K Series Immobiliser & Fuel Pump - Setting The Record Straight

[revilla]

I've seen it repeated time and time again on here and on other forums on the Internet that the K Series immobiliser cuts the fuel pump.

THIS IS NOT CORRECT

I've also seen several cases where this assumption has hampered diagnosis of a problem. It's just a myth that has been repeated so many times on the Internet that nobody thinks to question it. It's also an illusion created by the system behaviour.

I've done extensive testing both on the car and with the ECU and immobiliser on a test rig on the bench. Certainly for EU3, and on all EU2 ECUs I've looked at also, the following correctly described the way the system works:

• The ECU will run the fuel pump:
  • For a short period when the ignition key is first switched on to pressurise the fuel rail. This is repeated if the ignition is turned on again after a certain delay has elapsed. Whether immobilised or not.
  • When it receives a signal from the crank position sensor. Pretty much any signal, doesn't have to be a well-formed crank signal, anything switching between positive and negative voltage will trigger it, meaning that it will start the pump as soon as the engine starts to crank. If you take the crank sensor out with the ignition key on and just wave it around any old piece of magnetic metal, the fuel pump will turn on. Whether immobilised or not.
• The ECU runs the fuel injectors:
  • For a short grace period (around 2 seconds) on cranking or running after the ignition key is first turned on. This is repeated if the ignition is turned on again after a certain delay has elapsed. Whether immobilised or not.
  • When the engine is running and the correct immobiliser code has been received.
• The ECU runs the ignition coils:
  • Whenever it has a valid crank signal to give it ignition timing. Whether immobilised or not.

So the immobiliser (or lack of correct signal therefrom) causes the ECU to cut the injectors. Clearly when the injectors stop firing the engine stops. When the engine stops, the crank sensor signal stops. When the crank sensor signal stops, the fuel pump stops.

The fuel pump stops because the engine stops. The engine does not stop because the fuel pump has been cut.

• You can see this sequence on an oscilloscope. It all happens quite quickly so it's hard to see the sequence just watching the engine, but a scope resolves it easily. Injectors stop. Crank signal dies away as the engine slows down. Then pump stops.
• If you feed the ECU with a fake crank sensor signal when immobilised, the injectors shut down but both the ignition coils and the fuel pump continue to operate. Once you remove the crank signal, the ignition coils and fuel pump turn off.
• If you fit a fuel pressure gauge in the fuel rail feed, when the engine shuts down due to the immobiliser, the fuel rail is still fully pressurised. If the engine shut down due to the pump stopping, the fuel pressure would have dropped off killing the engine through fuel starvation.

The fact that the fuel pump primes the fuel rail, then the immobiliser typically allows the engine to run for a couple of seconds before shutting down, and the fact that fuel pump can no longer be heard when the engine has stopped gives the false impression that the engine used up the fuel pressure in the rail and died because the pump wasn't running. The two-second "grace period" appears to be to allow a Rover driver to jump into the car, stick the key in and start immediately; it gives time for the key fob to passively disarm the immobiliser and for the ECU and immobiliser to synchronise.

Rover's immobiliser system is quite robust. The immobiliser broadcasts an encoded key once mobilised and the ECU requires this number as authorisation to allow the engine to run. Each immobiliser has a different (programmable) key and the ECU is "trained" to know the key for the immobiliser to which it is matched. Removing the immobiliser will therefore not disable immobilisation, the ECU will not get the required key and will not run. Given this level of sophistication it seems odd to me that people would imagine that Rover would simply use the immobilisation to cut the feed to the fuel pump, which could then be bypassed with a single wire. Maybe this method was used on older, simpler engines but in the days of computerised engine management there are much more appropriate solutions available. By disabling the injection, it has disabled a complex synchronised time-based signal that cannot be simpy bypassed.

In summary:

THE IMMOBILISER CUTS THE FUEL INJECTORS - AND ONLY THE FUEL INJECTORS

THE OPERATION OF THE FUEL PUMP ITSELF IS INDEPENDENT OF THE IMMOBILISER

[revilla]

PS: If nobody believes me I'll post up some videos of tests on my car and my test rig.

[K7 VCT]

I believe u.

[Englishmaninwales]

Good detective work, Andrew!

[Jonathan Kay]

I assumed that everyone believed you and that there was no more to be said...

Jonathan

[Wrightpayne]

Andrew,

Excellent detective work. 

You may want to correct the typo - I think you meant to say ‘shut’ :-)

Ian

[revilla]

Hahaha. Oooops! Talking shut again. Thank you.

[StevehS3]

Excellent Andrew. Printed and placed in file. Long live the K!

[rj]

I would have been very surprised if the fuel pump was cut. Cutting the injectors is the only thing that makes sense.

[John Vine]

Impeccable, Andrew.  As always.

JV

[anthonym]

That heart chilling silence: turn the key and

no zzzzzzzzzzzzzzz

which I know only too well.

[StevehS3]

and that lovely pre-blat chatter of the IACV