Car still idles rough/runs rich after 2 months work - I'm at a wall

(mike90) #183


I just wandered onto this list and have been reading your story with interest. I tend to side with Philip (Jagwit1) that this is overfueling due to a vacuum leak somewhere. I also suspected the vac line to the ECU. Thing is, I had the very same symptoms some time back, wayyyy rich, fouling 4 or 5 sets of plugs before I could figure it out. Turned out to be a break in the vacuum line, as it ran over the trans tunnel, a problem I caused when I installed the 5 speed in the car.

But, to the issue at hand: another way you can get excess fuel into the manifolds is via the fuel pressure regulators; one of these is connected via a rubber line to the manifold, so that the pressure can be affected by vacuum levels. As these age, the diaphrams fail, and this permits fuel to run through that tube into the manifold…pull the line off at the regulator and see if it is wet. Most cars this age may show degradation of those regulator diaphrams. Not sure its enough to cause your problem, but…

Good luck,


(Carl Hutchins, Jr. ) #184

1 Yup!!

Poor ring seal does those things. Low compression and blow by!!! Signs of a well worn engine…

  1. However, the rich run is another thing. After a long trail awandering, as to causes of black smoke from my LT1, the obvious fixed it !!!

Cleaned the air cleaner!!!

did my errands in it yesterday., It purred and got admired a lady. appreciative of older steeds. Hers, very nice 88 MBZ 320!!!


(Greg) #185

OK, did a lot of testing today. Also cleaned and properly grounded injector resistor pack just to make sure.

So engine still runs rough and revs rough when cold, but runs (no change). Once warmed up to about 180F, it runs so poorly it wants to stall and is hard to rev (no change).

Disconnected full load throttle switch and full load vacuum switch, no difference.

Warmed up, I supplied fake 24 inHg to Bank B FPR (which is new), and it ran much better, but still rough, equivalent to when it was cold. (Sorry, don’t have my fuel pressure gauge). But w/ 24 inHg to FPR, it STILL runs extremely rich. Car holds 10 inHg.

Removed fake vacuum to FPR, applied fake 20 inHg to ECU, no difference whatsover.

I have no equipment to test injector pulse width :frowning:

I did compression test in October, all good (180-195)

So i think I’m now down to :

  • (1) Bad engine (maybe compression has gone bad since testing?)
    • I will need to do another compression check
  • (2) Injectors are not being told to lean down the mixture either due to
    • (a) Bad ECU - I will need to send out ECU to be checked out.

    • (b) ECU signals to injectors not getting through -

SO, I did a resistance check with GROUND on ECU harness and got this (battery disconnected from car):

  • 0.0 ohms: Pins 1 (ECU Ground),16 (Injectors Ground),17 (Injectors Ground), 23 (O2 Sensor Shield), 26 (Start signal), 34 (Injectors Ground), 35 (Injectors Ground)
  • 7700 ohms: Pins 8 (Bank B Injectors ON),9 (Bank B Injectors ON),11 (Bank B Injectors HOLD),12 (Bank A Injectors HOLD),13 (Bank A Injectors HOLD),14 (Bank A Injectors ON), 22 (Main Feed), 27 (Bank B Injectors ON), 28 (Bank B Injectors ON), 29 (Bank B Injectors HOLD), 30 (Bank A Injectors HOLD), 31 (Bank A Injectors ON), 32 (Bank A Injectors ON)
  • 7800 ohms: Pin 15 (Fuel pump relay)
  • 6800 ohms: Pin 18 (Engine speed signal)
  • Infinite ohms: Pins 3,4,5,6,7,10,19,20,21,24,25

VIN 141xxx. Looking at electrical chart, the ground signals are properly grounded. Not sure what the rest means, Any electrical gurus out there who can make sense of the ohm readings?

** history of tests: CTS, TPS and ATS all put out correct signals. Wiring from those to ECU Harness all good. Fuel injectors harness good.
Spark good. Ignition timing good. **

So bad ECU, bad engine, or bad ECU wiring to injectors. I hope I’m getting close.

