Jaguar XJ-S. Service manual — part 65

255

FUEL PUMP FOREIGN OBJECT DAMAGE: Steve Holst had a fuel pump totally seized by a tiny strip of
cellophane, possibly from a wrapper or maybe the edge of the Lucas label on the side of the pump. For those of us who
like to tinker, it’s possible to fix this problem by completely disassembling the pump and removing the cellophane strip
-- but just your luck it’d be a metal chip or a piece of sand and the impeller will be scored beyond repair. The message
is clear: make very sure nothing can get into that pump inlet except fuel! Fortunately, the line between the pickup
screen and the pump itself is very short and simple. Just make sure that there’s nothing within that little stretch of
plumbing and you’ll be OK.

FUEL PUMP VARIATIONS: Peter Smith says, “The pump motor is used on several cars including Volvo, but the
rotors are said to be different to provide different flows for different cars.” Just in case there is any misunderstanding,
let me make the implications here perfectly clear: the fuel pump is not an item that you should replace with a pump
from a different car just because it looks the same. If you install a pump with a lower flow rate, the car might run fine
at normal speeds but run lean when floored and the ECU is operating in open-loop fuelling mode. The result will be
burned pistons. There won’t be any warning prior to burned pistons.

FUEL PUMP UPGRADE: In 2005, the author picked up a flyer in an auto parts store touting a new line of Bosch fuel
pumps. Instead of the roller pump found in the XJ-S pump, this pump features a turbine pump. John Robison, who
works at Bosch, says “If I remember correctly, the impeller runs up around 30,000rpm.” Supposedly quieter,
supposedly more durable. And, judging by the illustrations, is about the same shape and size as the OEM Lucas pump.
So, if your pump dies, perhaps you’d be wise to seek out one of these newer design pumps. If it’s suitable for any EFI
automobile of similar horsepower and you can get it mounted and connected up, it should work fine. Bosch might even
provide a cross-reference listing new pumps for old cars.

FUEL COOLER: Many people see that the air conditioning circuit in the Jag includes a fuel cooler, and assume that
this is a high-performance trick. Dragsters often include an ice bucket in the engine compartment to cool the fuel on its
way to the engine to get more horsepower.

It’s a great idea, but unfortunately not the case. The fuel cooler in the Jag is in the return line going back to the tank.
The pump moves much more fuel than is normally needed and most of it recirculates. The fuel being heated while
passing through the engine compartment as well as the pumping energy itself would eventually heat up the fuel in the
tank significantly, possibly causing vapor lock problems and problems passing EPA emission requirements -- heated
fuel gives off more vapor. The cooler is to counteract the heating effect. This may help explain why the ventilation
system has the A/C running during most conditions.

Why Jaguar doesn’t put the cooler in the supply line and reap both benefits is anybody’s guess. Perhaps it’s more
difficult to make a cooler to operate at the higher fuel pressure on the supply side.

FUEL COOLER - CONVERTIBLE: According to Michael Neal: “This vapor lock problem was such a problem that
the convertibles were modified to keep the A/C compressor running all the time. The aerodynamics on the convertibles
caused the engine compartment to run even hotter than the coupes. With the A/C compressor running the fuel cooler
kept the fuel temp to a decent level.”

This author suspects aerodynamics had nothing to do with it. In the coupes, the A/C was always on anyway -- you
can’t live in there long without it. But with the convertible, you can turn the A/C off and put the top down.

256

1992-on Fuel System

FUEL COOLER - 1992-ON: Richard Mansell quotes from a Jaguar publication that was sent to him: “It is titled
“Technical Guide - The New XJ-S 1992 Model Year Preliminary Information”. In it there is a paragraph about the fuel
cooler deletion, it says:

“With the introduction of the in-tank fuel pump, the fuel cooler is deleted. This affected the characteristics of the
system so that a muffler is added in the line to reduce noise and vibration.”

“I wonder why adding an in-tank fuel pump should allow the deletion of the fuel cooler. It also says in another section:

“As a result of the deletion of the fuel cooler, it is necessary to introduce a redesigned hot start system. The thermal
vacuum valve fitted to pre-92 MY vehicles is replaced by the following components:

1. Fuel rail temperature switch. The electrically operated switch enables the fuel rail pressure to be increased when
the fuel temperature exceeds 70 degrees Centigrade at hot starting.

2. 45 second timer module. The timer limits the time for which the increased fuel pressure is applied.

3. Solenoid vacuum valve. This controls the vacuum signal driven by the fuel temperature switch and the 45
second timer.

4. Vacuum delay valve. This controls the way the extra pressure is applied at hot starting to give a decayed
reduction in fuel pressure over 45 seconds, after which the pressure is switched to normal.”

