I am having a weird problem that I would appreciate your opinions on. To make everything better, it comes and goes. 1987 US spec V12.

Two weeks ago (July 4th) I had agreed to give a friend a lift and as the weather was nice I opted to take the XJ-S. The hour and a half drive out went without incident, almost all interstate. The trip back was on smaller roads, 50 mph, but with stoplights. This is where the problem first appeared. While stopped at a light, the engine cut out. It then caught itself and continued running. The second stoplight went without incident. At the third the engine died and didn’t catch itself. It restarted just fine. I changed the route to one with less traffic and fewer lights. Dropped the friend off, it idled fine there, and headed for home. The car wasn’t happy, lurching when I first stepped on the gas. It almost died once more, but I popped it into neutral with my left foot on the brake gave it enough gas to keep it alive. Made it home without a tow truck.
The next day I started it and it tried to die almost immediately. Again more gas kept it going but something was definitely wrong. I didn’t really have time to work on it, so it sat for a week. The next time I started it everything was perfect. Idling at about 730 rpm, nice and smooth, all the way up to temperature. Took it for a quick drive and everything seemed great. That was the 11th. Took it for another hourish drive on the 13th and all was well.
Today has basically perfect weather and I took the car to work. An hour/35 miles each way, mostly rural twisty roads with a bit of small city at the end. A lovely drive for a car like this. Going to work everything was good. Coming home I had problems. Shortly after leaving work I stopped at a red light. Engine speed dropped momentarily to maybe 250 on the gauge and then resumed normal performance. Coolant temperature had only just started to move the needle. Half an hour of driving later it died at a stop sign. I waved a couple of cars around me as I tried to restart it. The starter was spinning the engine but the car wouldn’t start. What felt like an eternity later, but was at most 90 seconds, the car splutters into life. It didn’t catch all at once, but it was quickly fully running. I take off and get up to the speed limit but the car isn’t happy. Any time I put my foot on or off the gas pedal there’s a lurch. If I let the engine speed fall too low it tries to die. I found that second gear between 2000 and 2500 rpm was a good compromise between the 40 mph limit and keeping the car running. But then, after about 15 minutes of this, everything was fine. A steady gentle response to throttle, idled smoothly at stop signs and red lights, I was even able to back up my driveway and into the garage. The problem had just gone away.

Now I don’t think the problem has actually gone away but I don’t know what is causing it. We’ve got spark, and we’ve got fuel. I don’t really suspect a clogged fuel filter because it runs well at higher revs. I checked the throttle disks and they’re between 0.002” and 0.003” and very even side to side.

My suspicion is the Throttle Position Sensor. I recently replaced the worn out original with the Hall effect one described in a thread here recently. My experiences with that can be summarized by “It’s great!” “It moved, I didn’t tighten it fully” and “it’s great!” The day it moved and went out of the specified range was the 4th, ie the first time I had this problem. However, I can’t move it now. I haven’t actually checked the voltage, but the sensor is tight and can’t be moved by hand. Last time it moved easily.
Is the sensor just bad? Can a misadjusted TPS cause the engine to die at idle? Or is there something else going on?
Before I changed the TPS I had the lurching when first opening the throttle, but it never died. That’s new, and it’s the only real work I’ve done on the car this year.

 

My understanding of the Hall Effect is the same as yours, or less.

TPS would be my go to.

Greg has a Hall Effect on his, and when the planets align, he will see this and offer some pearls.

Dont overthink it, the last thing you played with is 99% the culprit.

 

Not to be a Jeremiah, but I do not really trust cheap electronic components, especially in the engine bay deep in the V. My Hall effect TPS cost the equivalent of 220 USD from a motorsport specialist component company.
Having said that, the car will tickover, but not rev, with no TPS connected, however if the signal is somehow getting scrambled, maybe the fuel injector cutoff (trailing throttle above 1400 rpm) is not getting restarted by the ECU when the revs drop below 1400.
Therefore, following Grant's 99% dictum, it is highly likely the new TPS is the problem.

