Jaguar XJ-S. Manual — part 69

271

Tightening the locknuts requires three hands: one to hold the 1/4” wrench holding the ball socket still, one turning the
5/16” wrench to tighten the locknut, and a third to keep the crossrod from turning as you tighten. If you have a third
hand available, a pair of pliers might be preferable to expecting to hold the crossrod still with fingers. If you don’t have
a third hand available, a pair of vice grips will work if you carefully prop them on the fuel rail.

Make sure the ball sockets end up properly oriented 90° to each other when the locknuts are tight. The flats on the one
at the turntable end should be horizontal and the flats on the one at the outboard end should be vertical.

When done, check your work: put that .002” feeler gauge in the butterfly stop on each side (or one in both sides, if you
have two .002” feeler gauges) and very slowly and carefully turn the turntable (without touching the crossrods
themselves) until the feeler gauge drops. Note the gap between the turntable and the idle stop when they drop.
Officially this backlash at the turntable stop is to be between 0.062-0.094” (1.57-2.39mm)(the .002” feeler gauge adds a
bit to this, but it’s negligible).

Also make sure the two butterflies open as close as possible to the same time. If they don’t, the TSB allows an
additional 0.030” of turntable travel to get the second throttle to move off its stop.

Adjustment 5: Loosen the locknut on the full throttle stop screw on the throttle pulley, and back the stop screw away.
Hold the pulley in full throttle position, noting that the butterflies are both full open. Adjust the stop screw until it just
touches the pulley and retighten the locknut. This stop screw merely prevents stress on the linkage while the engine is
at full throttle and the kickdown switch is in operation, and is not meant to restrict full throttle.

Ensure that the throttle moves freely through the full range of motion. Finally, check for no gap at the turntable stop
with the throttle in the closed position. If the cruise control cable is too tight, it can restrict the throttle linkage moving
fully to idle.

You will need to warm up the car and readjust the idle speed, since the butterfly stop screw positions have been altered.

THROTTLE STICKING: Apparently, all cars have occasional problems with throttle sticking due to buildup on the
butterfly itself. A butterfly cleaning procedure from Randy Wilson: “You are cleaning the throttle plate and
surrounding area. The edge of the butterfly and the area of the housing right around it will be covered in black goo.
Prop the throttle open and wipe the stuff out with a rag. Use the weakest solvent you can to get it clean. I start with a
“WD-40” grade oil, and go to carb cleaner if it’s really bad. Oven cleaner is out. Sand blasting is not needed or
recommended.

“Warning: Some non-Jag cars, notably later Fords, have a teflon coating on things in an attempt to reduce this problem.
Most solvents will damage this coating.”

THROTTLE LINKAGE LUBRICATION: Jan Wikström says, “The ball-joints shouldn't be lubricated at all, unless
you want to give them a dusting with Teflon or graphite. The reason is that grease hardens and oil or grease picks up
dust from the air, increasing linkage wear and friction.

272

Electronic Fuel Injection

If you wish to understand or troubleshoot the Jaguar XJ-S electronic fuel injection system, it is highly recommended
that you obtain a copy of “Understanding Electronic Fuel Injection, Ignition and Engine Management” by Roger
Bywater and available from AJ6 Engineering (page 710). The book is written as a primer on EFI in general, but since
Bywater’s expertise comes from years of working for Jaguar many of the examples presented happen to be Jaguar
systems. Often, it seems he is talking about these cars in particular.

According to Dick White, “There is a book ‘Bosch Fuel Injection and Engine Management’ which I found very
informative. It contains many references to Lucas.” It is available from EWA (page 721), among other sources.

FUEL INJECTION SYSTEM TYPES: The 1976-80 XJ-S used a type of Lucas fuel injection system based on the
Bosch “D-Jetronic” design. From 1980 on, the cars used a system referred to as a “Digital P”, also based on a Bosch
design. The two systems are completely different. Among the differences: The earlier D-Jetronic system used a
trigger board within the distributor operated by a magnet in the rotor; the later Digital P system simply derives its
triggering from the ignition pulses. The D-Jetronic has a “throttle switch” in the turntable on top of the engine that
provides an on-off pulse about ten times as the throttle moves from idle to full open; the Digital P has a “throttle
potentiometer” in the same location, providing a smooth varying resistance as the throttle moves. The fuel regulators
on the D-Jetronic system maintain the pressure in the fuel rail at a constant value, while the regulators for the Digital P
system vary the fuel rail pressure according to intake manifold vacuum.

