Fuel Pump Removal, Fuel Gauge Readings
 

I've finally got round to looking into a problem I have with erratic fuel gauge readings.

Basically what happens is that the gauge suddenly drops to zero from just over half full when there is still around 35 liters left in the tank.
I have an xjr so it has two fuel pumps (to deal with the high demand under full throttle of the SC engine) and gauge senders, one each side of the saddle tank. Each pump delivers fuel to the engine and also transfers it to the other side of the tank.
So I've put the instrument cluster into test mode and looked at the individual sender readings and when the gauge drops to zero the readings are;
Right hand side (same side as the filler neck) reads 90 left hand side reads 160. When the tanks is full I believe these two readings are around the 240 mark although I haven't actually confirmed this.
I conclude from flooring it that both fuel pumps are working as there is no hesitation, the car goes as it should. There are no running problems at all, just the gauge readings are at fault.

Ok so my conclusion is that the right hand pump is doing what it should but the left hand pump is delivering fuel to the engine but is not transferring it to the right hand side for some reason when the level of fuel in the tank drops below the top of the saddle. This could be because of
1. a blockage in the pick up for transfer to the other side referred to a jet pumps, I think they are holes in the base of the plastic bucket housing the fuel pump in the tank
2. the transfer pipe has become disconnected from the outlet of the left hand pump
3. the transfer pipe from the left hand pump is blocked.

So I'm going to take out the pump modules, transfer pipework etc to investigate and hopefully solve this problem.
I have a tool to take off the plastic ring off the top of the pump module
I have the workshop manual procedure

So to my queries for you;
Any comments on the above and my conclusions are very welcome
Any tips, advice on getting the pumps out easily, safely and without damaging anything

Thanks to the forum which is the reason I know any of the above!!

Hi Chris,

I'm in a rush so I haven't thought through your analysis, but another possibility is that there is a dead spot on the resistive trace of the right hand fuel level sender. When you pull the pump/sender assembly, you can test the sender's DC resistance as you slowly move the arm through its full range of motion. Any high or low glitch in the readings indicates a bad spot on the resistive trace.

Cheers,

Don

Don,
Ok I'll check that.
What I did do to check the senders was to rock the car while I had the instrument cluster in test mode. I saw the readings from both senders go up and down as the fuel sloshed around in the tank. So I was thinking that this indicates that at least they aren't stuck on the track.
Also forgot to say in the previous post that I have the following codes at the instrument cluster;
B2879 Fuel tank jet pump fault
U2510 Invalid data for vehicle security
U2511 data mismatch
B1689 Autolamp delay circuit failure
U1900 Can communication bus fault anti theft system

I do have another problem in that my alarm does not set for some reason which is to do with U2510, U2511 and U1900. Another days work!!

B2879 clears Ok and stays away until the gauge does its drop to zero thing. From what I've read (can't remember where) this code is generated when the instrument cluster thinks the two readings from the senders are inconsistent.

I haven't seen it for a while but there have been many reports in the past about

problems with the fuel level indication.

I had a brief problem myself but years ago.

The root cause was poor contact of the sender due to corrosion etc.

You can buy the sender separately and it just clips on the side of the pump housing.

Hopefully your problem could be as simple as that.

Problems with CAN comms are often assoc. with poor contact in a connectors.

The Elect. Guide indicates where these are located and breaking and remaking the connector

will often resolve the fault.

Ok

So this afternoon I had a go at getting the pumps out.
First took out the left hand pump. This was the one which according to the IC test mode was lower (reading of 90)
Sure enough there wasn't a lot of petrol in this side of the tank. The petrol supply line to the engine came off the top of this one with a quick release connector. Inside the tank there were three pipes connected to the pump bucket with quick release couplings..
1. One from the right hand side of the tank which joined the pipe out from the left hand pump for the engine fuel supply
2. One from the right hand side of the tank for fuel transfer from right to left
3. One to the right hand side of the tank for fuel transfer from left to right

It was easy to disconnect the pipes and remove the plastic pump bucket although care is needed to avoid damaging the very delicate sender and float arm. No crap in the bottom of the tank and the large screen filter attached to the outside of the plastic bucket that houses the pump looked reasonably clean.
On the bench tested the resistance of the sender and was all to spec, varying from around 20 ohms (empty) to 160 ohms (full) across the sweep of the float, so nothing wrong there.

Moved on to the right hand side. The locking ring was very tight and it became apparent that someone had been here before as there was black sealant around the top. This side of the tank was brimming with fuel so had to pump out quite a bit so I didn't make a mess getting the pump unit out. Again this is consistent with the sender signal from the IC test mode which showed the level to be higher on the right hand side. When I got the top locking ring off the large grey O ring seal was quite distorted and it had obviously been over tightened. That's the reason for the sealant. O ring knackered and sealant and over tightening in an attempt to get a seal.

