Hi All
I've read all about the hot starting issues but this seems a bit different. Engine starts fine when cold but after a good run and 10mins rest it wont fire up. I have hooked up a cable form the pump connector and run it forward into the cabin with a 12v bulb on the other end so I can see when the pump is running. I have now figured that the fuel pump is NOT coming on when the engine compartment is hot, not for its usual couple of seconds prime nor during cranking. If I open the bonnet once it's cooled down enough the pump comes on as normal to prime for a couple of seconds, the engine starts like normal and the pump then runs continuously as normal. Any ideas what could be preventing the fuel pump from operating when the engine is hot?
Any ideas much appreciated.
Joss
1991 Lynx Eventer XJR-S 6.0l

 

Hi, I assume the car is running the Zytek ecu as per other XJR S models of the same age? Either way apart from bad wiring it should only be fuel pump relay related, so I would try a replacement relay first then check that the ECU is sending the ground signal ( pin 24 on the Zytek) to energise the relay and thus the pump. Not sure though where the fuel pump relay is on an eventer ! Good luck.

 

Fuel pump relay is located in the trunk area RHS forward of the battery. The relay is fine. When cold the pump runs fine and engine starts. When the engine bay is hot the pump does not come on at startup. Something is getting hot in the engine bay and I guess preventing the ECU from energising the relay.

 

This should not be too hard to find, subject to any Zytec idiosyncrasies. The ignition switch on an OEM Lucas system activates the fuel pump relay via the inertia cutout switch on the A pillar. I would start (once the condition shows itself at hot) by checking if current is getting to the inertia switch, and also to the coil positive, ignition on.
If it is, then go to the fuel relay, and if this has NOT got current, ignition on, test the loom between the intertia switch and the relay. If it has got power (and once again assuming the Zytec is like the Lucas in this regard) try running the orange wire from the fuel relay to earth. This should trigger the relay/pump to run whenever the ignition is on. On the OEM Lucas ECU this orange wire runs to a timer in the ECU and if the ECU does not receive a pulse that the engine is turning over, it cuts the relay earth after about 3 seconds.
This would enable you to find out which bit of the system is failing when hot, it may well be the ECU timer system is the problem as the ECU heats up, for example.

 

Thanks for the steer Greg. Where are you in France? I'm near Grasse. While hot and not starting there is power at the inertia switch and coil +ve with ignition on. While checking coil voltage with ignition on I noticed that the fuel pump came on to prime after leaving ignition on for a while. I repeated the test several times and it comes on after about 16 seconds and then the engine starts no problem. I have tried to jump the orange wire on the fuel relay to earth and pump runs whenever ignition is on but still the engine would not start. I will take the car for another run tomorrow to check again.

 

This wiring diagram may help a bit

 

Good on the tests, you wrote "While checking coil voltage with ignition on I noticed that the fuel pump came on to prime after leaving ignition on for a while. I repeated the test several times and it comes on after about 16 seconds and then the engine starts no problem."
Am I right in believeing that this means if you leave the ignition on for 16 seconds the engine will always start, hot or cold? If so, this looks like either an ECU timer-type fault or other mystery ECU fault, or a problem in the wire from the engine bay to the ECU that tells it what the engine is doing.
To be clear, again, if you jump the pump to get it running, the engine will still not start inside the 16 seconds, or not at all, even after 16?

In principle, if the pump is running and the engine does not fire, then either:

• Injectors are not pulsing, and/or
• the ECU is not getting the signal that the engine is turning, or
• the coil is not firing.

Next question, very important to answer before any more bright ideas:
Which ignition system do you have? is it the specialised Zytec, or Lucas (black box on rear top of B bank inlet manifold) or Marelli (two small flat amp units on radiator closing panel, plus position sensors on flywheel and crankshaft pulley)?

I am in Burgundy, by the by, do drop in if ver you are passing on the way north.

 

If the engine is cold it will always start first time. When hot if I turn ignition on and I can hear that fuel pump hasn't turned on to prime I now know to keep ignition on and after 16secs pump will start it's prime and if I turn key engine will start.
Not sure what ignition system this one is - here is a pic of the engine bay;

 

Hi, nice looking engine bay! Its looks all standard Zytek and as you will see if you study the wiring diagram I posted everything , ignition included , is controlled by the Zytek Ecu.
The fact that when hot it will start after a set time would say to me that its not a rogue wiring issue as that would be more random in its timings, although its worth tracing through the wiring and earth points. As Greg said, when hot , but before the 16 seconds are up, I would test for injector pulse and spark from coil. If these are not functioning then logic would say its ecu related.
I think there is a Jaguar Sport group that would have more knowledge of the Zytek idiosyncrasies, might be worth a search, good luck.

