Land Rover Fuel Injection System Manual

							
LAND ROVER 
 
Throttle linkage
The arrangement of placing the plenum chamber with its air intake toward\
s the drivers side meant 
that it was possible to reuse the existing throttle cable. 
Figure 23 – Throttle fabrication
 
However, the existing throttle assembly steel frame fouled the intake me\
tal air pipe running from 
the plenum to the mass air flow sensor assembly and so had to be discard\
ed. The throttle 
assembly was cut back, and welded strengtheners were added to the existi\
ng frame. A bracket was 
then constructed (see below) and welded to the assembly to accommodate\
 the throttle cable (the 
mounting bolts were then removed and the holes filed with weld).
Figure 24 – Throttle cable mount bracket
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (21 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
 
The arrangement permits ingress of the fine throttle cable through the s\
lot, at which point the 
whole throttle cable assembly can be moved forward to permit the mount s\
crew thread to enter 
the main hole shown above. Once in position, the slot does not permit th\
e throttle mount screw to 
exit. 
 
The throttle linkages on the plenum are sprung at wide open throttle whe\
n the back stop is 
reached. In other words when the throttle blade is wide open, the thrott\
le actuator can be opened 
slightly wider (by about 5 degrees of rotation). The cable was adjuste\
d so that that springing state 
is only just engaged – which means that WOT is possible and accommoda\
ted accurately by the 
linkage system. 
 
Mass air flow sensor and air filter
The mass air flow sensor is the electronic unit used to measure the prec\
ise amount of air entering 
the engine. It does this by using two wires in a balanced bridge circuit\
. One wire is exposed to the 
air flow entering the engine, and the other is encased and hidden. As ai\
r flow passes over the 
exposed wire, it is cooled, and its resistance alters. The bridge circui\
t measures the difference 
between the two wires to determine relative air flow.
 
Figure 25 – Mass air flow sensor schematic
 
Filtered air enters the MAFS unit which is cabled into the injection loo\
m via a four wire connector. A 
hose connects the outflow side of the MAFS unit directly to the plenum i\
ntake. 
 
There is one interesting oddity about the design – which is that the \
diameter of the MAFS outflow is 
different to that of the plenum inflow. The MAFS unit is 80mm OD, wherea\
s the plenum is 72mm 
OD – consequently the standard land rover plenum flexible hose used t\
o make the connection has 
two different sized ends. 80mm diameter pipe is fairly easy to source, b\
ut 72mm is difficult to 
obtain and so this aspect of the design, coupled with the rather tight a\
lmost-but-not-quite C shape 
and long length has been the most problematic to resolve – all the mo\
re so given that it is vital 
that the hose connection linking the MAFS to the plenum is air tight in \
order to ensure that all air 
entering the engine is metered by the MAFS unit. 
 
The solution (which passed through three different unsatisfactory desig\
ns) consisted of the 
following. First, a 20mm deep stainless steel sleeve was manufactured wi\
th a 72.1mm inside 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (22 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
diameter, and 80mm outside diameter. The sleeve was designed to be an in\
terference fit on the 
72mm collar of the plenum chamber. Silicon sealant was then applied to t\
he inside of the sleeve 
and it was then driven carefully onto the plenum – and left to set fo\
r 24 hours.
 
Two brushed aluminium pipes with 80mm outside diameters were then purcha\
sed, one a right 
angle, and one a 45 degree. Both had been bent using a mandrill press re\
sulting in a smooth inner 
radius. These two pipes were then cut, dressed and smoothed, and joined \
to the plenum, at the 
centre and to the MAFS in three places using silicon 80mm ID soft pipe a\
nd large jubilee clips. 
 
One complication to this solution was that the 90 degree radius pipe whe\
n fitted fouled the throttle 
assembly frame. In order to resolve this, the entire throttle had to be \
disassembled, cut and 
welded to permit adequate clearance while ensuring it was strong enough \
not to flex.
 
Figure 26 – MAFS air path
 
With the main air intake pipe routing fitted and with clearance ensured \
all round the intake pipe 
(especially near the power steering reservoir canister – allowing fo\
r engine torque motion) the 
mount for the MAFS box and air filter was then created.
 
