1997 XJ6 non supercharged automatic tranny. Runs really good when engine is cold. After 5 or 6 min. it starts running bad, as if someone was turning the ignition off and back on very quickly. I have tested all the pin voltages coming out of and into the Engine Control Computer. There are 2 voltages that correspond with the problem. The ground that controls the Air Injection pump relay, and the ground that controls the fuel pump relay. Every time the car acts up both of those grounds are dropping to 0 volts and snaps back to full ground (-) in ms (I am using a vantage pro dual trace graphing meter from snap on tools) I have also exchanged both the relays with no change. I thought it was the Computer, so I replaced it. still no change.
I am a Master technician that is lost. I have tested everything possible that my info and years in this field allow me to test. If anyone could help I would be most appreciative to them. Thanks : Autotec40

 

If I read correctly, you've already replaced the Engine Control Module? There have been numerous reports of clogged sunroof drains leading to significant corrosion at the ECM connector. I'll assume you didn't note any when you exchanged the module. This thread may be of interest to you: https://www.jaguarforums.com/forum/x...-worse-125486/
Particularly Vee's suggestion in Post #9.

 

The ECM was dry without corrosion. I checked all the connections to and from the Ecm, also checked the wiring resistance to all sensors, powers, and grounds. The crankshaft sensor has a normal signal.
The 2 control grounds I am talking about in 1st post are sent from the ECM. and I am testing them at the ECM. That is why I replaced the computer. Does anyone knows of a reason it would dither the ground signal to those relays ( fuel pump, and Air pump ) ?

 

I don't quite understand the terminology you have used.

However, consider both ends. If you lose the +12V feed at the other end of the control circuit, your ground reading is going to be strange as well. Try starting the car, then supply the other end with +12 using small test jumpers. If the problem disappears ... start looking over there.

Alternatively, start car, supply ground at the control terminal at the controlled device. If the problem disappears, start supplying ground progressivley further away.

Also look for connectors that loosen when warm.

 

Basically the computer sends a ground out pin 19 of the black high voltage connector to order the fuel pump relay to close and send power to fuel pump. It also sends a ground out pin 9 of the same connector to order the secondary air injection relay to close and send power to the Air Pump Motor. When the vehicle starts acting up Both Pins 9 and 19 signals starts jumping from full ground to 0 ground. This is what I believe is causing the engine to surge, misfire, and stall. What I need to know is what tells the computer to activate and deactivate the ground on these two pins. As I said before I have diagnosed it as a faulty ECM, replaced the ECM, and the problem is still there. I hope this clears up my terminology. Thanks

 

I still don't know what you mean by full ground and zero ground.

The ECM does not really send a ground anywhere.

It sinks a signal to ground. That is, it provides a path to ground under certain circumstances. This is usually done using a driver ic or transistor.

The above is why I suggested going back to basics and providing grounds and power with test jumpers as a diagnostic measure. Doing so at the relay eliminates the rest of the particular path.

 

I am saying just what you said. ECM provides a ground to relay in order to control the power to the said Fuel Pump or air pump. I will take your advice and jumper terminals 19 and 7 to ground and see if it runs better. But If it does fix the problem, then I am back to why is it losing that signal to the relay from the computer. I will be in touch tomorrow with results. Thanks again.

 

The reason I am being vague is that I don't have the same series, so no wiring diagrams or circuit descriptions at hand.

If you are consistently losing the signal at the ECM, maybe you have a bad joint on the circuit board that opens with heat or a bad driver chip. Some of them are ganged, meaning one chip has multiple sections.

Try varying temperature by blowing a fan at the ECU or surrounding it with ice packs to keep it cooler. If it lasts longer, then likely ECU internal.

Could be almost any chip/component/junction getting hot. Some people are good at finding them by visual examination with a loupe. Another technique is to blast suspect components with an air can nozzle to chill it.

 

It was one of the coils. I was reading another forum here that was talking about Ignition coils, after reading them I thought what the heck. I ordered 6 new coils ( Made in Japan )and installed them. The car has not messed up in 3 days of test driving.
I assume the coils were allowing a spike of high voltage to feed back to the ecm and cause it to go crazy. Anyway thanks for all the help.

 

Chalk up to yet another crazy coil induced problem. Wow

...one has to wonder how these coils can cause so many odd issues such as dash disco lights, etc.

.

 

Back around 15 years ago when I worked at the Jag Dealer we had a customer car in the shop that did the same thing. We watched DATALOGGER with WDS and the ECM would momentarily 'blink' and lose signal. The engine would 'hickup' or shut down and we FINALLY discovered a faulty ignition coil arcing and sending a BIG SPIKE back to the ECM causing all the problems.

New coil and all was well.

I have related this story a few times on the X300 forum over the years.

bob gauff