1987 XJ3 Start Injector

Jag didn’t want to start. Checked the start Injector and sure enough it was not injecting anything. Thus I sent it off to get cleaned. But now wondering what kind of gasket or seal was used under the injector? Where I sent the injector, they had no info regarding any gasket or seal. Thought perhaps they would have that item. Now wondering what goes between. Surely something does. Where can they be obtained?
Lee Noga

Paper gasket. Make your own. Doubt they get the injector clean but they will have good equipment

Well they said they did not have a way to test this one out but stated that it sprayed well. Don’t know what that means. But if spraying that’s better than what it was doing before. Thanks for the info.They are sending it back today. Anxious to see how this affects starting. I’m sure without the start injector the auto is very difficult to start right?

There’s a trick in a pinch. It will start but not as easy in the cold. If it won’t spray it’s the switch

Certainly so in cold weather, yes.

In mild temperatures…60-70ºF or higher…probably not so much. Years ago I inadvertently discovered my cold start injector wasn’t working and I had no idea how long it had been broken. That was in spring.


Did you test the electric inputs to the CSI injector, Lee - to eliminate a TT fault, or other fault?

There is also a pitfall here; once the injector has been activated it needs some time (minutes?) to reset for repeat operation. And of course; if coolant temp is above 35C, the injector won’t spray…

It’s a bit suspicious that the engine didn’t start - even without the aid of the CSI the engine would eventually start on fuel delivered by the other injectors…

xj6 85 Sov Europe (UK/NZ)

On the V12 XJ-S, the cold start injectors were deleted in the early 1980’s. There was a rumor that the dealers would start the car on the showroom floor several times a day just to show people the quiet idle, and eventually it would get flooded because they didn’t let it run long enough to clear out all that starting fuel.

Of course, I always wondered why you’d need cold start injectors when you could just program the ECU to fire the main injectors to accomplish the same thing, and wondered if that’s what they did hence the deletion of the cold start injectors.

The principal reasons for the CSI is to saturate the manifold with petrol during cold starts, Kirbert - to compensate for initial petrol loss due to condensation in cold manifolds. And also to partly compensate for imperfect combustion in an initial cold engine. This is sort of difficult to quantify, and would complicate ECU programming - which is calibrated for running the engine during warm-up and hot.

I agree it is a sort of botch, but common on all EFI engines until more sophisticated computer programs, or direct injection, arrived. CSI does indeed shorten cranking time - and omitting it on later V12s may be based on experience? An eventual delayed start considered acceptable - or the characteristics of the V12 made it less recalcitrant?

That a cold engine requires rich mixture is a fact of life for reasons stated, and the CSI solved one initial cold start problem - varying injected amount with temp. And omitting the function with the coolant above 35C. That failed CSI prolongs cranking on a cold xk is well recognized (though it will usually eventually start) - whether delay, with or without CSI is the case, more or less, on the V12, is best answered by their drivers…:slight_smile:

xj6 85 Sov Europe (UK/NZ)

But the ECU is varying injected amount by temp anyway. The existence of a CSI implies either that it doesn’t vary the injected amount enough, or that starting requires an additional amount of excess fuel over and above what cold running requires. Coating the inside of the intake manifold would make sense on a carb car, but on an EFI engine the injectors are right at the intake valve anyway, no fuel should be lost to condensation on the insides of the intake manifold.

Atomization maybe?

Davids point is well taken, Kirbert - the fuel is delivered in droplets, carbs or EFI, and burn from the outside inwards. Which takes time and requires extra fuel - but as engine heats up the fuel start vaporizing and burn more efficiently. All of which is tackled by the EFI mapping…

But even with injectors there will be a brief moment of condensation; all injectors spray simultaneously but only one, if any, inlet valve is open for direct fuel delivery. The injectors are not synchronized with valve motion - one injection per revolution, lasting typically some 50 milliseconds, must serve all cylinders. Which involves some transport and delay for condensation - and it is not confined to the manifold, but also the inside of the cylinders…

Main point is that this involves a very brief instance during cranking, slow engine rotation, where extra-extra fuel is needed to counter the effect of cold engine parts. This can of course be included in the ECU programming, but being very transitory, and the ECU only gets information of coolant temp, it just complicates ECU construction. And also requires sensing of repeated, non-start, cold cranking to avoid flooding…

The CSI takes care of all this; a brief burst of extra fuel at initial cranking, and no delivery on repeats - simplifying ECU construction. The same idea of simplification is behind the use of constant fuel pressure; it relieves the ECU from taking fuel pressure into account - using only injector opening duration for delivering the correct amount of fuel relating to engine temp and air volume…

The problem facing ECU construction was that only analogue technique was available at the time - and even with simplification; development hit lots of snags to be ironed out. With digital computing vast computing capacity became available - and the part digital inclusion in later V12 ECUs may be a reason for omitting the CSIs…?

Not that I disagree with you on CSI, but ‘our’ set-up was state of the art at the time…

To wit; the engine’s fuel (and ignition) requirements are unaltered by whatever fuel/ignition system is used - unless the engine gets what it wants; it will sulk…

xj6 85 Sov Europe (UK/NZ)

1 Like