My gut says, the fuel injectors are pulsing like it’s full throttle. I just don’t know why. It ALWAYS runs rich, and misses/rough at idle, but the more RPMs I give her, the smoother it runs and the misses almost go away. I just can’t rev past 3000 sitting in the garage. Maybe at 6000 it would run smooth as a top.

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #186

Looking at all the information to date I would guess the wiring and engine sensors are probably O.K.
No reason to suspect injectors or FPR.
The most obvious reason for very rich running looks like the vacuum sensor inside the ECU is faulty.
If so that would account 100% for rich running, and maybe the 10ins of vacuum is a byproduct of the richness.

The only other problem which comes to mind is timing error. That is about the last thing you would normally suspect, however it just happens I had similar richness symptoms when timing 60deg out, although never checked the vacuum.

Not sure we ever had reports of rings so bad that they could explain rich running and low vacuum. The small bleed holes in the pistons behind the oil control rings always seem clogged, but the 5.3L does not have a reputation of being an oil burner. Plus the compression tests early on seemed good.

(Greg) #187

Yeah, supplying a good 20 in vacuum to ECU and no change leads me to the same conclusion. I hope, ECU is an easy (although expensive) fix.

I 100% know the ECU can hold a vacuum, no split hose.

When faking high vacuum to FPR, noticeable difference, got 10 in vacuum. Without fake vacuum and thus higher fuel pressure, it ran terrible at only 7 in vacuum. So too much fuel definitely lowers vacuum.

I spent so much time, I’m so sure I static timed it spot on at TDC, and my timing light confirms. I also looked at my plug wires three times. 1A goes to 1. (Now, bank A is the left side? Just kidding!!!)

The pickup coil was a red herring.

(phillip keeter) #188


(Greg) #189

Is that an agreement or crazy laugh?

(Greg) #190

This Jag is helping me accumulate some cool tools!

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #191

Yeah, supplying a good 20 in vacuum to ECU and no change leads me to the same conclusion. I hope, ECU is an easy (although expensive) fix.

If you join or contact the local Jag club you might find somebody can lend you an ECU to try.
Not sure how interchangeable the various models of ECU are, but as long as the wiring configuration is O.K. chances are any one for an HE 5.3L would at least get you running without being too rich.

If that proves the point your existing ECU can be repaired if there is anybody with a stock of the vacuum sensors. Since the days when Lucas was making ECUs there are more types available.
Typically a modern 15 PSI type which is what you need costs US$10 - $25. Trouble is you have to match the output to what the ECU expects. I reckon any outfit used to fixing Lucas ECUs can do that.

For the Motec system in the coupe I use an external vacuum ( MAP ) sensor, which is a more flexible arrangement. The type used is a common GM or Delco unit, 5V supply and 5V maximum output. Just for always having a spare bought 3 of them from Chinese Ebay source for less than US$10 each.
It would not be hard to use one externally to fix a Lucas ECU.

(Greg) #192

Anybody in the NW US / Seattle area know who can repair Jaguar XJS ECUs? AJ6 in the UK would be my choice if I can find nobody local.

(Philip Lochner) #193

Why not first try a known good ECU from another car before you start down the repair route? Also, I should think it should be rather easy to find a used good ECU for sale?

(Kirbert - author of the Book, former owner of an '83 XJ-S H.E.) #195

The 6CU and the 16CU are interchangeable, although Bywater points out that the 16CU is the better unit. The later models, 26CU, etc., are NOT interchangeable.

(Dzia) #196

A couple months ago there was a thread about repairing ECU’s. I believe Veekay was involved with another guy in which they found a possible repair shop.



I’m willing to do ohm/continuity tests on my 86 16cu vacuum sensor for a reference point to compare to if you’d like to do a prelimiminary ECU evaluation. Static resistance, maybe multiple vacuum levels resistance? We could determine the curve.