Getting this one out was much more awkward as the three pipes couldn't be disconnected from the bucket and had to be pulled across from the other side of the tank.
There was quite a bit of crap in the bottom of this side of the tank and the large screen filter attached to the outside of the plastic bucket that houses the pump looked noticeably dirtier than the one on the left.

Right hand sender also tested OK on the bench, resistance in the correct range over the sweep of the float.

There were no pipes disconnected or split on either side of the tank.

So now I took the bucket things apart and checked that the transfer pipes weren't blocked.
I briefly put 12 v to each pump and they both ran they, sounded the same. Tested resistance across the pump connections and both came at around 2 to 5 ohms. Is this what a good pump/motor should show? I don't know.

When you take the lid off the bucket and take out the pump there's another screen/filter on the bottom/suction of the pump. The right hand one was noticeably dirtier than the left. Blew through all the connections to the pipes inside the left and right hand buckets and all clear.

So I'm non the wiser, both floats and pumps seem to be working OK. But for some reason the fuel is not being transferred from right to left.
The only things that of note of difference on the right hand side were the botched seal to the hand locking ring the dirty 'outside the bucket' filter and the dirty 'inside the bucket' pump filter and the crap in the bottom of the tank.
Could these be the reason for the lack of fuel transfer?
It is possible that there was a blockage somewhere that I have dislodged in the process of taking the stuff out and apart?

I'm a bit at a loss now as to what to do other than put it all back.
Any comments or suggestions very very welcome.

I am guessing that a previous owner had the same problem with the RHS fuel level and

that's why he tried to seal the tank.

I have an idea that there is a NRV between the two pumps, did you find one and if so was

it working?

I can only suggest that maybe the RHS pump is strong enough to feed the engine but

not feed the LHS also.

I think you need to work out how the system works eg does the software rely on the LHS sender value to tell it to stop feeding and how does it

do that and still feed the engine?

3. One to the right hand side of the tank for fuel transfer from left to right

What is no 3 for?

I am assuming the filler tube is on the RHS and that is why there is more dirt there?

Yes there are two non return valves, one on the outlet to the engine from each pump before they join together. I think they are working OK.
This is a description of how the set up works for the earlier SC engines from the Jag X350 technical guide with the two pump set up.
https://cimg5.ibsrv.net/gimg/www.jag...91d213039a.png
Description;
https://cimg7.ibsrv.net/gimg/www.jag...736df54109.png
Pipe no 3 is there because fuel is transferred constantly left to right and right to left.
Yes filler is on the right hand side which is the drivers side on my Irish/UK RHD spec car.

I've just read the start to this post which describes pretty much my problem.
https://www.jaguarforums.com/forum/s...-4-2-a-106285/

I had found this and read before but jumped into the middle of the thread and didn't read the start of it DOHH!

So Roger had the same problem on a supercharged S type which have the same pump arrangement I assume. He didn't get round to acting until the car refused to go anymore.
He had found low fuel pressure at the rail. He took out both the pumps cleaned all the filters, 'outside the bucket' filters to the jet nozzle inlets and the ones attached to the suction of the pumps inside the buckets and it solved his low fuel pressure and gauge dropping to zero problem.

So this looks like progress. It seems that the dirt build up on the right hand side filters affects the pumps performance enough to redcue flow to the other side of the tank via the jet nozzle system. I'm hoping so anyway!

Having tested the senders again the right hand one does seem to misbehave a bit mid way in the sweep with the resistance jumping up and down outside the 20 to 160 ohm range. So I'm going to replace this one. I'll clean all the filters, buckets and the right hand side of the tank. Fit new grey rubber gaskets/sealing rings and reassemble. Hopefully this will solve the problem.

Incidentally there is another thread discussing fuel pump replacement to the early x350 xjrs here;
https://www.jaguarforums.com/forum/x...solved-153085/

My pumps are not the same as those that Lagonia removed from his xjr, Mine have a two pin plug connection rather than two separate spades, a metal base rather than plastic although they are very similar.

Before reassembling I'll take a few photos and see if I can find any part numbers etc on my pumps for peoples future reference.