Figure 27 – MAFS mounting bracket
 
There are four bolt holes in the MAFS mounting bracket. Two bolts mount \
the bracket firmly on to 
the adjustment swing arm used to hold the alternator. An additional two \
bolts are used to mount 
the MAFS air filter case to the bracket. It is important to realise that\
 there are two levels of 
adjustment deliberately built into this bracket. The mount holes for the\
 MAFS air filter case are 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (23 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
slotted – permitting the entire MAFS assembly to move. In addition, t\
he left hand curved mount 
bolt hole (see above – closest bolt to alternator) is also slotted,\
 whereas the right hand lower bolt 
hole is fixed. That means that the entire assembly can rotate around the\
 right hand lower bolt. The 
combination of the two adjustments allows the MAFS position to meet the \
intake pipe, while coping 
with minor adjustment changes to the alternator position. 
 
During the fabrication phase, some time was spent reviewing adequate air\
 filtration for the MAFS 
unit with an investigation of K&N filters and other third party units. H\
owever, the manufacturers 
filter case has the significant advantage of being equipped with a stron\
g MAFS coupling which any 
third party filter would not. In addition the land rover air filters are\
 very much cheaper than after 
market dress up models, and supply more air flow. 
 
The donor vehicle air filter unit was a large metal canister using a sol\
id clip coupling to the MAFS 
and a lid with nozzle for the raw air intake. There was interestingly no\
 air preheating feature. The 
outer canister steel was badly rusted on the donor parts even though the\
 base was in good 
condition. An elegant solution involved grinding off the outer canister \
welds leaving a base with 
proper clip couplings to the MAFS and an elongated bolt fixing mechanism\
 for the air filter (a 
standard range rover filter). 
 
 
 
Figure 28 – Mounted air filter (outer canister removed) clipped to \
MAFS
 
Positive crank case ventilation (PCV)
The standard PCV system is designed to draw un-burnt hydrocarbons result\
ing from blow-by past 
the piston rings, into the combustion process. This process removes hydr\
ocarbons from the engine 
where they would otherwise reduce lubrication efficiency increasing wear\
 (and emissions from the 
engine) while also creating a mild depression inside the engine to assi\
st the gasket sealing process. 
 
The original carburettor PCV system employed a filter (actually a plast\
ic fuel filter) free standing 
and connected via a hose to a metal pipe on the rear of the engine valle\
y. Additional hoses 
connected half inch ports on both rocker covers separately to a pair of \
flame traps and from there 
to vacuum ports on each carburettor. This design applied engine vacuum t\
o both rocker covers – 
while drawing fresh air into the crankcase via the small plastic filter.\
 
 
The design may have been adequate for a 1970’s vehicle, but it is all\
 but unusable for a fuel 
injected engine. PCV acts as a secondary source of air entering the syst\
em, and so considerable 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (24 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
care is required to meter the precise amount of air it passes, while ens\
uring that it functions as a 
self contained system of crankcase ventilation. 
 
The standard PCV system designed by Rover for a fuel injected engine req\
uires the use of different 
rocker covers (a new set had to be sourced from eBay). The key feature\
s of the new rocker covers 
are a built in PCV breather metered orifice (see (c) below), a screw\
 thread capable of 
accommodating the oil separator (see (D) below), and a separate scre\
w thread fixed in a raised 
tower for oil entry. A schematic of the final PCV plumbing is shown belo\
w.
 
Note that the oil entry screw thread on the passenger side rocker cover \
isn’t shown and that the 
original pipe on the back of the valley is blocked off and sealed.  
Figure 29 – PCV schematic
 
The passenger side rocker cover is fitted with a fixed small round metal\
 T piece raised above the 
cover by perhaps 1cm (see (C) above). The stem of this T is drilled \
with a small 0.5mm hole 
passing into the inside of the rocker cover. A plastic shroud encases a \
gauze filter which clips onto 
the top of the metal T piece. Metered airflow draws into the engine via \
this filtered path and from 
there passes through the head into the main crankcase where it combines \
with oil and any 
hydrocarbons. It is then drawn out of the driver side rocker cover via a\
 screw-in oil separator (see 
(D) above) which as the name suggests allows oil in the air flow to d\
rain back into the engine while 
also acting as a flame trap in case of backfire. Airflow from the oil se\
parator is drawn into the 
plenum via a T shaped plastic connector (supplied by Rover) and hose p\
ipes.
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (25 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
Figure 30 – Rover PCV coupling T piece
 
The plastic T piece has a third metered orifice (a 3mm hole) which is \
connected to an engine 
vacuum port on the plenum, via hose pipe. This path is designed to provi\
de a metered amount of 
vacuum to the PCV system when the engine is idling, with the throttle sh\
ut. It is worth noting that 
the amount of air flow is substantial, even with this metering orifice –\
 a fact which has a bearing 
on the engine idle speed as the PCV system behaves as an alternate sourc\
e of air flow into the 
engine at idle. The implementation includes an additional copper pipe in\
 the path (see (A) in the 
PCV schematic figure) which further reduces the air flow via a 2mm dril\
led solder restriction. 
Figure 31 – View of the PCV system pipe routing
The fuel system
Fuel injection relies on high pressure fuel being delivered to the injec\
tor rail under all conditions. 
When the injectors open, the high pressure fuel instantly atomises into \
a fine spray and this effect 
greatly eases the traditional problem of vaporising fuel prior to combus\
tion. 
 
