eb1

Been on the hunt for four weekends. I think I'm tracking down a sensor issue but I need some advice, please...

90xj w/ 4.0L and auto.
Stumbles at light acceleration -- loss of power; substantial hesitation.

I've performed Tips 1,3,4,5,7,8 and cleaned the IAC (but not TB).
Fuel pressure at idle is 2psi low; was 5psi low before I replaced the FP regulator yesterday.

Just took it for a test cruise:
* Started fine
* Idled fine; not perfectly smooth, but not stumbling
* Put into gear and go for a spin -- hesitation started as soon as I got out of the driveway; seemed to be constant for the entire short casual drive; did not try to stall out, just didn't have power; was stuttering and stumbling;
* Back in the driveway it sits and idles fine again.

Further testing with car idling in the driveway:
* With trans in P, slowly depress accelerator to increase RPM...
* 800 - 1500 rpm = smooth running, no hiccups, tach operates normally;
* 1700 - 2300 rpm = NIGHTMARE, runs really rough, hiccups and hesitation, tach goes to ZERO in this entire range;
* 2500+ rpm = smooth running again, tach functions normally again.

So, the "trouble" is just for a specific range of rpms AND the "trouble" includes loss of tach function.

-- maybe the TPS is bad? I tested the TPS and it tests ok; even tried the sweep test multiple times really slowly (but I'm using a DMM not analog).

-- maybe the CPS is bad? I replaced the CPS 3-weeks ago even though the original tested ok (180 Ohms, .65V); and the NAPA replacement tests ok too (200 Ohms, .48V).

I'm thinking about replacing the TPS. But only if it makes sense to someone else with more repair experience. Any advice would be appreciated.

cruiser54

Sounds like a fuel pressure/volume issue.

With maybe an ignition issue thrown in.

Got a spare coil/ICM to test with?

eb1

Thanks for the reply, Cruiser. I do not have a spare coil or ICM. But I do have some new data since my last post:
* I caved to wife's pressure to "Do something!" (the XJ is her ride) and bought/installed a new TPS from Napa; definite improvement in the engine stumbling/hesitation -- about half as bad as it was and nearly drivable now; "trouble" rpm range shrunk a bit and is now about 1600-2000 rpm; tach drops a bit when in the trouble range but no longer goes down to zero.
* I measured vacuum: 12 at EGR solenoid; 13 at FP regulator. This is low, right, and should be about 15? The hits just keep on coming.


[For the record, in addition to the "tips" completed, I also performed a basic tune-up three weeks ago (plugs, cap, rotor, fuel filter; did not do plug wires or air filter) and changed the oil/filter two weeks ago.]

Does the new data help with the problem diagnosis or just make the water murkier?

cruiser54

Why no plug wires?

eb1

The wires are due to be replaced. I was thinking about upgrading, possibly with a coil too, but haven't done the homework on best ones for a 90 XJ. So I skipped buying the wires when I got the rest of the tune-up parts. (I should have posted that earlier so you didn't have to ask an obvious one -- my bad.) Are you pro or con on an aftermarket coil? And if pro, do you have a suggested make/model?

And a another set of driveway test data:
* Much worse performance than yesterday; very similar to original bad performance; not roadworthy; but it does indicate the old TPS was probably not part of the trouble;
* Starts well after about 4 cranks;
* Idle varies between 5-10 second periods of 750 rpm and 950 rpm with slight background miss here or there;
* The exhaust smells a bit rich; not dripping gas out the tailpipe;
* Next, a slowly increasing acceleration test with tranny still in Park...
** 750-1000 rpm = no stumbling/hesitation, a background miss here or there;
** 1100-2400 rpm = increasingly poor; missing, stumbling, power loss, tach bouncing downward;
**** 1750-2300 rpm = NIGHTMARE; complete chaos; missing, stumbling, no power, tach going to zero often;
** 2500+ rpm = no stumbling/hesitation, a background miss here or there.

I've got an old set of very lightly used wires that I'll swap in just for comparison testing; can probably get to it tomorrow evening.

cruiser54

http://www.taylorvertex.com/product-search/

eb1

Thx for the link; nice stuff; set of Thundervolts ordered.

Back to my engine trouble... been thinking about why the performance seems to improve so dramatically at 2500+ rpm and how could I use that fact to help solve problems.

Decided to RTFM a bit more; found the section on Modes of Operation in Fuel System chapter; very enlightening for me because it ties rpm ranges to modes of operation and lists the sensors/inputs for the ECU in each mode -- just what I needed.

One question: does WOT start at 2500 rpm? That's the assumption I'm using for now.

I get distinctly different results in the different modes; Engine Start mode and WOT mode work best, Cruising mode the worst. Seems I should focus on the sensors/inputs that are used in Cruising mode but not used in WOT mode: NSS; O2; AC on; AC Request.

