Jaguar XJ-S. Manual - part 72

 
 

283

The gaskets between the intake manifold and the heads are common leak sources.  It appears to be necessary to 
retighten the manifold-to-head nuts occasionally, especially shortly after a reassembly.  In fact, just go out to the garage 
now and torque ‘em all down. 

 

D-Jetronic 

 

RECOMMENDED READING:  Roger Bywater wrote the definitive guide to the D-Jetronic system, published in Vol. 
2, No. 2, Jaguar Quarterly (now Jaguar World -- see Kelsey Publishing, page 721), 1989.  It is highly recommended 
that owners of cars with this system obtain a copy of this article.  

Ed Sowell says 

 

http://www.914fan.net/djet.html

 

is a site providing info on the D-Jetronic system in the Porsche 914 by Kjell Nelin. The EFI systems from that era were 
similar, so this makes good reading for the early XJ-S owner. 

 

TRIGGER UNIT:  The trigger unit originally fitted inside the distributor had magnetic reed switches and a three-wire 
connector that plugs into the EFI wiring harness.  However, the replacement trigger unit uses Hall effect circuitry and 
has a fourth wire, separate from the connector; three-wire units are no longer available.  The fourth wire must be 
connected to a 12V power source that is switched on with engine ignition.  Unlike reed switches, the Hall effect sensors 
require a power supply. 

Reportedly, the official Jaguar procedure is to connect this wire to the white wire at the ballast resistor.  This, in fact, 
should provide the proper switched 12V source -- on paper, anyway.  However, Ray Reynolds reports that the 12V 
available at the ballast resistor was so noisy that it caused the injectors in his car to fire wildly, resulting in flat spots and 
even stalling at some throttle positions.  He simply reconnected this wire to a more reliable 12V source and his 
problems cleared up.  

If you have the Hall effect unit, the testing procedure in the ROM won’t work.  John Testrake provides a revised testing 
procedure:  “The Hall switch has 4 wires.  One is power for the switch to make it do its Hall effect thing.  Another wire 
is power from the ECU.  The two other wires go, one from each switch, to the injection amp which sends power to the 
injectors.  You connect one lead of the VM to the ecu wire and the other lead to one of the injection amp wires and 
crank the engine.  You should get zero resistance when the rotor passes over the corresponding side of the switch.  On 
an analog VM, the needle will bounce toward zero and then rest again on infinity.  Now connect the VM lead to the 
other injection amp wire and crank again.  You should get the same result.” 

 

ROTOR:  Mike Morrin reports:  “I found that the D-Jet trigger pulses were not present, it appeared that one of the reed 
switches was stuck closed.  Not having a spare trigger assembly, I called my local supplier of used Jaguar parts, and he 
produced one of the later Hall-effect trigger units.  I took it home and fitted it, and found that there were still no pulses, 
but meanwhile my original trigger unit has started working again. 

“So I took back the 'new' unit, but I thought that perhaps it needed a different rotor with a stronger magnet, based on 
experience I had with Hall-effect devices in another application.  My supplier then produced a rotor, which was clearly 
different to mine, with a different shaped magnet.  I took that home and tried it, and found that the 'new' trigger unit 
now worked. 

“So beware, there are two types of D-Jetronic distributor rotors.  The ones with the narrow magnet (P/N 54422422) 
only work with the old style trigger units, the ones with the wider magnet (P/N 54400048) appear to work with both old 
and new trigger units.” 

 

 
 

284

TRIGGER UNIT CONNECTORS:  Peter French claims to be one of the few people that still has reed switches in his 
car rather than the Hall effect sensor.  “I also had a good problem earlier with the 3-way connector to the injector 
switches.  The centre (0v) pin had two wires going to one pin and they were "over-crimped".  One wire had broken 
such that the reed switches had no common return unless the idle switch contact on the throttle switch was closed.  Net 
effect was that the car would start and idle fine but cut out as soon as the throttle was opened!” 

 

TRIGGER BOARD SCREWS:  As mentioned on page 139, Richard Dowling tapped out the holes in the distributor 
body to take M4 screws.  “My trigger board had 4 special rubber grommets with holes to take the screws.  I had no 
trouble with putting nylon M4 screws through.” 

 

MANIFOLD AIR PRESSURE (MAP) SENSOR:  One man’s pressure is another man’s vacuum.  Yes, the conditions 
within the intake manifolds on non-supercharged engines are commonly referred to as vacuum, but engineers normally 
refer to air pressures in terms of absolutes rather than relative to atmospheric, so the sensor measuring the air pressure 
within the intake manifolds is referred to as the “Manifold Air Pressure” sensor. 

