I had to do this job today in order to replace a throttle body that keeps jamming shut, overcoming the clutch in the traction control actuator, resulting in complete loss of throttle movement.
I was working on the engine last year and at the time I took the decision to re-route almost all the parts of the engine harness that pass between intake runners 3 and 4. The only exceptions were the intake air temperature sensor and the oil pressure sensor. I was also careful to position certain hose clips so that the screw of the worm drive was at just the right angle. All of these preparations made today's job so much easier. If anyone ever has cause to take the inlet manifold off I would strongly recommend re-routing the wiring for the knock sensors, throttle potentiometer and idle speed control valve, so they don't go through the inlet manifold. You will be glad you did it if you ever have to take the manifold off again.
If you unpick the harness to do this modification, you will find that the knock sensor leads seem excessively long. Do not be tempted to shorten them, as the length is critical to the correct functioning of the knock control system.

 

I wonder why the wires were not originally routed under the manifold.
I think I have seen a pic somewhere of a manifold with a slot cut in it for the wires to pass through. Do you think that is a bad idea?
On my car somebody has previously cut the knock sensor wires and rejoined them with a plug and socket. It had not occurred to me that the length would be critical and I did wonder why so much screened cable was bundled up in the loom. I assume its a timing thing?

 

The standard harness routing is quite neat and easy to fit when the engine is on the engine production line, not in the car. The cable length determines the impedance of the knock sensor circuit. The output signal of the knock sensors is very low, so changing the impedance changes the signal received by the ECU. Shortening the leads could change a signal that was too low to be considered knock into one which caused the ignition timing to be retarded. The leads to the knock sensor are longer than they needed to be because I had to finish the calibration of the knock control system before the routing of the harness was fixed, so I had to do the calibration with leads that were longer than they would need to be whatever harness routing was finally chosen.


I preferred not to put a slot in the manifold flange in case it affected its sealing ability. The intake system of an XJR6 that is subjected to manifold depression is quite complicated, so chasing down an air leak is to be avoided.

 

Interesting...I saw that as an odd design and thought about rerouting them when I had mine apart for its "20 Year Service" (Wow what a task...replaced all hoses, belts, rubber lines, water pump, water rail gaskets, ect) but in the end I put them back the way they were. I also noticed a splice repair in that knock sensor wire. I have an old invoice for warranty replacing several of the valves at the dealership when the car was less than a year old. So I guess the technician at the dealership did it? Everything I changed appeared to be original Jag parts and looked its age so I doubt it had been apart since.

Do you think the splice is an issue? I removed the tape and the connections were done well, soldered and clean so I re-taped it with good 3m electrical tape.

 

I'm afraid that I'm not an electrical engineer, so I can't really comment on the effect of adding a pair of soldered joints into the knock sensor circuit. Hopefully, if it was done by a Jaguar dealership, it was considered an acceptable practice.

 

Thats kind of what I thought. Best left as I found it.

 

Ends up looking like this?

 

Is there a way to indicate when the ecu senses knock? I would hope that the combustion conditions would cause knock sometimes, to which the ecu would respond by retarding the ignition timing, but how can I tell if and when its happening?