Such an unlikely failure. No apparent damage to the PCB. Suspicious. I’ve searched well, and I find multiple discussions of various ECU vacuum related issues. None resolved to the sensor. Hoses, dizzy pickup, amp module, wiring, yes. Sensor no. Not saying it isn’t possible, but I have doubts. The sensor is, I believe, a coil attached to a diaphragm moving around a magnetic core. What’s to fail except the diaphragm? This unit tested holding a vacuum right? Any visible issues with either side of the PCB?

Regarding our previous discussion of the pulse durations:

On the 16cu injector pulse durations can vary from 1ms to 8.2ms.
There are 8 different vacuum load-point-based fueling maps in the 16cu. These are at 25, 22.5, 20, 17.5, 15, 10, 5, and 0 inches of mercury.
At idle speeds, these maps yield 1, 1.25, 1.6, 2.1, 2.5, 3.5, 4.8, and 6.5 ms respectively.

Additional duration adjustment is added per lookup tables for air temp, coolant temp and WOT.

Given you are pulling 6-7 inches of vacuum you’d be at least 3.5 ms with a good ECU. 350% of required fueling.

(Greg) #198

Thanks for that, but I’m not sure I’d know what I"m doing with the electronic board.

Perhaps the vacuum sensor is fine, but something else in the ECU is not working, causing the vacuum sensor to appear bad? Perhaps the electronics tied to setting the pulse duration? I still remember in the book, someone saying when the resistor pack wasn’t grounded properly (mine didn’t seem to be), it kept frying the ECU. A burned out diode in the ECU could appear fine visually.

Regarding a used ECU, they seem to go for $200-$400 on ebay, and I will have no idea if they are 100% working correctly. According to AJ6’s website in the UK, fixing mine would cost about $300. If the ECU is fine, he just charges a minimal fee for looking at it. That may be my best bet at the moment, but I’m in no rush.

(Aristides Balanos) #199


6CU Part Numbers and Applications
DAC1926 - pre-HE Cat
DAC1936 - pre-HE non-Cat
DAC2596, DAC 3062 HE non-Cat
DAC2597 - HE CEI O2 Sensors
DAC3586 - HE CEI O2 Sensors

16CU Part Numbers and Applications
DAC1926 6CU 1980 pre-HE OPUS 9.0:1 O2 Sensors
DAC1936 6CU 1980 pre-HE OPUS 10.0:1 NO

DAC2596 6CU 1980 HE CEI 12.5:1 NO
DAC2597 6CU 1980 HE CEI 11.5:1 O2 Sensors

DAC3062 6CU 1981 HE CEI 12.5:1 NO

DAC4118 16CU 1985 HE CEI 11.5:1 O2 Sensors
DAC4119 16CU 1985 HE CEI 11.5:1 NO
DAC4478 16CU 1985 HE CEI 12.5:1 NO

DAC4585 16CU 1986 HE CEI 11.5:1 O2 Sensors
DAC4586 16CU 1986 HE CEI 12.5:1 NO

DAC6335 16CU 1989 HE CEI 11.5:1 O2 Sensors
DAC6336 16CU 1989 HE CEI 12.5:1 NO

DAC6337 16CU 1989 HE Marelli 11.5:1 O2 Sensors
DAC6338 16CU 1989 HE Marelli 11.5:1 NO

On my 1986 Canadian spec V12 XJ I replaced the 6CU (DAC3586) with a 16CU (DAC6335) and there was indeed some improvement.
I found it on ebay, got lucky I guess, not expensive and worked fine.

Sounds like a good plan and you will know it works.
You could even go for the enhanced version.

(Philip Lochner) #200

An excellent post!! Thank you!

(Greg) #201

Hmm, interesting. Mine is dac4118.

Yeah, I’d roll the dice and purchase a used one for $100. But not for $300.

(Aristides Balanos) #202

So you have a 16CU.

Any of them will do.

(Greg) #203

So a decent 15 inHg at idle would = 2.5 ms. If my ecu is not seeing vacuum = 6.5 ms. Is that difference enough to make it run extremely rich?