all for your help and see you next time, I’m sure!
**
Which of course may relate to faulty injector(s), David…
**
It’s a bit iffy; the injectors are always powered via the resistor pack, but clicks only when grounded on the other, ECU wire. Which is grounded by the ECU with triggering signals from a running engine. However, with ign ‘off’ there might be some ground path within the ECU. But the tests imply that the injectors are unharmed - as does a running engine…
Incidentally; to eliminate the CSI you just clamp the hose to it - just disconnecting the plug is inconclusive…
If everything now is ‘normal’ - you may have just experienced the actions of a gremlin of very unusual breed…
But how would they fail linke this, Frank, mechanically… also, fuel pressure remains somehow stable, which also - for now - rules out the csi: it is closed and tight for now, keep in mind the flooding occurred rather violently and is not related to a small leak.
If there is a false ground en route to the resistor pack and the resistor pack connects the three, we would have 12V from the ecu and earth from the resistor pack: Open.
Maybe a strange gremlin after all, who knows. Hopefully it’ll resort to smooth running again when I go for a drive later today; and from now on I‘ll give it some miles lest it fails again. Imagine all injectors suddenly opening on the highway… Shudder.
I‘ll keep you posted if something blows up again.
Thank you for now,
The false ground must be on the connections between the injector and the ECU to trigger the injector(s), David…
The ECU is not designed to ‘deliver’ power to the injectors, only ground. While 12V is indeed applied to the injectors between ground triggering signals, it’s a voltage only without any current capacity…
The ‘regulated’ fuel pressure will vary with manifold pressure - which is why the pressure test is carried out with only the pump running (or the regulator’s vacuum hos removed or clamped). A leakdown test will reveal seeping/leaking injectors - but with an instant drop; the non-return valve must be eliminated first…
Without any clear suspects, you just have to drive - and hope that if there a second event, the cause may be revealed…
Didn’t consider that. If the ECU is capable of providing the amperage it probably won’t be for long — no, wait, electricity flows through there as well so it should be fine. Or is my thinking flawed? Grounding the injectors lets the power flow from the resistor pack via resistors and through the ecu to ground. Provided the ecu has a mosfet or other switching capability there will be some room for flow. The leak to ground might have been anywhere and still no clue why, but hey, it works.
Forget about the mechanics of the fuel rail, everything there is good.
Edit//
Opened the resistor pack. Very boring. It is not more than a loop of printed circuit to drop the voltage, and that’s it. A lot of aluminum, extremely overbuilt for what it is!
It is designed for two units of four injectors each. They are interconnected. You can tell I don’t want the ECU to play a part.
Best guess is that there was a short to ground, or that there is an intermittent case, and this connected via the resistor pack.
You were extremely helpful Frank, as usual. Tonight the oil is draining (has a very thin about 0W consistency for a 20W50) out and the car will be run tomorrow. Another check whether all is well and then she’s back on the road.
Whoever made that oil filter and everything else in the vicinity so hard to reach on RHD.
**
Reputedly, not even the constructor of the ECU fully understood how it worked, David…
However, if you connect a test lamp to the resistor pack side of the injector plug - it will be fully lit. If you try it on the ECU wire it will not…
To open the injectors the ECU wire must be grounded - either by the ECU or a short. A ground short at the resistor pack side will simply burn the wires; the main relay, feeding the resistor pack, is connected ‘directly’ to battery power - there is no fuse…
A noid over the injector plug will reveal an ECU side leak by being lit/glowing with ign on. It should be dark until the engine starts turning, then flicker…
Ohming the ECU wire from the plug to ground will give some reading - caused by electronics in the ECU, but not ground…
With a gremlin at work, you can only change the oil to get rid of contamination - and hope the gremlin is busy with somebody else…
No one knows how the ECU works.
Right, the resistor pack would have shorted…
The gremlin must be on vacation; it started and ran like a dream today!
He can have the whole year off, and paid retirement.
For now,
David