 

OK, we are getting somewhere. You have a Lucas ignition system, or at the very least a Lucas amplifier triggering a Zytec ECU!

This is the Lucas amplifier:


The next thing to do is to make absolutely sure what ECU you have. In a normal coupé the ECU is tucked up,inside the RHS buttress on a bracket:

The standard Lucas ECU is fixed just inside the black RHS strengthener bracket, and can partly be see here marked up.
Somewhere hidden in the Eventer will be this unit, or some sort of other one. If it is a Lucas unit it will have have the code 16 CU on it. If not post a pic of what you have.
In principle the Lucas system is far easier to diagnose and is better understood and less complex than later Marelli or Zytec ones, so this is good news.
Now, assuming you DO have an all-Lucas system, the question is, what is causing the delay in the ECU starting the pump? My honest best guess, but is is a guess, is an ECU fault, because as far as I know, the pump starting on the ignition being turned on is a circuit completely independent of anything else in the system. Once we know the ECU you have, it is an easy swap to install another and see if that solves the trouble. Anyway, progress is being made!
Granr Francis (turn and bow to the Southern Cross at the name of the Great XJS Prophet) will have expert views to contribute in due course once the Franco-Oz hotline warms up!

 

By the way, nothing in my above post is in any way contradicting SCD's post about there possibly or even certainly, being a Zytec ECU in your car. For all I know the Zytec used the standard Lucas amplifier and dizzy!

 

Hi Greg, you are right, the pre facelift Zytek cars used the Lucas amp and a Lucas distributor body with the sequential ignition bits cased inside so they look like standard Lucas systems. The zytek ecu is located in the passenger footwell under the heel board plate a la Marelli ignition ecu, so should be easy enough to find out what system the car is running.
Regards Simon

 

Joss
As SCD has said above, to find the ECU in your car: If you pull up the carpet under the passenger toe area you will find a hi-tec plywood plate! Pull this away and there will be the ECU. I suggest that you first unplug it and clean up all the pins and sockets on the plug/ECU with aerosol contact cleaner and a soft toothbrush. You may even find a bit of condensation in the area as the screen can leak down there too.
Also, any connectors in the engine bay connecting to the distributor and amplifier would be worth carefully separating and cleaning too.
This sort of stuff:
https://www.amazon.fr/Sonax-%C3%A9le...%2C141&sr=8-16

 

Greg,

Joss's car is a 1990, not a 1991-manufactured car, and is a well known car. It was built as a standard pre-facelift XJR-S with Zytec and not converted by Lynx until 1993. Although it's had a lot of work done over the years, including a full restoration by KWE in 2016, I'm not aware that the Zytec was changed.

As we know, there are not many people around who understand the Zytec system, but there are still a few specialists. There's a guy who did a lot of investigative work and identified that the XJR-S Zytek ecu is an Hitachi Motorola Clone CPU - HD6303X.

Hope you all get the problem sorted.

Cheers

Paul

 

Joss, quick thought, you haven’t by chance got an immobiliser fitted to the car? Again good luck with the fault finding, regards Simon

 

Just a note here (and if you want as deep a dive into the Zytek ECU's guts as I think has ever been on a forum, you need to look at our thread in the Jag-Lovers XJ220 group forum... very very long. As comprehensive as we understand to date.)

And to the note: Unlike the Lucas EFI system where the EFI triggering is DOWNSTREAM of the Ignition, the Zytek system is upstream of all of it. I.e. the ignition (and injector control) is dependent on the dual hall effect pickup in the distributer to sense both engine position and RPM to both output sequential control over the injectors AND control over the Lucas AB-14 amp (really just a GM HEI pickup - though possibly a module using a square wave trigger).....thus providing control over the coil. In other words there is no backwards amp signal back to the EFI ECU to tell it the engine is spinning over and thus to turn on the fuel pump. So Greg's question remains. Have you checked for ignition output during hot cranking? (there is a small chance that the amp module is failing under heat soak but I wouldn't suspect that immediately.) Without the Zytek computer interface box and old DOS computer, the next best thing to have is a cheap oscilloscope to check that it is getting, and or receiving signals. The HALL sensors in the distributer have been known to fail, but I've never heard of a heat soak failure that regularly corrects itself when the engine cools. Dry solder joints or corroded contacts? Yes. Heat soaked amp? Yes. ECU fault? Yes. And I'm sure I've missed something. Also, though you have checked that the FPRelay works, have you checked that the MAIN relay works as it is ignition triggered, and passes power to the EFI ECU, Coil, Injector power, AND FPRelay control-side power. The simple thing to check would be to see if the coil instantly gets power with key-on (hot), or if there delay or any reason to suspect the main relay. It's an easy check at least all-be-it unlikely.


~Paul K.