The D-Jetronic has no oxygen sensors, lacking the sophistication for feedback control. The North American Digital P
has two oxygen sensors, one in each exhaust downpipe, and Roger Bywater reports that the ECU will control the
mixture in each bank separately. The Digital P systems in countries without emission regulations had the feedback
capability omitted from the ECU, and had no oxygen sensors.

Bywater elaborates on the various systems: “D-Jetronic ran from 1976 to 1980 in the XJ-S but, just to confuse
things, ran until 1981 in the saloon. Digital P 6CU first appeared in July/August 1980 on the 10:1 compression flat
head engine rated at 300 b.h.p. in the XJ-S. Most XJ12 saloons continued with the earlier 9:1 D-Jetronic engine
until the arrival of the H.E. engine in July 1981 but a few saloons were fitted with the 10:1 engine and Digital P. I
understand that the improved fuelling accuracy of the Digital P EFI system was what made the higher compression
engine viable.

“The 10:1 engine performed very well and was certainly the most lively version of the 5.3 ever produced but there
were two things which could make it a bit fragile at sustained high speed. Firstly the cooling ability of the radiator
was a bit marginal, secondly the distributor build quality was not too good and we measured errors of up to 10
degrees excessive advance at 6000 revs, which is obviously likely to cause trouble. Not surprisingly, there were
tales of piston failures on German Autobahns.

“Those early 6CUs triggered from the coil negative terminal and the ECU was protected internally from the high
voltage firing spikes by a 100K resistor in the input circuit. This was not necessary with the Constant Energy
system introduced for the H.E. and was replaced by a link. Another important difference, apart from the fuel
requirement of the 10:1 engine being different from that of the later H.E., was that full load fuelling was set into the
mapped fuelling values so there was no need for a separate full load enrichment switch circuit. This means that
fitting an ECU intended for an H.E. car to one of these engines is likely to magnify any tendency to burn pistons
because the full load condition will be excessively weak. A number of people have found this out the hard way.
Strangely the ECU seems to survive the high voltage firing spikes despite not having the protection resistor.

“The H.E. engine arrived in July 1981 in both the XJ-S and XJ12 saloon so that is when D-Jetronic really ended.”

INJECTOR GROUPING: The wiring harness is set up to operate the injectors in four sets of three. Within the ECU
these circuits are combined to form two sets of six. In the early D-Jetronic systems, each set of six injectors included

273

three on each bank, and this injection pattern is explained in great detail in the repair manuals. However, in the later
Digital P systems (which include oxygen sensors), each group included all six injectors from a single bank so the ECU
can monitor the oxygen sensors separately and control the mixture of each bank as a group.

OVERRUN CUTOFF: On the Jaguar V12, some EFI systems feature overrun cutoff and some don’t; see the section
on overrun valves on page 266.

When the foot is all the way off the throttle but the engine speed is well above idle, the pilot is clearly not interested in
power; any fuel provided is a waste of fuel and an unnecessary contribution to emissions. So, if the ECU detects that
the throttle is at idle and the RPM is above some set point (1500-1800 or so), fuel to the injectors is totally shut off. As
the RPM drops past another set point nearer to idle speed (1100-1300 or so), the fuelling is turned back on to provide a
smooth transition to idle -- or to rev it back up to the cutoff RPM. Roger Bywater: “It's easy enough to check on the
car by just flicking off one throttle pushrod and opening that throttle to lift the speed without the throttle switch being
moved (assuming the switch is set correctly).”

Regarding which EFI systems have overrun cutoff, Bywater says “...all 6CUs do indeed have over-run cut off above
approximately 1500 r.p.m. but it is activated not only by a closed throttle signal from the throttle potentiometer but also
by the vacuum signal. On a Federal 6CU cut-off commences at nominally 17.5" Hg and fuel reinstates when the
vacuum falls to 15" Hg. Cut off on European versions is active over a wider vacuum band which also varies with
engine speed.

“16CUs only need the throttle closed signal to activate cut-off above the required speed.”

This feature can easily confuse the novice mechanic that may not be familiar with it. What will sometimes happen is
that the car is running poorly, and the idle screw will have been opened up to maintain a reasonable idle speed. Then
the problems with the engine are corrected, and the engine then tries to idle faster -- but hits the cutoff RPM, shuts off,
drops back to the cut-in RPM, kicks back on, and sits there cycling up and down indefinitely. The mechanic thinks he
has really screwed something up this time and begins all sorts of ill-conceived corrective measures when all he really
needs to do is close down the idle screw a bit. It’s even more confusing, of course, when the overrun cutoff feature is
temperature dependent so it doesn’t cutoff when the engine is cold but starts cycling upon warmup.