The first thing to do is to check the voltage throughout the rev range on the TPS. If that seems OK, I would be inclined to refit the old one you removed and see if the problem goes away. It could be that the new one is breaking down when it gets hot and somehow fooling the ECU.

I would also carry a can of Easy Start, and the next time it happens squirt a GOOD amount down the inlet trumpets and try the starter. If it starts/coughs, you know it is a fuel problem, which, had you not changed the TPS, I would have been my first thought.

 

Unfortunately I cannot refit the old TPS. I cut the lead off when I did the modification. This let me keep the original plug, which seemed like a good idea. The original potentiometer has been in a landfill for a month or more at this point. It was very worn out, with voltage fluctuating wildly when I tried to adjust it, which is why I was so enthusiastic about replacing it.

I have ordered a different sensor, this time from a major electronic supplier; Newark. This one is designed for harsh conditions, including transportation, and only slightly more expensive at $40. Should be here in a few days. In the meantime I will go check my solder connections, in case its something basic like that.

 

Have you a link to the Newark one please?

 

I was going to wait until I knew it worked, but here you go!
https://www.newark.com/amphenol-pihe...med-NA-Product

 

That sounds identical to what mine was doing before replacing the TPS, I would definitely check your output voltage and make sure the sensor hasn't gone bad

I'm not sure how much sense it makes but the only other thing I did when fixing this issue was adjust and clean my overrun valves (the triangular doo-dads on the front of both intake manifolds), if yours are untouched like mine were you'll likely find they're the source of a massive vacuum leak at idle.

Please do let me know if the sensor you found seems more robust and what you find on the other one, I would hate to be recommending something unreliable!

 

Interesting. I definitely haven't touched the overrun valves, doubt anyone before me did either. I'll check them out. What is involved with adjusting them?

 

the valve assembly is basically just a nut and bolt with a spring and plate captured between them, tightening the nut should make it seal tighter. While you're in there I'd do new gaskets (2 per side) and clean the plate and mating surface with a brillo pad or similar. It took me a couple tries to get them adjusted to a point where they didn't leak, I think going too tight will cause backfires but I'm not 100% sure, there's not much info I could find about them.

You can easily check if they're leaking by removing the air filter, starting the engine, and putting a finger over the small hole directly forward of the throttle bodies, it should not draw in any air.

 

Today was the first day this week I've been able to get out to the garage, its also the day the new sensor arrived. I did a bit of poking around and found a few issues.
One of the vacuum ports on the overrun valve was open to the world, so I put a cap on it.
Started the car, it idled fine. Gentle pressure on the accelerator resulted in a gentle increase in engine speed. What was unexpected though was that a quick sharp stab at the accelerator resulted in the engine cutting out, although it caught itself every time. I was doing on/off, I'm not sure what would have happened if I stomped on the gas and held it.
Turned the car off and took the throttle capstan apart so that I could get to the TPS. Multimeter connected, ignition on, reading 0.31V. Technically out of spec, but not wildly. Gentle turning of the the throttle disk resulted in a steady rise in voltage. Flicking the throttle disk gave differing results, sometimes it would jump up and sometimes it wouldn't move. My multimeter is not a high end one, so its possible that it just couldn't keep up, but I somehow doubt that.
The new sensor does seem to be better, its also lower profile. I'm now going to modify the design of the adaptor to suit it.

 

There is good news and there is bad news. The good news is that the new sensor fixed the quick throttle dying. The bad news is that this sensor isn't really suitable for this application. It took a bit of digging through the documentation to find it, but the minimum voltage output is 10% of the input, or on my car 0.45V. Its kinda hard to set it to 0.34V if the lowest it will go is 0.45V. There is also a large angular range where it outputs 0.45V. I managed to position it where it was just starting to increase with rotation and tested it. The in-garage test went well, even with the higher voltage. I'm going to print another adapter, I need to reposition the bolt holes, then I will test it on the road.

I thought it would be ok to test it with the screws only holding on a little bit, but I don't want to drive it like this.