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

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #163

Sorry Philip, my Motec setup does not mention allowance for opening time of the injector, or for the closing time under spring force. It does need the type of injector to be nominated and possibly that has an offset allowance built into the times in the fuel map.
If you use an A/F meter to tune and set the fuel map times it all comes out in the wash anyway.
I have batch firing in 4 groups of 3 which I think are evenly spaced, meaning every 90deg of engine cycle. I recall the Jag ECU fires 4 groups of 3, but 2 groups at a time every 180deg of engine cycle.

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #164

The ECU always has an upper limit to injector pulse width which will be set by the frequency of opening at max RPM, and this is something like 5 times the idle pulse width. Which would be giving a very rich mixture unless the throttle was wide open and the engine under load and needing that 5 times more.

I think the point about pulling the CTS with a cold engine making little difference is the fact most of the fuel entering the cylinder condenses on the cold metal and the actual vapour in the cylinder is not beyond a combustible mixture.

With a hot engine that fuel does not condense so the vapour is way over rich and will not ignite.
When my coupe pre HE engine had a faulty CTS the engine just stopped dead while driving along on moderate throttle.

Pulling ther CTS at idle with a warm engine may not add fuel beyond the allowable limit. It may cause the ECU to add, say, a maximum of 50% to the base fuel map which is tolerable, no matter if the CTS indicates absolute zero.
Then again, the ECU may also override too much extra fuel if it detects via the TPS that the throttle is closed. A very easy piece of software to add if the programmer bothered.

Looking back at Greg’s posts he did mention pulling the vacuum hose to the ECU made no difference.
You can look at this two ways.
First it seems odd, because with no vacuum the ECU at idle will then increase pulse width by a factor of 5 ( if not overridden as conjectured before ) and produce a very rich mixture.
Secondly the mixture is already far too rich for whatever reason, so suddenly seeing zero vacuum is a non event.

The fact that Greg’s engine will idle and rev a bit when cold helps the “condensation” theory since when warmed up the engine is much worse.

There is the possibility the small vacuum sensor in the ECU has failed and is giving out a signal to the ECU which translates as low vacuum needing lots of extra fuel. I cannot think of any easy way to test this except try another ECU. Having eliminated all other influence factors that is a last ditch stand.

Another spanner in the works is the report of such low vacuum at idle. With closed throttle the engine is a simple vacuum pump, trying to suck all the air out of the manifold. It should be pulling at least 15ins and at first glance why would an over rich mixture make a difference to vacuum ?
I don’t have a vacuum gauge to see what happens to my convertible if I pull the CTS at idle and create a rich mixture. Anybody want to give it a try ?

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #165

Many people have had the shielded wire fail in the part which is in the hot engine compartment. If the inner wire shorts to the outer shield there is no pulse to the ECU and no injector sequence.
Likewise if the inner wire goes open circuit.
There is the outside possibility if the signal is noisy, with extra pulses, the ECU could create too many injector sequences and give extra rich mixtures. I doubt this can happen in practice for many reasons, and if it did it would be fairly inconsistent.

The PO had the shielded wire trouble on my convertible, and made a very untidy but 100% effective fix.

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #166

More for the never ending thread - have we set a record yet ?

Follow up on the vacuum issue. My Motec fuel map says at idle with only 10ins vacuum instead of regular 15ins the injector pulse width is about doubled.
So in the question of is it cause or is it effect, if being too rich was the effect of only a low vacuum, pulling the vacuum hose to the ECU would tell the ECU to pour on a lot more fuel up to 5 times the regular idle level. I think that would be noticeable.

Maybe sucking on that hose to create lots of vaccum would be a worthwhile experiment.

(Greg) #167

Thanks for more responses…yeah, sorry for the longest thread ever. So I’m either the slowest Jaguar repairer, or I’ve got something strange.

