Gauges - no bus - strange unsolved

98 XJ Automatic. Starts and runs well.

Key in turned to on position - 'no bus' and no gauges(any). Disconnecting the TCM brings the cluster back to life and odometer appears. Plug TCM back in and start xj and drive for a while and then cluster dies again...

I have swapped out the TCM and PCM and Instrument Cluster with known working modules and exactly the same every time. Disconnecting the TCM brings the cluster back to life. Sometimes flicking the wipers makes it come back to life as well.

Strange thing is the same thing happens if I turn the ignition to the on position, and then disconnect the PCM C3 connector, all the gauges come back to life - just presented with no bus, and plugging it back in makes the no bus disappear...

I have found and cleaned to bare metal every ground except G108, which I cannot find under the dash... but the pin on the cluster C2 harness tests good for that ground. (How do I get to g108??)

I have cleaned/scraped the contacts behind the cluster. Removed all the connectors held in by 10mm bolts, cleaned the contacts and tightened up. Cleaned the contacts in the 3 wire connectors near rear of engine compartment passenger side. Checked every fuse.

Is there any reason in particular that disconnecting the tcm or the pcm c3 connector and plugging it back in would bring the cluster and bus back to life????

No real reason, but just a gut feeling, unplug your CPS the next time it happens and see if your gauges come back. Numerous sensors share the 5 volt supply circuit from the PCM. Just a guess but maybe one of the sensors has a partial short that sometimes overloads the 5 volt supply and/or after the 5 volt supply has a chance to cool down some it will work again for awhile. I picked the CPS because I've seen it before, never threw code, but it was shorting the 5 volt supply circuit enough to kill the bus. Volt test your 5 volt sensor supply circuit.

This diagram is for a 99, I don't have a 98 diagram. The configuration (which sensors share a circuit tree or in other words are parallel wired) changes some from year to year, but is more often than not mostly the same.


bild upload

(Key in/ON position) Disconnected all the sensors one at a time and all together... No change..

Voltage on harness to sensors tests 5.2 volts.

With all sensors disconnected (or connected), Voltmeter set to 20V, simply touching one pin on the CDL (pin 11, ccd bus) with one probe from the voltmeter woke up the gauges and bus. Other probe wasn't touching anything....

Seems like the bus isn't getting a wake up signal...? Totally stumped

So after days of cleaning and renewing grounds and repeating such. Replacing cluster harness, Trying known working PCM, CPS, ACM, TCM, etc.. Disconnecting all sensors one at a time. Numerous resistance and voltage checks..

Problem is solved. It was a very intermittent issue and some of the tests seemed to have varying results which made it very hard to track down, but eventually found by probing the CCD bus wires at the Diagnostic Splice Connector found the WHT/BLK wire that runs from the DLC to the cluster had a fluctuating resistance when it should be less that 1 ohm. Found the problem location and spliced in a new wire twisting according to specs and now works fine!

Thanks for all the help!