Mike Morrin:  “I would strongly recommend that all owners of cars with this system periodially check the manifold 
vacuum sensor for internal leaks.  I stress this because when I needed a replacement for mine, I was told that new units 
are no longer available for the “early” calibration systems, and so I needed to find a good used one.  This resulted in me 
driving around London (I was officially in the UK on business) for 2 days sucking on manifold vacuum sensors at all 
sorts of dodgy Jaguar parts shops.  Of the dozen units I saw, only 5 did not leak, and only one of these was the correct 
calibration for my car.  There are obviously a lot of cars on the road with leaky sensors. 

“The sensor is the aluminium cylindrical thing right in the front right hand corner of the engine compartment.  The test 
is simple, disconnect the center hose from the tee piece between the manifolds (above the distributor) and suck on it as 
hard as you can, there should be no hissing sound from the sensor unit.  Then stick your tongue over the end of the pipe 
while releasing the vacuum in your mouth.  The hose should stick to your tongue for several seconds before it (the 
hose) drops off. 

“Internally the sensor has a barometer bellows, but there is an extra brass (?) diaphragm which provides the vacuum 
seal.  This diapragm flexes every time the manifold vacuum changes and it is not surprising that metal fatigue sets in 
and causes a split around the outer edge.  This is an example of Bosch design which would make Lucas look good (if 
Lucas had not insisted on sticking their name over the Bosch logo!). 

“If the sensor is leaking, then the engine will be running rich, particularly at part throttle, but this varies depending on 
the size of the split in the diapragm. 

“I did attempt to repair my faulty manifold sensor, both by adding a rubber diaphragm and by soldering up the split in 
the brass diaphragm, but both of these repairs resulted in a significantly lean mixture.” 

Roger Bywater of AJ6 Engineering (page 710):  “If there is a break in continuity to either winding on the pressure 
sensor the entire system will be dead.  Really all that most people can do is check the two windings for isolation and 
resistance (nominally 350 ohms between terminals 8,10 & 90 ohms between 7, 15) and the assembly as a whole for a 
vacuum leak.  Obviously that will cause excessive richness at light load whilst full load will be unaffected, although in 
practical terms a minor leak may not be all that important.  Electrical failure is extremely rare and I don't suppose the 
aneroid bellows fail very often, unlike the full load diaphragm on the European version which is prone to splitting 
around the edge (which can be repaired by careful soldering in the early stages.  These devices come apart quite easily 
once the 4 rivets are drilled out.  The pressure sensor being effectively a variable inductance transformer, function 
testing would entail firing pulses or low voltage AC through the primary winding and measuring what happens at the 
secondary.  In fact the most practical way to test one is to link it to an ECU on a bench rig representing the vehicle 
installation (such as we use) and measure injector pulses at different loads at a set speed.  Obviously not many people 
will be able to use either method.” 

 

 
 

285

D-JETRONIC REPLACEMENT:  Roger Bywater of AJ6 Engineering (page 710) says “it is not that bad a system and 
shouldn't be that troublesome really if it is properly set up (correct throttle balance is vital).  Although odd bits like 
trigger boards and pressure sensors do fail occasionally, and injectors tend to jam if exposed to water contamination, 
the ECUs don't often go wrong (unlike early HE 6CUs) and really the wiring and connectors are the main problem 
areas as they age.  It is a lot easier to go over the connectors, cutting back the wires and soldering them to the terminals, 
than to strip the lot out and install a new system.  

“If the system is beyond economic repair and really does need to be replaced then I would have thought that by far the 
easiest route would be to simply fit all the hardware, including manifolds, plumbing and Constant Energy ignition, from 
any HE engined car from 1981-88 and use a 16CU controller (1986-91) and have us install the correct program for the 
early flat head engine (Super Enhanced option available).  This setup has been used successfully on a number of V12 E 
types converted from carbs to EFI and has the great advantage that all the parts are familiar and easily available.  

“Aftermarket ECUs sound great but in practice involve a lot of time and effort to get right and would be hard to justify 
for a standard engine.  It would definitely involve rather more than "just wiring in a new processor" and any 
improvement in performance or fuel effcicency would be minimal. 

 

Cold Start Injectors 

 

SCHEMATIC CORRECTIONS:  In the Haynes manual 478/49015, you should mark Fig. 10.124 and 13.92 at the 
thermotime switch #298 to show a ground connection on the line at the bottom between the heater and the contact.  
Without this ground shown, it’s really difficult to figure out what’s going on! 

Also note that the thermotime switch contains a thermal switch and an electric heater.  The schematics make it look like 
a relay.  It should show a resistance rather than a coil. 