And the wife mentioned this regarding XJ's performance on one of her recent trips: severe hesitation/stumbling for first quarter mile; she was about to pull over but instead shifted into Neutral and revved a bit; engine revived and started running well, no hestitation/stumbling. NSS involvement? (Sounds like a conspiracy theory, doesn't it?)

So I just went and did a driveway test: started XJ; with tranny in Park executed an acceleration sweep test using pedal; Cruising mode was stumbling/hesitating severely; WOT was not stumbling/hesitating.

Next step: Idling in Park, depressed brake and shifted into Neutral; ran acceleration sweep test; no hesitation/stumbling in any mode; shifted back to Park; now, no hesitation/stumbling in any mode. NSS involvement???

I don't know much about the NSS, but out of the four inputs I listed above it seems like the one that could be involved in this scenario where the major delta was shifting from P to N to P on the tranny.

Could a bad NSS cause stumbling/hesitation issues?

cruiser54

NSS only involves starting.

WOT is over 80% throttle opening.

I wonder if your MAP sensor or it's vacuum source is bad.

eb1

For NSS, manual says it feeds gear selection info to ECU during Engine Start plus other modes; in Cruise mode ECU uses NSS input to up idle to compensate for engine load in the drive range. I have no idea if that comes into play in my driveway testing or test cruise results.

MAP seems like a possibility. When the XJ is running roughest, it's a lot like when you pull the MAP vac line from the base of the TB, as far as hesitation/stumbling. I've tested the MAP and checked the vac line -- seem okay; tested fine; huge difference when vac line pulled from base of TB. I'll inspect it some more.

Vacuum in general is definitely an unsolved issue, though; running slightly lower than 15; 12 and 13 at a couple of random spots. Might be time for a new set of vac lines? Current set is 5-years old. Set prior to that lasted about 8-years.

The XJ is running a bit rich based upon tailpipe odor. Maybe cracked manifold and/or bad gaskets faking-out the O2? If so, at least the O2 is working properly. Maybe MAP vac is low and MAP tells ECU to send too much fuel? If so, MAP is working right. I'll try to measure vac at the MAP.

Click-Click-Pull

After letting it run for a while, shut it off and carefully remove each spark plug and inspect. look for excess carbon or excess fuel. I suspect you have a (one or more) fuel injector issue. Do a search or someone can post a link for testing them. Good luck and please give an update when you find the issue to help the next person.

eb1

Thanks for the idea, Click. Changed plugs 3-weeks ago; all 6 were nearly identical, all slightly rich looking; normal deposits on electrodes and tower w/ some carbon at top level of threads and thread deck. Nothing to indicate differences between injectors.

Re-inspected the MAP and MAP vac line; looks right; engine idles fine until I pull the vac from the MAP sensor, then very rough until I put it back; realized I could not measure vacuum there without some sort of adapter.

Ran continuity checks on the NSS connector: all gears registered as expected except Reverse -- the XJ's backup lights are out! Also tried starting in all gears and got expected/good result.

Ran a test cycle:
* started fine
* idled in Park at about 850 rpm steady
* took it for a spin
* back in the driveway the idle is 750 rpm steady.

During the drive I got consistent results: under little or no load the engine ran well; more aggressive throttle would cause immediate hesitation/bogging/stumbling with tachometer bouncing around; extreme throttle would drive the tach to zero. While rolling in Drive, when I shifted the tranny to Neutral I could rev away with no bogging/heitation/stumbling and with normal tach reactions.

So the car was almost driveable; initial acceleration from a stop required patience and a feathery approach to the throttle; cruising at 25-30 mph was okay; getting on the gas in an attempt to go faster quickly caused immediate hesitation/bogging/stumbling.

I guess I'm back to the two primary suspects:
1. Vacuum is low at idle (12-13)
2. Fuel pressure is low at idle (29/37)

What's up with the tachometer bouncing around and going flatline under engine load? Is the tach tied to vacuum or fuel pressure in any way?

cruiser54

Tach gets it's signal from the ICM

eb1

So it's definitely an ignition issue that is driving the erratic tach.

Am I back to looking at ECU inputs and/or sensors? The CPS is involved with the tach if I remember. Or is it the TPS? I've replaced both, just in case; the originals tested well and so do the NAPA replacements. Maybe low vacuum to a sensor, like MAP, is tied in somehow?

The tach only gets chaotic when the engine is under load; then the tach really bounces, can be driven to zero under heavy load. Same for the stumbling/hesitation -- just during engine load.

And it all clears up under no load or very light load or above 2500 rpm. That's the other aspect that's odd to me but maybe a good clue: above 2500 rpm the engine all of a sudden knows how to get out of its own way. No stumbling/hesitation and the tach is working normally. Have you heard of anything similar?

cruiser54

I'm thinking your coil is breaking down under load.

eb1

Yeah, could be; it's the original 25 yr old coil.

Napa's got an Eichlin in stock for $25; I'll report back after installing and testing.