My first guess was he had a connector making flaky contact someplace or a ground making flaky contact. Usually when the ballast resistor starts to go, it will starve the fuel pump of current (low fuel pressure) or even more common the motor will start off of the starter relay (fuel pump power bypass), then the motor will quit as soon as (or a few seconds after) the key is returned to the run position after the motor starts.
Could be his fuel pump power bypass wire from the starter relay has an open circuit and he is getting enough fuel for a start from the prime or when the fuel pump relay closes- Just a wild arsed guess.
If you get a bad ground someplace no telling how the current will flow, backwards through another sub system, your guess is as good as mine, whatever the path of least resistance is..
Jumping the ballast resistor for a test is common practice, the ballast resistor is really unnecessary, but it does have to work as long as it is in there. They say the only real function the ballast resistor has it to cut down on pump noise. My 87 never had one and worked fine, my 88 does and I had to replace the ballast resistor once. My 88 would starve for fuel at interstate speeds when the ballast resistor was slowly failing..
Well , there it is . Back to basics . Hope Tristan owns a computer compliant 10 mega ohm impedance VOM to start checking grounds and voltage drop as its back to basics . Somewhere on the sheet metal of that XJ is corrosion that is wreaking havoc and inviting gremlins to party on with his electrical system . Join the crowd as we all battle electrical glitches and enjoy less than proper performance from our mighty four wheelers . Not so mighty when we step on the gas and it misses or just lags along and the gas gauge goes down faster than the speedo goes up.
Very poetic , but in reality , time to get dirty . Time to disassemble connections starting from the battery , take measurements and clean all the way to the negative and o2 and PCM grounds.
With have to reach under and clean that ground connection at the sending unit even though it mainly serves to make connection for the gas gauge. But yeah , sounds like some engine harness insulation by the power distribution center is brittle . A good cleaning of the ballast connectors too as they are always over exposed and corrosion just infests them . PBS blaster makes a product that is supposed to be good for this , wonder if that will help him ? At least he knows not to fall into the "parts changing trap " without proper diagnosing first . That a boy !
We'll make a fine four wheeler of him yet ! I remember , " bless me father for I have sinned , this is my first four wheel drive ..." Lol. We all learn that jeeps are just not maintenance free or they would be made of vinyl . I would say he should start by inspection of wiring too and move on to cleaning connections . With any luck , that may do the trick . If not , time to give the VOM a workout . Just don't rest on anything hot or pull on the leads so it falls on hot or moving parts . We lose more meters that way ! I abused the hell out of my high school radio shack meter
i learned from then . COMPUTER SAFE is the ticket . He will have his hands full , this will no doubt start a thread of a few megs . Well , let me not put you to sleep , I know what splits are . They consume life. Ok , Tristan ? Whatever questions you may have , I think you got some time under the hood but what ever you may need to ask , I'll do my best as MudderChuck Has been working hard and we got to take it easy on the man. Stay tuned and as EricTheCarGuy says , stay dirty .
P.S. - my brother always wears gloves when he works on his car , he's a guitarist to and yells at me when I don't . Can you say , dexterity ? Hate gloves . Greg