Note also that Section 19.22.32 of the ©1975 ROM shows a schematic of the cold start injector system that is poorly 
labelled at best.  It shows “12V” connected to 86C via a WR wire, when it would have been much clearer to mark the 
wire “from starter relay” -- there’s no power on it except when the starter is engaged.  The thermotime switch is at the 
bottom left, unlabelled and no innards shown.  The cold start injectors are at the bottom right, unlabelled and no innards 
shown.  The cold start relay is at top center, unlabelled with the innards shown incorrectly; since it shows five terminals 
including an 86C, it is evidently attempting to illustrate the system with a bootstrap circuit (see below), but there are no 
internal diodes shown and there is no internal connection to terminal 86.  Ken Gray and Roger Bywater claim the 
illustrations in the same section in other ROM’s such as the early XJ12 ROM show the non-bootstrap relay (four 
terminals) largely correctly. 

 

COLD START RELAY VARIATIONS:  Basically all of the schematics showing the cold start injector system on the 
D Jetronic EFI system show the cold start relay as a simple relay: one coil and one contact, 4 terminals.  If the car is 
really wired that way, the cold start injectors will only operate while the starter is engaged.  This makes good sense; 
once the starter is released, it can be presumed that it’s because the engine has fired and therefore there is no more need 
for the cold start injectors to be spraying.  In this configuration, the only reason the thermotime switch cuts off after a 
few seconds is to prevent flooding in case the owner holds the starter on for a long time. 

Ken Gray says the early cars were wired differently, though:  “In reality, (76 model XJC5.3 Oz spec & also 75 Daimler 
Double Six brit spec.) the relay is a five terminal relay.  It has two internal diodes, and the extra contact is labelled 86C. 
 Diode 1 goes from 86C to 86. This allows current to flow to the relay coil from the start relay ie when the engine is 
cranking.  Diode 2 goes from 87 to 86.  This allows current to flow into the relay coil from the main injection relay to 
allow the cold start injectors to stay energised after the starter has been dis-engaged.  Diode 1 blocks any current from 
feeding back to the starter solenoid once the starter key has been released.  The thermotime switch opens after it has 
timed out and the injectors are de-energised.” 

 
 

286

Basically, what Gray is saying is that the relay has a built-in bootstrap circuit; once it has been energized by the starter 
being engaged, it will hold itself energized even after the starter is released until the thermotime switch shuts it off.  If 
your car has such a relay, the schematics should be altered to look like Figure 14. 

 

Figure 14 - Alternative Cold Start Relay Schematic 

Interestingly, while the Jaguar cars with this system were reportedly connected this way, it would actually make more 
sense to connect the wire to the heater in the thermotime switch to 87 on the cold start relay instead of 86.  It would 
function exactly the same way, but there wouldn’t be as much current going through the diodes.  Roger Bywater found 
schematics of other cars with similar Bosch EFI systems that were wired that way. 

The diagrams that this schematic revision might apply to are as follows:  In the ROM ©1975, the “Fuel Injection 
Schematic” following Section 86, and whatever diagram shown in section 19.22.32, as discussed above.  In the Haynes 
manual 478/49015, Fig. 10.121, 10.124, 10.129, 13.92, and 13.96.  Check your own car, and if it has the relay 
described, mark the appropriate schematics accordingly. 

Gray:  “The relay is a Bosch manufacture with the Lucas sticker stuck on the side.  The Bosch number is 0 332 003 028 
- 12V, 10A.  The Lucas marking is 25RA which I think is the series (not serial) number, because all the relays that are 
in the same can configuration bear this number, and 33353A which I believe is the important part number.” 

Gray goes on to point out that the Haynes manual clearly shows the 5-terminal relay he’s talking about in Fig. 3.50.  He 
claims this relay is also illustrated in his XJ12 ROM; “My ROM is the red covered Jaguar XJ12 /Daimler Double Six 
version. Publication Part No. E190/4.”  Section 19.22.31 of the XJ-S ROM ©1975 shows the cold start relay, but not 
clearly enough to indicate how many terminals it has. 

If your car has the relay Gray describes, it will be a large metal can that looks similar to the starter relay with individual 
wires connected to spade terminals and it will have a terminal labelled 86C.  If your car has the simpler system 
illustrated in the schematics, it may have 5 terminals but one will be labelled 87a or some such; there will be no 86C. 

Roger Bywater isn’t thrilled with the bootstrap concept used here:  “I know what can happen when the cold start 
injector(s) are allowed to spray with the engine running.  A weak 'spit back' can ignite the fuel from the cold start