Citroen XM Forum

I've started on a new project car, not an Xm this time.

A sneak piccie of the rats nest that awaits. If anyone is interested I'll put more detail up.

Now, that looks like an Italian Spaghetti bowl. What sort of car is this?

It's a Jaaag XJS with a complicated history.

The issue is it had an engine swap
The car is designed for the later Magneti Marelli ignition system, the replacement engine is designed for the earlier Lucas constant energy system.
The options are to use the Lucas system and rewire the car and fit an associated Efi ecu, or, as is my plan, make the engine work with the more advanced, Marelli ignition system and the rest of the car stay as built.

The second image above shows the crank pulley for the Mareli system, with one of three timing trigger fingers shown.
I have fabricated a bracket to mount a crank sensor and drilled the engine block to be able to fit a flywheel sensor.
The sensors have to fit very close to the trigger fingers and flywheel ring gear (0.48-1mm clearance).

That is a nice car as well, but please stick to mainly XM

An interesting project, I have quite a lot of history with Jaguar V12s and have rebuilt a couple of engines, but I never had one with Marelli ignition.

What is the compression ratio of the engine? Designated by the 2nd to last letter of the engine number. Marelli engines were all 'S' = 11.5:1, but most UK spec Lucas HE engines were 'H' = 12.5:1 and require a different ignition curve and fuelling.

I have all of the components of a Lucas ignition system and a matching ECU.

No idea of the compression ratio, but I should be able to find the engine numbers as paperwork for both cars should be available. Where is the engine number located on the block?
I note the Marelli loom is missing the "strategy link", which apparently sets the fuel octane rating between 91Ron, or 95Ron, so presumably the timing is less advanced.

The distributor and associated ignition components for the Lucas system are in the boot of the car and I understand there is also the Efi Ecu, but only the wiring loom fitted directly to the engine was saved. The swap was made 7-8 years ago, but the engine has never been able to run since.

My understanding is the Efi Ecu are different between Lucas and Marelli as the earlier one uses an analogue sine wave from the distributor as the speed trigger, the later one receives a digital pulse from the ignition Ecu, but I'm not sure that matters given afaik, both use the same 6CU Efi Ecu and the wiring diagram shows both systems to be wired the same, not withstanding the Lucas system uses a shielded co-ax cable as the speed pulse wire to Efi pin 18. I've just read a link https://www.jag-lovers.org/xk-lovers/li ... umbers.htm
which has a note
"Note that for V12s 1971 through 1979, LA and LB suffix was 7.8:1 CR and SA and SB suffix was 9:1 CR. For 1980, suffix HA was 10:1 CR. From 1981 ?? suffix was 11.5:1 CR and ?? suffix was 12.5:1 CR."

The car is My 1989 and iirc, the engine was from a 1985 car.

The engine number is stamped on top of the block just in front of the bell housing, behind the oil pressure sensors.

A 1985 UK-market engine is sure to be 12.5:1, they are marginal for pre-ignition running with any fuel less than 98 RON. Be aware that they can suffer from 'silent' pre-ignition which erodes the cylinder heads without any audible pinking. I have run one (a 1983 XJ-S) on 95 RON with original Lucas ignition, but it was not driven hard, and I probably retarded the ignition by a couple of degrees.

The ECU model was updated from 6CU to 16CU before 1989, but the ECUs are specific to emission control version (with or without O2 sensors), and compression ratio. I believe that 6CU and 16CU are interchangeable if engine specification is observed. I had a NOS Lucas injection harness (the one that runs the length of the car), but sold it last year.

I think that maybe the Marelli ignition gives a different number of pulses per rev than the Lucas system, but I don't remember now.

Kirby's book is the best source of information, I contributed a few things to it back in the day...

I'll find the engine numbers one way or another.

The pinking issue exists for either ignition system and assuming I get it sparking will try it with no strategy link, so set for 91Ron.
My fear is the Marelli system can't fire the higher compression ratio, but I think that's unlikely as it uses two air cooled coils, one per bank, with each coil having twice as long to recharge as coil(s) firing all 12 spark plugs.

The Lucas system creates 12 trigger pulses per 720 degrees of crankshaft rotation, whereas the Marelli system creates 6 pulses per 720 degrees of crank rotation.
It takes a bit of thinking about, but this isn't a 12 cylinder engine, it's 2, six cylinder engines side by side...with two separate ignition systems being triggered by the Ecu and crank sensor.
The rotor arm has two completely separate contacts, 180 degrees apart, one for A bank, one for B bank.