Another common cause of trouble is the infamous auxiliary air valve (page 263). The car is running fine one day, but
the next day the auxiliary air valve sticks open and the engine starts ramping up and down upon warmup. The
mechanic drives himself crazy trying to figure out the ramping, while what he really should be looking for is the cause
of the high idle.

For the overrun cutoff feature to work right, the ECU clearly needs to know the throttle is at idle. On the D Jetronic,
there is a special contact in the throttle switch for idle. In the Digital P, the throttle pot must be adjusted properly (page
295).

EFI GROUNDING PROBLEMS: Michael Neal sends this experience: “...the radiator had been replaced and a ground
for the fuel injection harness at the right front of the engine compartment had not been re-secured. It had blown two
ECUs before I found this. I don’t know why but it took the ECU a week or two to blow. After I replaced the ground
and ECU there were no further problems.”

Reportedly, if the engine clearly is running very rich or possibly won’t even start because it’s too rich, it’s time to start
looking for grounding problems in the EFI wiring. John Napoli relates one experience: “Although ground to the ECU
was one of the many things we had checked early on, the owner later went further and checked the ECU connector. He
found an intermittent ground on pin 1 at the connector. This pin, according to our schematics, should ground to the
frame of the car at the ground points near the battery. This wire was identified in the loom and a splicer to ground
added. The car started fine and at this point is OK.”

274

ELECTRONIC CONTROL UNIT: The following is a list of the ECU part numbers, hopefully mostly correct. Thanks
go to Jeffrey Gram, AJ6 Engineering (page 710), and other sources.

D-Jetronic: Type 3CU

LRZ103 and LRZ113?

Digital P: Type 6CU

catalyst version: DAC 2597 and DAC 3586

non-catalyst version: DAC 2596 and DAC 3062

Type 16CU

catalyst version: DAC 4118, DAC 4585, DAC 6335, and DAC 6337

These are the low compression version - presumably 11.5:1

non-catalyst version: DAC 4119, DAC 4478, DAC 4586 and DAC 6336

These are the high compression version - presumably the 12.5:1

Additionally, there is a low compression non-cat version: DAC 6338.

For one year (1981), the XJ-S had Digital P EFI on a pre-H.E. engine. Clearly, the fuelling maps for such a car would
differ from those for the later H.E., and the ECU’s would not be interchangeable. It’s not known if any of the part
numbers above apply to the pre-H.E. Digital P.

Note also that the Marelli ignition system provides a cleaner signal to the EFI system, so the ECU’s fitted to Marelli-
ignitioned cars have less filtering in the signal input circuitry. This may help account for the profusion of 16CU part
numbers listed above. The pre- and post- Marelli ECU’s are not interchangeable.

Richard Mansell quotes from a Jaguar publication on the changes for the 1992 model year:

“The new fuel control system, the Lucas 26CU, replaces the 16CU and provides the following benefits:

- Improved starting performance at all temperatures.

- More efficient engine operation during warm up.

- A diagnostic facility which stores fault data for later retrieval.

- Low quiescent current drain.

- Eprom socket to allow modifications.

- JDS diagnostics serial link.

- Fuel used output.

- Enhanced software.

- Enhanced limp home capability.

With the pre-facelift ECU, (Lucas 16CU), one injection is delivered per engine revolution. In the facelift condition, the
Lucas 26CU, there are three injections per engine revolution. The number of pulses is reduced to one per revolution
after a set number of injections.

After-start enrichment - The enrichment is now coolant temperature dependent. It decays away over a number of
engine revolutions instead of over a time period and is no longer linked to cranking fuelling so that better calibration is
achieved.

Acceleration enrichment is reduced; calibration is improved; over enrichment is bettered in the occasions where the
driver opens the throttle very quickly.

Full load Enrichment - Where full load enrichment was determined by vacuum operated switches sensing inlet
manifold depression, it is now determined by throttle potentiometer position and engine speed.

There is a single 2.5 bar mini fuel pressure regulator instead of two.”

Roger Bywater continues: “The 26CU controller introduced in 1991 for 1992 model year was superseded by the 36CU
type part way through 1993. 36CU was the next stage of development of the 16CU, 26CU theme and was itself
superseded at the end of 1994 by the much more advanced Pecus / Nippondenso system of full engine management.”