Jaguar XJ-S. Manual - part 76

 
 

299

 And you can provide suitable holes and slots in the plate to securely support the pot and allow adjustment of the idle 
position.  In fact, it wouldn’t be too difficult to apply this method to installing the later Jaguar pot EAC9634 in case you 
want to save the cost of their adapter, and it would probably result in a more professional-looking installation! 

Roger Bywater of AJ6 Engineering reports, “Some years back when we needed a clean throttle signal we used a 5K 
wire wound pot on a bracket attached to the LH air filter ahead of the throttle and connected onto the end of the throttle 
spindle by a short length of 5/16 bore hose and clips.  Might be a cheap solution to renewal of original.” 

That idea of using fuel hose as a coupling is a good one -- regardless of where you mount the pot.  A pot doesn’t take 
much torque to turn, so the friction in a hose coupling will easily drive it.  It doesn’t matter if the shaft has a flat or a pin 
hole or whatever, just slide a length of hose over it and clamp.  And the flex of the hose will correct for misalignments. 

 

THROTTLE POTENTIOMETER UPGRADE:  Matt Emmons thinks that using a potentiometer here is a bad idea due 
to inherent wear problems on the wiper.  He thinks a Hall Effect sensor would be more reliable because it has no 
wipers; it works via a magnet moving over a circuit. 

Yeah, right.  Where are you going to find a Hall Effect sensor that will go in here?  And who’s going to figure out how 
to connect it to the Digital P ECU so it works right? 

Folks, you simply have to see this to believe it.  It’s very difficult to find a generic potentiometer that can be adapted to 
the Jaguar throttle linkage, but it’s no problem at all to find a Hall Effect sensor that will fit perfectly!  Emmons 
provides a few web sites: 

 

http://www.electrocorp.com/o1-4.html

 

 

http://content.honeywell.com/sensing/prodinfo/solidstate/

 

 

http://www.nasatech.com/Briefs/Oct99/MOC10991.html

 

 

http://www.clarostat.com/NEWhalleffect.htm

 

 

http://www.wabashtech.com/throttle.htm http://www.beiduncan.com/html/news/

 

 

http://www.delphiauto.com/pdf/intellek/rotary_sensor.pdf

 

 

http://www.aecsensors.com/sensors/noncont.htm

 

 

http://www.bourns.com/bournsautomotive/tps.html

 

Note that last one is actually Bourns.  “There are more. The Honeywell site was particularly informative about Hall 
Effect devices in general. 

“Most of the manufacturers would not sell them to me in such small numbers, but the one I got is made by Clarostat.  
It's called HRS100 Hall Effect rotary position sensor and I bought it from Allied Electronics.  Here's a link to their 
catalog page with details of the device: 

 

http://www.alliedelec.com/catalog/pf.asp?FN=630.pdf

 

Connecting it up?  Believe it or not, you just connect it up.  While the pot took the 0 and +5V leads from the ECU and 
split off a voltage from somewhere in between, the Hall Effect sensors use the 0 and +5V to power the electronic 
circuitry and generate a signal between 0 and 5V depending on position. 

What’s more, Hall Effect sensors are apparently cheap -- cheaper than the OEM pot. 

It would not surprise this author to learn that some of the Ford and GM TPS’s described above are actually Hall Effect 
devices.  Some of them don’t feel as though there’s a wiper involved when you have one in your hand. 

 

THROTTLE POTENTIOMETER -- LATER CARS:  Charlie Fritz says, “the 95 XJS has a double potentiometer in the 
TPS with one signal for the transmission and the second for the ECU as before.  It has six wires as in the attachment, 
but I don't know which goes where or how to set it.” 

 
 

300

 

HESITATION:  Mark Sampson had a bad battery in his car.  “I was also experiencing some hesitation when warm, it 
has also gone away with the new battery.  Interesting.” 