FUEL PUMP: John Goodman reports: “The latest XJ-S’s had pumps inside the fuel tank, so Jaguar may have done
this for noise related problems. I suspect that these pumps are lower output because the ’93 model XJR-S had twin in-
tank pumps, and must be ordered in a matched pair.”

FUEL PUMP WIRING: Nigel Sly, who owns a ’93 4.0, reports on a fuel pump wiring issue: “Pull the fuel tank and
remove the pump cannister/housing, then prise that apart, remove and replace the pump. Took me about 5 hours all in
all. Just the pump unit will cost about $100. Don't celebrate at that cost, when mine failed it was the connector which
had toasted, apparently a design fault. New pumps have a different connector, which will not fit the female connector
in the evaporative flange.”

“First, cut off the connector from the pump, and then solder on an extra 2 to 3 feet of wire to the pump cables, this gives
room to allow the pump to be refitted without the evap flange getting in the way. Pop out the through connector which
goes through the evap flange (black in colour), cut the right angled bit off the connector, and then drill a small hole
down through the connector, which will allow the pump cables to be passed through. Fit the connector back into the
evap flange, and fill the connector with mastic/sealant to seal the hole. You will probably need to remove the connector
from the external harness as well, unless you can find a matching connector to fit to the pump cables. If cannot find
matching connector, cut the connector from the external harness, and then fit matching connectors to the external
harness and the pump cables.”

“I tried for weeks to get the original pump connector, even tried the makers of the connector, only to be told it has been
out of production for years. Jaguar had the pump connector, and hence the evap flange changed in 1994, mainly I think
because of numerous failures of the original connector.

“Incidentally, the tank does not need to be completely removed from the car; it needs to be moved back about a foot in
order to get access to the evap. flange.

“A replacement pump (in canister), complete with new redesigned evap. flange, and new harness is about $900, but
now as rare as rocking horse s**t, you might see them listed on various sites, but stock is zero.”

Sly may be satisfied with his fix of running the wires through the flange and sealing them up, but in general this type of
thing is not a good idea. It’s possible to end up with leaks through the wires themselves, vapors or liquid passing
between the copper conductors and the insulation. In general, this type of problem calls for a liquid-tight bulkhead
connection, something that the pump plugs into on the inside and the harness connects to on the outside. Better

257

electrical supply houses should be able to provide such connectors, but make sure they’re compatible with fuel and
ethanol. If all else fails, you could conceivably make such a thing by embedding straight pieces of solid brass in the
sealant and then connecting the leads to them on the inside and outside. You might even use connectors to hold the
brass pieces in place while the sealant sets up.

Alternatively, you could make one small hole per conductor, and install a small bolt and nut with insulating washers.
You must make sure that the head of the bolt seals against the insulating washer, otherwise you will get a leak along the
bolt threads. It may prove a challenge to get the electrical leads connected without causing a leak.

FUEL PUMP PLUGGING: Noel Keefer reports: “I was a 1000 miles from home and the fuel pump plugged up with
crap in my '92 with only 42,000 on the odometer!! Paid big money to get it repaired. New pump including plastic can
and o'nite shipping was $700!!

“Pried open the can and guess what? It had been previously changed. The pump in my low mileage '92 convertible
was a WALBRO fuel pump. It had a small fine screen filter and the outside filter never got a chance at this small stuff
in that fine filter. My new Jag pump had a somewhat larger mesh filter. I cut it out completely. There are 2 filters
feeding the canister before it gets to the pump. Give the outer filter a chance, or remove it!

“But a WALBRO pump ought to be available at NAPA, for cheap. It was a perfect fit.”

Rory McKenzie reports: “My friend Mike Dallaire ('92 xjs v-12) told me his in-tank pump fried out and he had to do
similar work. He said when he pulled the in-tank fuel filter out, it was all clogged up with the remnants of little
polyethylene ''packing pillows'' which had not been completely removed between the shipping and installation phases.
He also quoted a fuel pump price of around $900 to $1000!”

Vapor Recovery System

VAPOR RECOVERY SYSTEM: When the car is running and drawing fuel out of the tank, there must be a vent
system to allow air into the tank to prevent a vacuum from being created. It is no longer acceptable to merely have a
small hole in the gas cap; such methods allow fuel vapor to escape into the atmosphere all the time, whether the car is
running or not, and would contribute to air pollution. Raw gasoline is a serious air pollutant, and besides that it smells
bad.

Unless you live in California or somewhere else where they have intelligently-designed gasoline pumps, you pump 20
gallons of fuel vapor out into the air every time you pump 20 gallons of liquid gasoline into your car. It is simply
inexcusable that, decades after the public became aware of the problems of air pollution, every filling station in
America doesn’t have the same vapor recovery filling systems that California has. But, back to discussing the fuel tank
vent system in your Jag.