Like I said, can’t get to car until Saturday. But here is my list of things to do:

  • Check ECU internal vacuum and signs of any blown diodes, etc.
  • Inspect/clean resistor pack connection for corrosion and fix the ground
  • Change oil because I’m paranoid about fuel wash, will test vacuum afterwards
  • Test wires from TPS to ECU (TPS checked out fine)
  • Test fake vacuum to ECU (I now have a pump w/ gauge, so can simulate 18 inHg exactly)
  • Check intake manifold for leaks just for the hell of it

(Paul Wigton) #168

Not quite: IIRC, there was a thread of over 1100 posts!

@Andrew_Waugh can correct me, if I got the number wrong, and can link to it!

(Andrew Waugh) #169

I think the longest is What did you do to your E-Type today?

(Paul Wigton) #170

A new winnah!!!

(Robert Laughton) #171

I’ve been following this, as I too have a v12 running rich. I won’t highjack this thread with the details.

I’m also I bit lost where we are, partly due to the length of this thread, and my lack of understanding.

Someone above - Richard I think, mentioned a “dirty signal” to the ECU. This made me remember an E Mail I got from AJ6 regarding my problem, that mentioned this, and requires a scope to check.

I don’t think I saw this test addressed here, so I hope it helps.

see here:

Hi Rob,

If it is running rich like that then some things to check are:

  1. Fuel pressure – should be 2.5 bar with no vacuum connection to the outgoing regulator.

  2. Coolant sensor or wiring – it won’t be open circuit or the engine would just die but if there is excessive resistance that would cause richness.

  3. Engine running too cool – this can cause excessive acceleration enrichment which can bog the engine down on acceleration. Thermostats should be 82 C minimum and preferably 88 C.

  4. Faulty ignition trigger or amplifier producing too many trigger spikes. This can be seen with an oscilloscope attached to the coil negative terminal to check for any defect in the primary waveform. Sparks may still seem OK.

  5. A poor earth or power connection in the EFI wiring.

The throttle potentiometer can just be disconnected if in doubt – the engine will just be a bit flat and the idle trim won’t work.

An exhaust gas analysis is always helpful to figure out what is happening.

Best regards,

Roger Bywater.


I started the thread over. Reading your original post, and subsequent exchanges.

You are running so poorly you’re making no vacuum. That pooches the ECU control. Acts the same as if you had the ECU vac disconnected. No obvious ECU damage, no tale of alternator weirdness or lightning strike. I’m going to assume the ECU is good by default because they are not fragile nor prone to just fail.

The only indication of running rich is the smell. Poor running may or may not be due to over fueling, and missing multiple cylinders would cause the smell too.

Given that everything checked so far is functional, I’m beginning to wonder if you aren’t just off a tooth on the dizzy and firing at 18BTDC but sparking to the wrong electrode.

18 teeth on that dizzy shaft I believe. That’s 20 degrees per tooth.

I cannot wrap my brain around the results of firing an adjacent piston so far off time, but if you haven’t already tried remeshing the dizzy I suggest you add it to the top of your list.

(Kirbert - author of the Book, former owner of an '83 XJ-S H.E.) #173

Exactly. It’s what happened when mine failed, and it’s what happened when others failed and were discussed on this forum. On this thread we’re hearing conflicting reports, including that essentially nothing happens when the CTS is disconnected on an engine that’s running fine. This is not possible. I really think we need to get to the bottom of this, both to solve the mystery at hand and possibly to identify a general truth about the CTS and related ECU programming.

It strikes me that one possibility is that Jaguar, somewhere along the line, realized that a failed CTS causing a parked car on the side of the road was unnecessary and changed the temperature map so it simply did no more enriching beyond, say, 0degC, on the theory that nobody drives their Jag when it’s that cold anyway. So, when the CTS goes open circuit – or is disconnected – the engine runs rich, but it still runs.

(Greg) #174

This is what I would believe. Unless most Jaguar owners were driving in Antarctica, it makes no sense to default fuel mixture based on sub-freezing temperatures.

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #175

It is possible because that happens with my stock convertible at idle. Driving the car with a CTS suddenly pulled is very likely a different story, unless the ECU is programmed to cater for this.
When I had the pre HE engine in my coupe the ECU did not cater for a failed CTS.