The 14CUX system requires the following fuel circuit
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (26 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
 
Figure 32 – Fuel system schematic
 
14CUX production implementations generally employed wet fuel pumps mount\
ed inside the fuel 
tank along with non-serviceable gauze intake filters. The pump pressuris\
es fuel to about 3 bar and 
delivers it to an inline replaceable filter and onto the fuel rail – \
connecting to all 8 injectors. At the 
distant end of the fuel rail, the pipe connects to the inlet of a fuel p\
ressure regulator designed to 
regulate the pressure between 24 and 36 psi, relative to engine vacuum (\
the higher pressure 
obtained when vacuum is lowest). The outlet of that fuel regulator then\
 feeds to a pipe designed to 
return excess fuel to the tank at relatively low pressure. 
 
The land rover solution adopted is essentially identical – except tha\
t a dry inline Jaguar XJ6 fuel 
pump rated at 3 bar is mounted on the mid horizontal chassis rail along \
with a high pressure fuel 
filter (taken from a Katterham Seven vehicle). 
 
The plumbing around the pump is worthy of note due to the odd mix of pip\
e sizes.
 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (27 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
 
Figure 33 – Fuel pump and plumbing schematic
 
The fuel circuit was plumbed using predominantly 8mm fuel hose (rated t\
o DIN standards for fuel 
injection). A small amount of 12mm hose was used to couple the inlet of\
 the fuel pump.
 
Where the fuel line runs up the firewall to the rear of the engine, it c\
onnects to an aluminium block 
threaded with three holes – two designed to accept 8mm hose ends, and\
 a third tapped to suit a 
1/8
th NPT non-permanent fuel gauge. After testing this gel filled gauge was r\
emoved and replaced 
with a threaded block plug. 
 
 
Figure 34 – Fuel pressure tap off point
 
 
 
 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (28 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
The fuel tank
During the early part of August 2010, the relatively new fuel pump faile\
d as a result of an electrical 
fault (not linked to any plumbing problem). The pump was replaced (an\
d a second spare pump 
purchased at the same time).
 
Later that month, a series of odd running problems hit at more or less t\
he same time – but which 
were clearly linked to the recent rework of the induction intake routing\
. There had been a 
noticeable smell of fuel in the cab – which couldn’t be traced and\
 which lasted about a week. At the 
tail end of that week a series of intermittent misfiring problems occurr\
ed on part throttle cruising 
during a long drive into town – a condition which would appear and ju\
st as quickly vanish. During 
low speed parking manoeuvres the engine exhibited surging and stumbling \
with revs dipping below 
500rpm. The 14CUX reported error code 48. 
 
Based on the smell of fuel, and the surging (and code 48) it seemed re\
asonable to assess the fuel 
system for problems. 
 
The one area of the system which hadn’t been closely assessed was the\
 fuel connections to the 
tank itself – so the panel under the driver seat was removed to gain \
access. It was immediately 
obvious that at least some degree of fuel was leaking from the tank seal\
s – probably when the tank 
was filled to absolute capacity. 
 
There are five pipe connections to the tank. A large fuel filler pipe co\
upled with a smaller breather/
vent hose connect to the tank just behind the driver chair back. Closer \
to the front of the vehicle 
there are three brass units screwed to the tank and sealed with rubber g\
askets. The first is the fuel 
level sender unit. The second is the return fuel line brass housing, and\
 the third is the flow fuel line 
brass housing. 
 
All three units were removed and cleaned – as was the surface of the \
tank. The return brass fuel 
pipe housing, and the electrical sender were immediately refitted using \
silicon gasket sealant on 
both sides of the rubber gasket surface to ensure correct sealing. 
 