Avoiding polluting the air via a properly-designed fuel tank vent scheme is easy and doesn’t hamper the performance of
the car one iota. The only disadvantage to the owner is in the initial cost of installation, but since the car comes with the
system from the factory and has therefore already been paid for, there is no reason to remove or disable it.

In simplest terms, a “vapor recovery system” connects the fuel tank vent to a carbon canister. Air can flow freely
through the canister into the fuel tank, but when vapors from the fuel tank try to escape through the canister they are
absorbed by the activated charcoal. Of course, the charcoal can only absorb so much fuel. Therefore, whenever the
engine is running, there is a system by which the engine draws fresh air through the canister. This draws the vapors
back out of the charcoal and burns them in the engine.

When the car is not run for extended periods of time, the amount of vapor generated in the tank could be considerable.
Gradual changes in ambient temperature and barometric pressure would cause the vent system to “breathe”, running a
large amount of vapors through the canister. To absorb all this vapor, the canister would have to be prohibitively large.

258

To reduce the amount of vapor the canister has to deal with, there is a “2 psi valve” in the vent line between the fuel
tank and the canister. This valve will allow air to flow in either direction, but only after a certain pressure or vacuum
has been reached. If the contents of the tank expand and try to escape out the vent, it will prevent any flow until the
pressure reaches 2 psi or so, and then it will allow it to pass to prevent damage to the tank or hoses. Similarly, if the
contents of the tank contract and try to draw air in through the vent, the valve will prevent any flow until the vacuum
reaches a set value -- usually lower than the pressure value -- and then it will allow it to pass to prevent the tank from
collapsing. As a result, the vast majority of the smaller expansions and contractions do not open the valve at all, and the
amount of vapor the canister must absorb is greatly reduced.

Of course, this means that at any given time the fuel tank and hoses may be slightly pressurized, even when the car is
not running. Even the tiniest leak anywhere in the system may become quite significant.

I am sorry to report that the vapor recovery system on the XJ-S is unreliable. Many owners have reported problems,
usually indicated by a large “whoosh” but often dismissed as normal until the gas tank cracked or other dangerous
symptoms arose. Fortunately, if the symptoms are recognized and addressed promptly, it’s never difficult to make the
system work as it should; it’s simply not that complicated or expensive to fix.

VAPOR RECOVERY SYSTEM -- AFTER RECALL: The original XJ-S vapor recovery system pretty well followed
the description above. However, there was a tendency for the tank to develop stress cracks from excessive pressure
fluctuations as well as indications it did not vent as well as intended (including numerous complaints about fuel odors),
so there was a recall to address the system. In general, the recall involved the installation of a “Rochester valve” in
place of the 2 psi valve. David M. Johnson: “The Rochester valve (Part CBC 7714) is a cylindrical disk (approx 2 inch
diameter) with an inlet from the fuel tank vent pipe and outlet for the vent to connect to the carbon canister. A third
outlet at the top is to connect to the manifold vacuum.” A Rochester valve works the same way as the 2 psi valve when
the engine is off, but when the engine is running the vacuum signal pulls the Rochester valve wide open to freely vent
the tank.

The recall also installed one or two (depending on VIN) vacuum-operated purge valves to provide a positive purge
when the engine is running. The left side throttle butterfly housing was replaced with one that had suitable taps for
controlling the Rochester valve and purge valves.

For those who have disconnected the vacuum lines from the butterfly housing and want to reconnect them properly, I
can give this guidance based on the dealer instructions for the recall that installed the system in the first place: Among
the two vacuum hoses that go from the butterfly housing to the area in front of the left wheel well, one is supposed to
have a black and white delay valve in it, hanging in the vicinity of the coolant header tank. A hose from the black side
of this delay valve should be connected to the front port on top of the left butterfly housing. The hose with nothing in it
should be connected to the rear port on top of the butterfly housing. Of course, there were several variations on this
system as applied to different cars, but I think that description should cover most of them.

The carbon canister, the Rochester valve, the purge valves, and the associated plumbing are all located forward of the
left front wheel; remove the forward panel of the inside wheel well to inspect. The only part of the vapor recovery
system not in this location is the vapor separator, which is high in the right side roof pillar on the coupe.

I am sorry to report that, after the recall has been applied, the vapor recovery system in the XJ-S is still unreliable. The
installation of the Rochester valve sounds good in theory, but unfortunately the Rochester valve itself is apparently
extremely unreliable. When it fails, it apparently fails jammed shut, which is the worst thing that can happen to a vapor
recovery system.

CARBON CANISTER OVERBOARD VENT: When the recall was applied to the vapor recovery system, one change
done was to remove a hose that vented the carbon canister out the bottom of the car. It then just opens into the
bodywork cavity. This demonstrates an interesting attitude; of course, if the system is working as it should, there
shouldn’t be any odors coming out of this opening. But if, on the off chance, some gasoline vapors make it through this
canister, you might be happier if you installed a new hose venting it overboard.