Squadron 58 - 2004 - Saloon Floor Removal

[jfm]

Well, in principle those electronic routine checks should be able to identify also this type of fault, because obviously that cylinder would have zero contribution at all times, so the rpm would be unaffected regardless of how much/less fuel the ECU tries to squeeze in that injector...

I don't think so - if an injector is flooding it is putting in more fuel than can be burned. Opening the injector for a nanosecond longer, as per LS, will not increase rpm. Hence, the ECU when doing a LS test cannot tell difference between a closed injector and a flooding injector. It could however tell the difference if it had a short term memory whereby with a blocked injector it would see the injectors open for 6/5ths (for a 6 cyl motor) as long at any given rpm as they were at last week. I doubt it does that, though.

 

[Bouba]

There should be a fault code of P0200 for injector circuit open, where P0201 is cylinder one and P0202 is cylinder two etc.
Of course the fault will also include an electronic fault to the injectors as well as a mechanical failure

 

[jfm]

So, I suppose that the check routine which we are discussing must consider separately not only rpm, but also the governor-driven load adjustments, to eventually identify the effects due to single injector adjustments.
That's just armchair software engineering anyway, and not so relevant once we agree on the principle.

I may be misunderstanding, but I don't see that. The LS method involves opening the injectors one by one in turn for say a nano second extra, and looking for extra rpm measure from the flywheel proximity sensor. The quantum of extra rpm is recorded for each cylinder, then the ECU compares them. If one cylinder shows nil extra rpm or 20 rpm extra, and all the other show +40 rpm then ECU knows the injector in that cylinder is faulty. This does not involve looking at any load data, because load is merely a derived result rather than a source of any truth. It involves overriding the governor and opening an injector for a nano second longer than the governor wants. Maybe that's what you meant, sorry!

 

[MapisM]

Nope, not exactly.
What I meant is that if it weren't for the governor constantly sensing the rpm and squeezing more or less fuel as needed (depending on the prop demand), the rpm would continue to change, at any given moment.
So, the "LS test" must somehow be able to distinguish between an rpm change induced by the test itself rather than by higher/lower prop demand.
Otoh, I suppose that depending on how fast this test can be, maybe any prop-induced rpm variation becomes irrelevant, I'm not sure.
No worries anyway, as you say we are just speculating, and the principle is clear anyhow.

Ref. a potential injector stuck close, I only meant that also in this case it should still be possible to understand that there is a fault, and in which injector.
I agree that understanding whether that's due to it being stuck open or close might be more difficult, though not impossible, following your own train of thoughts. :encouragement:

 

[MapisM]

Of course the fault will also include an electronic fault to the injectors as well as a mechanical failure

Well, logically they are completely different things, since from a logical viewpoint the first doesn't even require the "LS test" we are discussing, while the latter obviously does.
No idea about what VP ECU actually does, anyway.

 

[scubaman]

I'm trying to get my head around this, but isn't this related to the automatic compression test whereby fuel feed into each cylinder is turned off one at the time. How would the ECU be able to tell the difference between a faulty injector and poor compression?