Here are some photos of my xjr fuel pumps

Only writing reads;
3C5U-9350-AB
07/02/03 10-19-43

Haven't searched yet to see what comes up

https://cimg7.ibsrv.net/gimg/www.jag...2c0e1bd25c.jpg
https://cimg7.ibsrv.net/gimg/www.jag...1545a221a6.jpg
https://cimg4.ibsrv.net/gimg/www.jag...d4280030bc.jpg

So after cleaning the fuel pump filters,and the inside of the bucket things I put it all back together this morning and back into the car;

This is the right hand module which has the three pipes that travel to the other side of the saddle connected to it;
The large sock filter is connected to the small chamber with the two induction jets, one of which pushes fuel from the right to the left of the tank
https://cimg8.ibsrv.net/gimg/www.jag...3ba0c474d5.jpg

This is the new gauge sender I fitted to the right hand module, slightly different and a bit more robust (but not much more!) than the original
https://cimg0.ibsrv.net/gimg/www.jag...48f2c26e2c.jpg

So getting them back in was a bit awkward.
Pumped out all the remaining fuel from both sides of the tank and cleaned it.
To get the right hand one in I had to take off the gauge sender (it just unclips from the bucket) as it was impossible to feed all the pipes in, across to the other side of the tank and get the bucket with the float/sender attached through the hole without damaging it. Clipping the sender back on while the whole lot was in the tank was a bit tricky but a third pair of hands makes it possible.

Connected up the pipes after getting the left hand module in, fitted a new grey sealing gasket to each side (one of the ones I'd taken out was ruined due to over tightening and not being seated correctly)
Torqued the rings up to 70 Nm in accordance with the manual after putting about 7 liters of fuel back into the left hand side. This is the tool I get off ebay to use on the rings on top of each module;
https://cimg7.ibsrv.net/gimg/www.jag...59ab1766ae.jpg
Dumped about 10 liters of fuel in the filler neck, and up she started.
Always a relief for me after messing around with things that the beast lives afterwards!
Went for a drive all seems ok, fuel gauge showing just above the red line.
I haven't seen it down there for a long time as its either been on zero or at half full or above.

So I've filled it up with fuel, looked at the two sender readings for the left and right side, both at 254, fuel gauge on the dash showing full.
Now we have to wait an see if the fuel gauge behaves itself!!

I'll let you know.

https://cimg6.ibsrv.net/gimg/www.jag...07702e513e.jpg

Dear all

Just noticed a mistake in my first post on this thread which might cause a bit of confusion for anybody reading through it later.
Basically got my left and right mixed up. So read left for right and right for left in the opening post!!

Back to the drawing board on this one I'm afraid, fuel gauge is not behaving!!!
Same thing happening as before. Remember my car is an R, so a fuel pump on ether side of the tank, both supplying the engine and both transferring fuel to the other side of the tank.
I have been driving around with the instrument cluster in test mode and watching the RAW 1 and RAW 2 fuel sender readings from the right and left hand sides of the tank respectively. My fuel gauge drops really slowly, then when its at half goes to zero with still around 35 litres left.

The RAW 1 and RAW 2 readings (0-to 254 scale is empty to full) dropped in unison until each at around 170, then RAW 1 (right hand side of the tank) stays at around 170 and RAW 2 drops until its around 90, then gauge drops to zero and fuel warning appears on the screen.

So its still happening. When the level goes below the top of the saddle (I think this is a value of 170) the fuel is not being transferred from the right to the left to keep the levels equal.
I bench tested both pumps with 12 V when I had everything out and both ran. Cleaned the filters and jet pump nozzles, etc etc see above.
I'm at a loss.
Any ideas?

Just wondering
Anybody know if there is a way I can monitor the operation of the fuel pumps when the engine is running?
I have a carsoft i30 scanner

Hi Chris,

Your i930 should be able to read the fuel rail pressure, but I don't know of a way for it to see the activation/operation of the two fuel pumps. As we may have already discussed and if I recall correctly, the pumps are controlled by the ECM via pulse width modulated signals that the REM doubles in frequency to drive the right hand pump, and the left hand pump is driven by a separate Fuel Pump 2 Module. One possible method of monitoring the pumps would be to backprobe their connectors and connect a two-channel oscilloscope or individual DMMs with the capability of measuring duty cycle, while also monitoring fuel pressure via OBDII.

Hopefully someone else has a better idea.

Cheers,

Don

This is a difficult one.

The RAW figures tie in with the symptoms ie fuel transfer stopped.

I am suspicious of those parallel valves but it's not clear to me how they work even

though I have read the description.

I think you could try replacing the RHS pump.

Hi Meirion1,

As the car is running fine I'm not going to change the pump or any other parts at the moment.
This is an annoying problem more than anything.
I am going to monitor fuel levels, gauge reading and pump operation in more detail for a while to try and figure out more about whats going on here!
The operation of what Jag call the 'parallel pressure relief' valves are to allow fuel flow in either direction but only above a certain fuel pressure in certain conditions;

1. So no pump operation, they're closed to flow in both directions and maintain fuel pressure in the line for easy starting. Also prevent the tank from emptying through a leak in the fuel line downstream of the valve.
2. No pump operation and temp of the fuel in the line/fuel rail increases, causing vapour to form and an increase in fuel pressure, valve opens to allow relief of excess pressure into tank.
3. Pumps operating stops flow short circuiting to the other side of the tank, open to fuel flow to engine
4. Pumps operating and fuel rail pressure too high, open to tank to relieve excess fuel pressure

Have you seen this? It relates to an X Type but could apply to an X350

http://www.jagrepair.com/images/TSB/...g%20Issues.pdf

Page 5 of the above says:

If the value of F2 decreases to a value of approximately 28 to 29 this indicates that fuel is transferring. 6. If the value does not decrease, change the transfer fuel pump assembly.