My Motec manual has a default table ( which I use ) suggesting 50% enrichment at 0deg C, 60% at -10deg C. It does not go below -10deg C which further suggests that is a programmed limit.

(Bernard Embden) #176

Following this thread intermittently, but it strikes me that we have not confirmed whats working properly.
Suggestion, put a scope or pulse duration tester on each injector. Measure pulse duration. Is the pulse duration correct? If no, then continue with the over fueling diagnoses. If the pulse duration is correct then you have a problem with leaking injectors. I know you just had then rebuilt. But stuff happens, don’t overlook the obvious.
My two cents (Smile)

(Robert King) #177

Another easy check is disconnect both the WOT micro switch (on the throttle pedestal) and the WOT vacuum switch (blue on right intake). They need to be open circuit unless at full throttle.

(Richard Dowling, 1979 XJ-S HE+5sp coupe, 1989 XJ-SC, 2003 XJ8 3.5L) #178

Here comes another theory. Strictly speaking just a conjecture.
I know when first starting up the HE engine in the coupe I had the timing set statically and correctly to fire plug 1A at TDC. I had the dizzy cap off but had marked exactly where the 1A post would be on the dizzy body because using the Motec EFI it was imperative to get the rotor exactly lined up to allow for advance.

As mentioned I had the plug wires all out by one so all cylinders fired 60deg later than they should. No matter what else I varied, the mixture was always too rich and fouled the plugs. That agrees pretty much with Greg’s situation. I did not spend much time actually running the engine before spotting the error. although spent a lot of time figuring out what might be a cause.

Greg has timed the engine, and that should be done with a strobe light on plug 1A. If you have the strobe pickup close to plug 1A you would not get that wrong. If you have the strobe pickup close to the dizzy it would not be hard to accidentally have it on the 6B lead which is next to 1A, and you are timing the engine 60deg out.

Looking at the valve opening times there is a 34deg overlap at TDC on the exhaust stroke. I guess that lets a small amount of exhaust into the inlet manifold but not enough to be serious. Suppose the 60deg error means the plug fires at 60deg ATDC and the mixture is now rarified and burns slower. When the overlap occurs there is a still burning mixture some of which escapes into the inlet manifold thus decreasing vacuum. In turn that tells the ECU to add more fuel.

Just conjecture.

(Nick Johannessen) #179

"I could make no sense of the problem. Then I just started from square one as if the engine had been given to me and nothing could be assumed to be correct."

That’s the way to go about it.

(Philip Lochner) #180

If I still had my DD6, I would have measured it but now we are debating this matter without engineering backup.

Kirby, even if enrichment at -150 is 100% we still have a 3ms pulsewidth, not a 5-6ms pulse as at WOT (or with vacuum signal disconnected).

My point stands that a disconnected vacuum signal to the ECU will have a MUCH GREATER effect on over fuelling than a disconnected/faulty CTS.

(Kirbert - author of the Book, former owner of an '83 XJ-S H.E.) #181

So, what’s the enrichment at WOT? Because it’s not as though the enrichment goes away when off-idle. All I know is that my car belched thick gray smoke at ALL throttle positions when cold, ran worse and worse as it warmed up, and stopped running altogether before reaching operating temperature. This is FAR WORSE than what typically happens when the vacuum signal is disconnected, although I agree with you that it shouldn’t be. People have run with the vacuum signal disconnected and apparently the car runs. And of course it runs BETTER when you put your foot in it, as that drops vacuum to zero anyway.

(Philip Lochner) #182

On these old ECUs, I would expect a single enrichment parameter to apply throughout, so if enrichment is 30%, that applies to idle and to WOT. Modern ECUs I expect will have much more complicated algorithms to ensure minimal emissions.

I find it quite surprising that you find that pulling your CTS has worse results than pulling the vacuum signal. Unless I do measurements on such a car, I can not explain it. The donor XJS I bought for its engine and gearbox (my V12 Cobra project) had its vacuum line disconnected and it did idle (not to badly, but not happy either) and it did drive. Fantastically heavy on fuel, but as soon as we connected the vacuum line, she was happy again.