The brass flow fuel line housing unit was altogether more interesting. T\
wo things were obvious 
after it was removed. The first was that it was coupled to a hard style \
plastic fuel hose pipe via a 
screw thread affair – which resulted in the 8mm pipe being reduced (\
easily) to less than 5mm 
diameter – a reduction constituting a very unwelcome restriction in w\
hat is the low pressure (suck) 
side of the fuel system. In addition, at the other end of the pipe (whi\
ch would be inserted down 
into the fuel) there was a soldered sintered mesh filter screen fitted \
over the pick up end of the 
pipe with a rather unfortunate 5mm square hole in it. An attempt had bee\
n made to repair that 
hole by filling it with silicon – which had come lose and was flappin\
g around in such a way to 
intermittently block the pick up.
 
Two steps were taken to alleviate these problems. Firstly, the upper fue\
l housing screw thread was 
discarded and drilled out to accommodate an 8mm interference fit copper \
pipe. A short length of 
pipe was then soldered into that hole. Secondly the sintered mesh was re\
moved from the pick up 
end of the pipe and the broken end was cut off. A second area of the mes\
h filter was then cut off, 
flattened and hand soldered to close the pick up end of the filter. The \
pick up pipe was dressed to 
remove its sharp edges, and then the completed filter was refitted and s\
oldered into place. It has 
to be said that the end result was not at all pretty, but it would form \
a capable filter which was 
vastly better than it had been before. The pick up was refitted and seal\
ed, again using silicon. Time 
will tell if there is any leakage when the tank is filled to capacity. 
 
An addition check was made of all the fuel line couplings at the back of\
 the engine to try and find 
the elusive petrol smell. With the engine cold, the injection system was\
 primed four times, at the 
end of which fuel could be found leaking from the flow side of the pipe \
work. All the jubilee clips for 
both the flow and return hoses were tightened and the leak was resolved.\
 The fact that the engine 
was cold made that problem vastly easier to locate. 
 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (29 of\
 55)16/10/2010 9:03:18 a.m. 
						

							
LAND ROVER 
 
 
Vacuum plumbing design
Connections to the engine intake are used by a number of key engine comp\
onents. There are five 
vacuum ports on the engine…
 1.  Primary vacuum feed for the brake servo
2.  Advance / retard vacuum feed for the distributor
3.  Vacuum feed for the fuel pressure regulator
4.  Vacuum feed for the PCV (crankcase ventilation system) from two so\
urces.
5.  Vacuum feed for a dash board gauge. 
 
The brake servo connects directly to a driver side port screwed into the\
 plenum intake runner 
manifold using silicon hose. It is exposed to engine manifold pressure w\
ith no restrictions. 
 
The passenger side of the plenum intake runner manifold has three ports \
screwed into the side, 
two of which have piped ports (one of which is fabricated in copper). \
All these ports are exposed to 
engine manifold vacuum without restriction.
 
The port nearest the firewall connects to a 10mm copper U bend to a hose\
. The hose connects to a 
small 10mm copper pipe blocked and drilled with a 2mm metering orifice t\
he other side of which 
connects to the white plastic PCV “T” piece. This is designed to p\
rovide metered engine vacuum to 
the PCV system when the throttle is closed. 
 
The second intake runner port (located in the middle of the manifold) \
connects to the dash vacuum 
gauge via a reducer. 
 
The vacuum feed to the fuel pressure regulator is taken from the plenum \
upper chamber port 
located directly under the bypass air stepper motor valve and which is e\
xposed to engine manifold 
vacuum without restriction. This feed uses red silicon hose. 
 
The advance / retard feed for the distributor is taken from a metered po\
rt directly on the top of the 
plenum throttle body. This feed also uses red silicon hose. 
 
 
 
 
 
Engine cooling
The plumbing used for engine cooling has been covered in the intake mani\
fold section of these 
notes. There are three related issues worthy of note. 
 
The first is that a standard 88 degree thermostat must not be used on th\
is vehicle as it will cause 
overheating. Instead a 74 degree thermostat, with no joggle hole, must b\
e fitted.
 
The second is that the original carburetted engine had the engine fan re\
moved and replaced by 
twin electric fans switched by sensors in the top coolant hose. However,\
 in warm conditions the 
arrangement is prone to overheating especially following a stop after lo\
ng soak conditions. In an 
effort to reduce the problem a fabricated bracket has been added to the \
water pump shaft and a 
solid coupling MGB 14” plastic fan has been fitted. It will provide l\
imited air flow over the radiator 
when the vehicle comes to a stop – and if it works well, a viscous co\
upling may be employed. In 
fact, that fan has now been removed – as it was wholly ineffective at\
 resolving the marginal cooling 
problems. 
 
http://www.conehead.org/Projects/Landrover/EFi/efi%20-%20web.htm (30 of\
 55)16/10/2010 9:03:18 a.m.