Don Snyder:  “My car would occasionally hesitate, and sometimes die, when accelerating while the engine was hot.  I 
replaced the oxygen sensors and added a bottle of injector cleaner, and have not had the problem since then.” 

 

INJECTOR RESISTOR PACK CONNECTIONS:  The resistor pack for the injectors seems to be fairly reliable, but 
unfortunately the wiring to it is typical Lucas.  Tom Amlie, who has a Ph.D. in Electrical Engineering and entirely too 
much time on his hands, relates an interesting failure mode that kept his car from starting:  “Put an oscilloscope on an 
injector.  Instead of a clean negative-going pulse, it oscillated like crazy.  Took the ECU apart and traced the circuits 
until I understood how it works.  There are four power transistors, two for each bank.  The "ON" transistor slams the 
injectors on with full voltage to overcome the inductance.  The emitters of the "ON" and "HOLD" transistors are tied 
together and go to a 0.2 ohm power resistor to ground.  When the current to the injectors reaches about 3.5 Amperes, 
the voltage across this resistor is fed back to a circuit which cuts off the "ON" transistor.  Then (in theory) the current 
through the "HOLD" transistor is adequate to keep this voltage to about 0.6 volts and keep the "ON" transistor off.  All 
of this depends on proper resistances and tight connections. 

“Get real!  This stuff says "Lucas" all over it!  The four precision resistors in the pretty box on the right side just aft of 
the radiator are connected with an 8-pin plug with round pins, far inferior to the American practice of using rectangular 
blades with much larger contact area and heavier contact pressure.  I must admit that the box is a pretty casting and 
nicely finished.  The "Lucas" decal must be worth a lot, but it is a piece of crap. 

“What happened is obvious:  The "ON" transistor did its thing and turned itself off.  The "HOLD" transistor couldn’t 
pull enough current because the resistance was too high.  The "ON" transistor comes back on, switches itself off, the 
cycle continues, and no fuel flows. 

“All is not lost.  Clean the contacts and use a very sharp X-Acto blade to spread the split pins a thousandth or two.  
Don’t overdo it and don’t cut yourself. 

“Wonder how many ECU’s have been rebuilt because of this.  To be fair, the workmanship in the ECU is superb.  Must 
be made by Bosch.” 

 

EFI WIRING CHECK:  Tom Amlie goes on to suggest a general check of the Digital P wiring:  “Disconnect the ECU 
and measure every pin going back into the car with a VOM.  Using a long wire with a small alligator clip to reach to the 
front of the car you can measure the continuity of every wire in the bundle.  In particular, make sure that the pins that 
are supposed to be ground are really ground.  A little corrosion can cause a lot of misery.  The pins for the "HOLD" 
transistor should measure close to 7.2 ohms referenced to the power pin on the main relay which is also in the boot.  
You can also easily tell if the injector harness or one of the injectors is open by measuring from the "ON" pins to the 
same relay.” 

 

AIR TEMPERATURE SENSOR:  In the ©1975 ROM, page 19.00.07 and section 19.22.23 show charts of air 
temperature vs. ohms for the purpose of checking the sensor mounted on the left air filter housing.  These charts apply 
to D Jetronic systems; if you have a Digital P system, these charts are not valid.  The correct resistance for the Digital P 
air temperature sensor, part number 73197, at room temperature is around 2800Ω. 

 

COOLANT TEMPERATURE SENSOR:  As opposed to the air temperature sensor discussed above, apparently the 
charts of coolant temperature vs. ohms shown in the ©1975 ROM, page 19.00.08 and section 19.22.19, apply equally 
well to both D Jetronic and Digital P coolant temperature sensors. 

 
 

301

Ian Hooton:  “If the car doesn't want to start, or when it does there is very little power with the throttle wide open and 
there is a strong smell of petrol, check the coolant temperature sensor on the left hand bank, just above the stat housing. 
 Remove the connector and bridge the plug with a piece of wire, try and start the engine, you will find it will probably 
start and run fine. Usually all that is wrong is that the contacts are corroded, clean them up and everything should be ok. 
 This happened to me late last year, and took quite a time to find the culprit. It would appear that the ECU thinks the 
engine is constantly cold and over compensates by increasing the richness of the fuel/air mixture.” 

