Adjusting ECU fuel mixture

XJ-S
21

Old thread, might have connections to my problem?

Hi

I drive an XJS V12 Coupe 5.3 liters 1993 with catalysts and Oxygen sensors

Have spent many hours on trying to find out why the car only spins when flooring it from stop during the first 5 or 10 minutes of driving (when starting from cold engine), and after that period of time it does not spin from stop anymore that day. The loss of power is significant, but only during takeoff and only during the first 1 or 2 seconds, during these seconds the car does not even manage to spin from stop on a gravel road!

This loss of power comes instantly, in the blink of an eye. Suddenly the engine becomes “tired”. But – only the first seconds when flooring it from stop.

The rest of that day (as long as the engine is running at working temperature) it will not spin again when flooring from stop. Have to wait many hours until egine is cold again to be able to spin the wheels from stop when flooring the pedal.

The acceleration is always normal/correct from 15 – 20 mph and upwards, the problem with loss of power is only during the first 1 or 2 seconds when flooring after takeoff.

Top speed of the car is correct.

TPS works perfectly fine; 0.32 Volt to 5.0 Volt), so does all other sensors. New sparkplugs, new ignition coils, new cables and new ignition amplifiers. Marelli system. Throttle linkage OK (and correctly adjusted), new air filters. Fuel pressure regulator (only one on my car) working fine. MAP vacuum hose to ECU works correctly.

When standing still at a red light I can hear by the sound of the engine if the car will spin or not, when OK the sound is more vivid, and the idle RPM is about 100 higher compared to when the engine is “tired”. I also notice that if I shift the gearbox to P or N and floor the pedal for half a second, the RPM “jumps” to 2.000 in the blink of an eye when the engine is OK, but when it is “tired” the increase from 750 to 2.000 RPM goes much slower. A huge difference!

Since this big loss in power comes instantly (after 5-10 minutes of driving from cold engine) I do not think the problem has to do with any mechanical issue or with faulty sensors.

Maybe the fuel mixture suddenly (after 5-10 minutes of driving) changes from normal to very enriched at idle, and when flooring from stop it gets even more enriched (open loop) and overwhelmed /“drowned” in fuel for the first 1 – 2 seconds from takeoff?

If so, how can that be confirmed? And how to fix that problem?

I have bought a graphical multimeter, so I can measure, and I can also watch the signals from the sensors. Not exactly sure what to look for though…

What is the best way to measure the signals from the sensors (air- and water temperature, flywheel, crank, TPS and so on)? Can I take the lid of the ECU and measure each of them on their respective incoming leads/PIN´s into the ECU?

I have measured the signals/readings AT the sensors, (seems OK) but I would prefer to measure at the ECU (to be sure that the wires/leads are OK between each sensor and the ECU)

Which PIN in the ECU connector is connected to which sensor?

My ECU is a 26 CU. DAC 7188.

Would be nice to also measure the output signal/s FROM the ECU to the fuel injectors. How to do that?

Hopefully your experience will help me with this problem. 13 years with a tired XJS is boring…

Kind regards

Henrik
Stockholm
Sweden

also XJR 2000 and Norton Commando 850 Interstate 1974.

22

At all times, engine runs evenly on all 12 cylinders.

23

Maybe back probe the o2 sensors and check that they are feeding the correct information back to the ECU?
Has your fuel economy become worse?

24

Hello Henrik and welcome.

Lots of info here:
http://www.jagweb.com/aj6eng/lucas_efi.php

Yes, its best to measure the sensors signals at the ECU plug to be sure, the above site has all the pin-outs.

Not familiar with the 26CU, but on the earlier versions there is the Feedback Monitor socket, located ajustent to the ECU, and you can adjust the ECU Base Fuel Map potentiometer with a digital voltmeter and a pair of capacitors in parallel.

From what you describe the engine performs well when stone cold, ECU is on open loop, but when it warms up and the ECU goes to closed loop, it underperforms, (and it would explain the very fast transition)
First thing I would check is the Coolant Temp Sensor (CTS) resistance vs temperature.
Second, the O² sensors
Third, Base Fuel map.
Fourth, Fuel Pressure when warm (maybe a tired fuel pump)

If all the above check ok, then it might be that the ECU is tired.

1 Like
25

Welcome! I also am a Norton motorcycle fan, having had many Commandos in both 750 and 850 displacements, along with a couple of Featherbed versions.

First check the capstan switch for wide open throttle and the vacuum switch; these cause the ECM to go open loop near full throttle. If they are working correctly, it removes the possibility of oxygen sensors being the problem, as the fuel is internally mapped at WOT.

Also have you tried manually putting the gear selector in 1st to eliminate the possibility that you are starting off in 2nd gear?

Another possibility would be inadequate fuel delivery, or seized/ inoperative ignition advance- as you have the Marelli ignition system, I recall there being sensors related to the ignition amplifier that could cause ignition retard.

Good luck!

26

This triggered a memory for me. The exact details are foggy, but…

Years ago, with my '88 XJS, is discovered that in some instances I could floor the throttle and leave two black marks on the pavement for 20 feet. Other times it felt like I was chained to anchor. No wheelspin, very lazy response.