I'm pleased to hear the Efi Ecu should be 16Cu, which are apparently more reliable than the 6Cu.

I am aware of Kirby Palms' book, but haven't read it.

Firstly, Xantia_V6, you were correct about the engine numbers being S and H respectively. As mentioned, the crank and flywheel sensors have to fit very close to their triggers. There had been a crank sensor bracket made previously and the flywheel sensor casting was hanging from the torque converter cover plate, so the sensor had about 5mm clearance.

I placed the casting in a suitable position and drilled and tapped the block, then installed the casting using bolts. The casting sits on two feet, which took quite a lot of fettling by hand to ensure the sensor stayed within tolerance once the casting and cover plate were bolted tightly: note the 1mm "no-go" gauge. The crank pulley sensor bracket wasn't made correctly and had to be remade. I'm currently waiting on a new crank sensor and will cover that when it arrives.

I have checked the crank and flywheel sensor outputs using both Dmm set to Ac Volts and an oscilloscope, which is obviously better. I swapped the working sensor between the two locations to take measurements and readings.

Flywheel sensor output once the cover plate was finally installed and the sensor connected to the Ecu. The closer the sensor is to the flywheel ring gear, the greater the output voltage is, hence the tight tolerance.
When using the scope one can even see the variation of voltage caused by the flywheel not being perfectly centered, so the distance to the ring gear varies slightly. That isn't visible in this snapshot.
Each ring gear tooth creates a positive voltage, each valley causes a negative voltage, thus the sine wave, A.C. output. Next was to check the distributor was aligned correctly. The drive gear inserts into the block and is driven by a lay shaft. One needs to ensure that when A1 cylinder is at TDC on the firing stroke the rotor arm is pointing towards the front of the engine and A1 ignition lead.

To ascertain A1 cylinder is at compression you can remove the intake manifold, the cam box cover and check the position of the cam lobes have both valves closed: I chose to not do that.
Unbolt the Ac compressor, remove A1 spark plug and jam some paper wipe, or polyethylene bag, into the plug hole then crank the engine in short bursts: the tissue will be blown out as A1 cylinder compresses the air, so must be on the firing stroke. Rotate the engine in the forward direction to align the TDC marker on the crank pulley with the TDC mark, which is vertically down from the pulley. I made my own marker as there wasn't one on the engine.
If the mark on the pulley has passed TDC rotate the engine forwards nearly two revolutions until the mark aligns with the TDC mark on the engine. Always rotate engines forwards to avoid the timing belt/chain jumping. Once the engine is at A1 firing Tdc, unscrew the distributor cap and check the rotor is pointing forwards. A quick check of the distributor cap showed it to be good order. Note the dual level HT connections, one set for each cylinder bank. Also check the secondary rotor arm contacts, which will be facing directly rearwards. It's always worth having a reminder with you of the engine layout and firing order...

Yesterday I fitted the new crank sensor , checked the clearance to all the trigger fingers was in specification, connected a Dmm just to see if it was giving pulses and turned the key.
I was more than a bit surprised when engine actually coughed and on a second attempt poorly limped into life and ran for about 30 seconds. The fuel is about 8 years old.

I sprayed some brake cleaner into the air filter boxes and the engine ran and then continued to run, but poorly and I only noticed smoke from the right (A bank exhaust).

This was when I noticed the gearbox was pumping Atf all over the floor, due to the missing oil cooler...

I checked fuel pressure, which when dead headed, the pump is able to achieve over 100 Psi and when cranking will maintain 3 Bar, but this depletes as soon as power is switched off.
I couldn't get it to fire again and thought the injectors might not be firing, but on reflection there is no B bank spark and having swapped the coil connections over, now have no A bank spark either, though need to investigate this further as to whether it's just the coils, or the ignition amplifiers are also faulty. I had already swapped A coil to B amplifier and it continued to create a spark, so I don't necessarily think the amplifiers are faulty.

My spark test caliper went missing years ago, but I am unable to make the voltage jump even a small gap, so think both coils are faulty.

Assuming the ignition Ecu is Ok, the shopping list is likely to be a pair of coils and amplifiers and a fuel rail pressure regulator. Having a parts bin to be able to dip into would be useful.

Sorry, no pics, or scope traces, it ended up being quite a challenging day.