According to the Jaguar literature, the coolant temperature sensor provides a “minor” trim to the fuel schedules.  
However, when the sensor fails open circuit or the wires to the sensor are broken or disconnected, the trim turns out to 
be major indeed.  An infinite resistance tells the EFI system that the coolant temperature is a zillion degrees below zero, 
so it responds by making it overly rich.  Not just a little bit overly rich; the car may not even start, and if it does start it’ll 
run very badly and pour out clouds of smoke.  And it just gets worse as it warms up, if you can keep it running that 
long.  David Johnson:  “When I had a bad wire going to my sensor the smoke was so bad I swore I blew the head 
gasket.  In a matter of 20 seconds there was so much smoke that I thought I had a fire!” 

An open-circuit coolant temperature sensor failure happened to this author, and he was able to diagnose it quickly using 
his own book!  One thing that caused some consternation, though, was the fact that the smoke was a light gray color; as 
any good mechanic knows, running too rich causes black smoke.  Johnson:  “When I took automotive theory (years 
ago) the instructor told us white "smoke" meant water.  Black smoke meant gas and blue smoke meant oil.  I also 
remember him telling us that on a really rich mixture the color is grey.  Hard to tell from "white" but it was because the 
engine is running almost as a fogger.” 

One would think this would still be obvious due to the smell of unburnt fuel, but that wasn’t apparent either.  Johnson:  
“The smoke had almost no smell.”  The smoke from this author’s car had an odor, but it wasn’t strong and not clearly 
raw gasoline.  The only plausible explanation so far is that the ignition within the cylinders burns the most volatile 
compounds most readily, pumping the less smelly fractions of the gasoline out as an unburnt mist. 

Just the same, it probably wouldn’t be a good idea to be smoking around the car.  If the air pump is diverting (the 
engine is warm), there’s no chance of fire within the exhaust system or catalytic convertors simply because there’s no 
oxygen in there; it was all used up in the cylinder.  But once the smoke clears the end of the exhaust pipe, there’s no 
shortage of oxygen in the surrounding air.  Things could look like the Hindenberg disaster if you get careless. 

If the air pump is operating, it would make sense that the symptoms might be different.  The engine wouldn’t run any 
better, but the dense smoke could be replaced with a red-hot exhaust system or catalytic convertors.  Just in theory, 
mind you; there are no reports confirming this actually happening to anybody. 

The coolant temperature sensor and wiring also suffers intermittent failures, with some Jaguar owners actually 
monitoring readings while driving to confirm that the problems were caused by this sensor.  One of the connectors 
should give ~2.4 volts with the ignition on and the engine cold (the other one gives 0 volts) and gradually decrease to 
~0.95 volts at normal operating temperature.  Sudden changes in voltage mean the sensor is breaking up. 

Michael Neal:  “Don’t use Lucas brand.  Get a Bosch part, cheaper and better.”  Good luck figuring out which Bosch 
part to get.  The following part numbers have been suggested as being usable: 

 

73198 -- original Jag part, 1979-89 

 

SNB 802 -- often listed as a superceding number for 73198 

 

0 280 130 023 -- what XK’s Unlimited ships when you order an SNB 802 

 

EAC3927 -- what the 1987 Parts Catalogue specifies for all engines 

 

DAC4737 -- Jag part, 1989-on -- reportedly offered as replacement for 73198 on earlier cars 

 

0 280 130 055 -- reportedly an exact replacement, but available only in Germany and very expensive 

 

0 280 130 026 -- apparently differs slightly in tolerance, but definitely works well 

The sensor should cost about $30.  If someone tries to charge you over $100, it’s time to look elsewhere. 