Your result:

dropped in unison until each at around 170, then RAW 1 (right hand side of the tank) stays at around 170 and RAW 2 drops until its around 90, then gauge drops to zero and fuel warning appears on the screen.

Am I correct RAW1=RHS=transfer pump?

F2 = value of transfer pump sender.

BTW How many miles have the pumps done?

What do you think?

Meirion1,
That's an interesting document, thanks very much, no I haven't come across this one.
I'll have a thorough read through it. My initial thoughts are that the x350 xjr is different from the x-type in that both pumps on the x350 feed the engine whereas for the xtype one pump feeds the engine(F1) and the other pump (F2) transfers fuel from the left hand side of the tank to the right for it to be pumped to the engine.

Yes RAW 1 is I think the right hand side of the tank. This is the side of the filler neck on my right hand drive.
The car has just gone over 80000 miles, so not that many.

I asked in an earlier post about a way of monitoring the operation of the fuel pumps with my carsoft i930.
I hooked it up just to see what I could find and started the engine.
I logged into the rear electronic module and hit Live Data Stream.
Scrolled through all the pages of items and came across the fuel levels on either side of the tanks and then a couple of pages later the 'fuel pump duty cycle to the rear electronic module' and 'fuel pump duty cycle 2 to the rear electronic module'

So I took a few snaps.
Car is idling, filled up with fuel about 80 miles ago, fuel gauge showing 7/8ths full.
The fuel levels are showing 74% fuel level 1 (right hand side) 59% Fuel level 2 (left hand side) on the i930;

At idle the fuel pump duty cycle for the pumps (No 1 right hand side, No 2 left hand side) look like this;

https://cimg5.ibsrv.net/gimg/www.jag...2f279a7c10.jpg

So if I'm reading this correctly ( and to be honest I don't know if I am) the left hand fuel pump is working twice as hard as the right?

Now I'm still in the garage but I floor the pedal, this is what happens;
https://cimg6.ibsrv.net/gimg/www.jag...7903527f02.jpg
https://cimg7.ibsrv.net/gimg/www.jag...4a408c7986.jpghttps://cimg6.ibsrv.net/gimg/www.jag...cafe3ea5d3.jpg

https://cimg7.ibsrv.net/gimg/www.jag...d33dcad97e.jpgThe duty cycle to fuel pump 2 went up to 100% at one point dropped to zero, was very erratic, then back to 52% when engine returned to idle
The duty cycle to fuel pump 1 rose up to 74.9% then dropped back down to the previous reading of around 25%.

So does it look like the signal or control of the second fuel pump is dodgy?

I then scanned the REM for codes and got a new one (or maybe I hadn't picked it up before) P1235. This is what the manual says about this code;

P1235;

Fuel Pump Control
Out Of Range
Rear electronic module,
fuel input (modulated
pulsewidth) - out of range
(low)

Carry out any pinpoint tests
associated with this DTC using the
manufacturer approved diagnostic
system. Refer to electrical circuit
diagrams, notes and check engine
control module to rear electronic
module fuel input (modulated
pulsewidth) is out of range (low)

Any comments welcome!
Not sure what this latest information may mean

This might help-there's 11 pages!

https://www.jaguarforums.com/forum/x...solved-137070/

Another day gone! lol

Hi Chris,

From the Engine Management System manual:

"The ECM communicates the fuel flow demand to the REM as a pulse width modulated (PWM) signal over a single line at a frequency of approximately 150 Hz and a duty cycle range of 4 - 51%. The REM amplifies this signal by increasing frequency and doubling the duty cycle, thus providing the variable high current drive for the fuel pump."

On the supercharged vehicles with two fuel pumps, there is an additional Fuel Pump 2 Module which mirrors the REM functions for the second pump. The Fuel Pump 2 Module is mounted behind the REM in the right rear inner fender.

According to the Technical Guide, both fuel pumps operate simultaneously, so presumably their duty cycles should be identical.

Here's the listing for P1235 from the DTC Summaries:

https://cimg5.ibsrv.net/gimg/www.jag...359e7ca663.jpg
https://cimg8.ibsrv.net/gimg/www.jag...683bd552b1.jpg


You can find the pinpoint tests for P1235 in the Powertrain section of the Workshop Manual, which you can download from the X350 HOW TO quick links thread on the home page of this X350 forum.

Cheers,

Don