I can’t remember if there was a hot/cold pattern. I do remember that everything felt normal under light-to-moderate applications of throttle. Only under heavy throttle did the problem occur.

Long story short…the torque converter had failed.

Cheers
DD

27

I recall that there is a stall speed test in the ROM; you could also note engine rpm when cold and compare to hot under hard acceleration- should be little to no difference.

28

Dear Gentlemen!

Thank you so much for your input!

Right now I am on vacation in Prague, will go back next week to Stockholm (longing…).

After a good nights sleep I have some more info:

• Forgot to mention that I often get FF44. I just read the article that Aristides mentioned (from Roger Baywater AJ6) and found that it is a common fault, often because of bad grounding of the heater of the Oxygen sensor/s. I will check that! Both sensors are brand new, but I will check the grounding one more time. Baywater writes:; “FF44 - Defaults to open loop operation”. Interesting!

• Since I have hade my problems since the car was “quite new” – I doubt the fuel pump is the reason. But I will check the pressure at the injector rails.

• I have checked Volt from Coolant temperature sensor (and also Air temperature sensor) AT the sensors, correct output at different temperatures. Now, as I have the wiring for my ECU, I will also check the readings at the ECU. Just to ensure that the signals from the sensors reach the ECU. Will do the same with signal/s from the oxygen sensors.

• Bought a brand-new ECU recently, same behavior. So, I guess both my ECU are working correctly.

• The problem with bad acceleration from standing still is not related to the gears. Same problem when I shift gear selector to 1 st and floor it.

  • I will check the torque converter - thank you Doug.

• Not sure if my car has capstan switch for wide open throttle or a vacuum switch. I will check that out! I read from Baywater that on the facelift (which I have) with ECU 26 CU, the TPS does this job. So I guess I do not have any of these switches. Will check…

• Baywater also writes that “Full load enrichment and switching to open loop is activated by the Throttle Potentiometer (TPS), no Lambda monitor outputs.”

• My ignition advance works just fine. I have made small inspection holes in each valve cover (just above the grove in the shaft – in the front – above sparkplug number one) , so with a strobe lamp I can actually see how the cam shaft/s shifts a couple of degrees in “position” (advance/retard) when the spark fires at different RPM’s . Really cool…

• Fuel economy is most of the time to high. I guess engine runs at open loop to often.

To be continued – want to get back to Stockholm and my garage ASAP (can not tell my wife – haha).

Will start with some easy things: What happens if I plug the MAP-tube from engine to ECU, and what happens if I disconnect the wires from the Trotte pot. Will be fun.

Henrik

29

If you plug the MAP sensor input, then the engine will have no way of knowing what fuelling to apply based on the air pressure in the manifold. It’ll fuel the engine based on the pressure reading it has when you sealed it, which is likely zero vacuum, or consistent with wide open throttle, so it’ll run rich at idle, run super rich at overrun and run right only at full power only.

Unhooking the TPS ought to have less effect if you are very gentle on the throttle. As the electrical response to a changing TPS is quicker than a change in MAP as you accellerate, it probably uses d(TPS)/dt for acceleration enrichment (or a mixture of d(MAP)/dt and d(TPS)/dt) to decide how to alter fuelling as the throttle is snapped open or snapped shut. (As you have pointed out already, this ECU monitors WOT via TPS directly and makes adjustments rather than from a separate sensor.)

Remember that your lambda sensors are only narrowband sensors. They can only discriminate lean/rich over a very narrow band around 14.7:1 AFR, so are probably only active over a restricted range of rpm or load - it is unlikely that the open loop parameters are set for a default of 14.7:1 AFR for all loads and all rpm values.

kind regards
Marek

30

I still have to revise this diagram, but it’s basically correct. The issue is the entire thing should be flipped top to bottom. I.e. the short row should be at top looking at the plug, and long row on bottom.

But this should help if testing what the ECU is seeing.

26CU ECU Connector Illustrated (Color)
26CU ECU Connector Illustrated (Color)1623×1027 365 KB

I’d suggest also checking the Ohm readings of the TPS during a very slow manual swing. The Ohms measurements should move consistently up or down with NO drop outs or odd readings. The TPS acts kind of like an accelerator pump, and if the ECU can’t see it’s movement in a clean way, it has to rely on the slower vac signal to catch up with the RPM signal.

Also the Marelli cars have a separate set of temp sensors and load switches for the ignition ECU that determines base idle mapping etc. It’s probably worth testing those sensors too although I don’t have the charts for them in front of me. Not sure if they have the same values as for the EFI sensors.

~Paul K.

31

Dear all!

So helpful!!

Can’t wait to start…

Will get back!

Henrik

32

Hi Gregg, which direction did you turn the potentiometer knob to lean the mixture?

Mine is a 76 D Jetronic so I’m hoping it goes the same way. A previous owner removed the tamper-proof cap so it can be adjusted.

My spark plug analysis reveals possible rich idle mixture.

Thanks. Malc