 
 

302

Note that the D-Jetronic system uses sensor 73142, and it appears this is not the same critter, not interchangeable with 
the later sensor. 

 

MANIFOLD AIR PRESSURE (MAP) SENSOR (Digital P):  The MAP sensor for the Digital P system is within the 
ECU in the trunk, so a vacuum line is run the length of the car.  On page 343 of Haynes manual 478/49015 it states, 
“Commencing with... vehicle identification number XJS 107102, a vacuum reservoir has been located adjacent to the 
ECU to eliminate resonance from the inlet manifold vacuum sensor pipe”.  This reservoir is a plastic chamber about an 
inch in diameter and three inches long, fitted in the line just before it connects to the ECU.  Would installation of a 
similar reservoir help earlier Digital P cars?  Who knows? 

 

ACCEL PROBLEMS:  Vince Chrzanowski reports that his car wouldn’t start accelerating until a couple seconds after 
he floored the accelerator, even if the tranny kicks down.  Otherwise, the car seemed to run fine.  “The vacuum line to 
the ECU in the trunk was partially blocked.  A shot of compressed air cleared it out, and everything went back to 
normal.” 

 

RICH RUNNING:  Richard Chapman sends this tip:  “If you're getting rich running and fouling check to see if the 
pressure regulators mounted on both ends of the fuel rail are leaking fuel into the manifold via the vacuum line which 
connects between the reg and the manifold.  With the engine idling pull off the hose at the reg, the engine speed will 
increase with the extra air entering the manifold through this pipe so block it with something.  After running the engine 
for a minute or so you may notice the presence of some petrol coming from the regulator vacuum connection.  If so 
replace it.  They are not the same unit despite similar looks.” 

 

INJECTION SIGNAL FAILURE:  Chrzanowski sends this report on a problem with his car:  “It seems that the engine 
revs information is sent to the ECU from an amplifier unit atop the engine.  The signal is sent via a thin shielded wire.  
Unfortunately, the wire is routed near some very hot areas and that heat melted the insulation between the center 
conductor and the shield, shorting the signal to ground.  Replacing the defective section of wire restored the signal.  In 
talking with some other local Jag owners, I’ve found that this is not an unusual problem.” 

It’s not; in fact, this author had the same problem.  One of the W/S wires from the Lucas ignition amplifier rearwards 
towards the firewall connects to what appears to be a solid white wire but then disappears into a harness wrapping in 
which it is the only wire.  This white wire is actually a coaxial wire, a tiny conductor surrounded by insulation, a 
grounded shield, and more insulation.  Within that harness wrapping, the insulation gets brittle and starts cracking, and 
the outer layer of insulation being cracked only encourages the inner layer to crack at the same spot.  Eventually, the 
thin conductor begins to short out against its own shield -- often intermittently.  Sometimes you can wiggle that wire 
and make the problem come and go.  All the while, the harness wrapping hides the wire’s condition from view. 

Since there is a resistor inside the ignition amp to limit the effect of such shorts, the ignition system continues to work, 
but the injectors don’t fire since the signal isn’t making it to the EFI ECU.  That’s why the problem is listed here in the 
EFI section; you can review the related ignition system issues on page 153. 

If you suspect trouble in this wire, you might unplug this connector from the ignition amp and the big connector at the 
ECU in the trunk and check between pins 18 and 23 for continuity, which would indicate a short.  However, since the 
short is only too likely to be intermittent and therefore look perfect while you have a VOM in your hand, you probably 
should go ahead and peel that wrapping off this wire near its connection to the amp.  If you find the wire inside cracked 
and brittle, keep removing the wrapping until you get to non-brittle wire, then cut it off and splice on a length of new 
shielded wire.  It doesn’t need to be as skinny as the original stuff, but it’d be nice if it was more heat-resistant.  Don’t 
ground the shield; it is grounded at the ECU, and grounding at two different places can cause ground loop problems 
wherein the loop picks up interference and then transmits it into the conductor it’s supposed to be shielding.