S3 XJ6 - it ran until I tried to make it run better

(Beau LeBlanc) #21

Hi all, it’s been a while, but time for an update.

I rented a fuel pressure test kit from a local auto parts store, but it was made such that it required using one of a specific set of adapters to hook up; it wouldn’t simply attach at the fuel rail. The only way I could get it to work was via a bulky “T” adapter that could have fit behind the fuel filter in the trunk. I detached the line from the fuel filter, and the gas that leaked out of the filter was dark brown/red. A fuel tank drain/flush seemed inevitable. According to the PO, the car had only been run from the left tank for as long as he could recall, so I focused on that one.

I was able to get the smaller drain plug out, and at first nothing came out of the tank. I inserted a small screwdriver through the opening and it felt obstructed. A bit of pressure seemed to free the obstruction and rust-laden fuel trickled out - but only for a moment. I poured some clean fuel into the tank, and it drained immediately (along with some fine rust particles).

At that point I realized that my non-start issue may have been caused by nothing more than an empty tank. I knew that the fuel gauge was not working, but the PO represented that it he had put 8 gallons in. Nevertheless, I flushed the tank with clean fuel until it no longer carried rust with it, and it came out clean. I could not, however, get the larger drain plug free.

I changed the fuel filter, added a few gallons of clean fuel, and tried to start. No luck, though - it sputters after a few tries, but won’t fire up. I can hear the fuel pump working. I would really like to get the larger drain plug out and check the in-tank screen. Also, I’m thinking that the obstruction at the bottom of the tank may have been a sunk float embedded in the rust (just a theory, though). Has anyone dealt with a reluctant drain plug before?

Otherwise, I’ll check the spark plugs, as I haven’t yet done so since getting the car.

(Aristides Balanos) #22

Remove the rear light and behind you will find a large plug that is the fuel sender, remove it and you will be able to see inside the tank.
Next, make sure that the hose from the tank to the change-over valve, the valve itself and the hose from the valve to the flter is not blocked.
Also check that the valve is actually swiching.

Best ,

(Paul M. Novak) #23

I usually connect my fuel pressure gauge between the fuel rail and the cold start injector by removing that hose, repkacing it with two longer pieces of fuel injection hose, and then adding the “T” connector with the gauge at the ends ofbtge new hose. Sometimes I use enough hose that I can observe the fuel pressure from the driver’s seat while driving. Where did you try to connect the fuel pressure gauge “T”?

I have removed those large fuel tank drain plugs over a dozen times on my three Series III saloons as well the three Series III XJ6 parts cars that I have disassembled, and they rarely comes out easily. That small set screw comes out easily with my standard socket set but I usually have to get my large breaker bar to get enough of a moment arm to remove the drain plug. Sometimes the drain plug is stuck on so tight that I wondered if I was going to damage the tank as I applied more snd more force to the breaker bar to loosen the drain plug. Then slowly the drain plugs started to turn and eventually came out. I apply anti-seize to the drain plug threads before putting the drain plugs back in hoping that will make the next removal easier.

That obstruction you felt through the set screw opening was the bottom of the in tank strainers that sits at the bottom end of the fuel pickup tube.

I recommend that you have replacement drain plug seals (C26310) on hand before you remove the drain plugs because the current seals are likely to be cracked and usable.

BTW, some of us have added two aftermarket fuel filters between the two fuel tanks and the changeover valve to protect the changeover valve and fuel pump from in tank debris. There are numerous posts in the archives about this.


(Con Saris) #24

Hi Beau,
Reading your story seems like an exact duplicate to my own a couple of months ago. Had all sorts of problems getting my beast to run. I would recommend trying to get the larger plugs out and then removing the intank strainers all together. Then fit extra aftermarket filters between the tanks and the change over valve as Paul has suggested. I did and have not had any problems since.
Keep at it, it just needs patience

(Frank Andersen) #25

A sunken float will not obstruct fuel feed, Beau - it will just make the fuel gauge unreliable…

In the present situation it is uncertain why the engine won’t start; clogged in-tank filter, clogged lines and/or clogged changeover valve, faulty pump (fuel pressure) - or an engine problem.

Accessing in-tank problems, including in-tank filters and proper debris removal, require removal of the large drain plug. Which will require brute force - which may risk tank damage.

Some ways to eliminate doubts; clamp hose from changeover valve and detach from pump. Feed the pump directly by a can of clean fuel - if the engine stats and runs; pump and engine is OK. If the engine does not start; the pump/engine iteself may be the problem - and a fuel pressure test is required…

Clamp and disconnect a tank hose from the changeover valve and connect directly to the pump. This will bypass the changeover valve and test flow from the tank - if the engine starts and runs; the changeover valve may be the culprit. If no start; there may be blockages tank-to- changeover valve.

You may further try each tank in turn this way - for further location of problems/blockages to only one tank. Depending on results; further checks on the changeover valve is called for.

The point of tests is to possibly pinpoint the problem for further action. Which may then require removal of the recalcitrant big drain plug - if an in-tank problem is indicated.

The easiest access to fuel pressure test should be the hose connection at the fuel rail. Plugging the pressure gauge directly into the fuel hose will give ‘raw’ pump pressure - 60/70 psi should be ample for running the engine; a perfect pump would read 100+ psi.

Below 40 psi indicates a defective pump to be changed. This doesn’t measure ‘regulated’ fuel pressure - but a ‘T’ at the fuel rail connection will…

Of course; checking/cleaning the changeover valve as Aristides suggests may be a shortcut. Remember that all three hose connections at the changeover valve should be clamped before any disconnections - or you may have a flood of petrol…:slight_smile:

xj6 85 Sov Europe (UK/NZ)


(David Jauch) #26

Yes, but between the pump and the regulator the pressure should be the same at any point.

Does the fuel return into the selected tank in a good clear stream? (Open the flap under the filler and look). Does the exhaust smell like fuel after some time of cranking?
Look for the bullet connector somewhere around the fuel rail (coil to injector harness). Ignition on, disconnect and reconnect it a few times. The injectors should click. Then try to start the engine right away. Any luck?

That was the jerkiness I had; thats long ago. But I guess okay for a cold engine… and the 0W20 or whatever came out. Look at the oil pressure, middle of winter. Now a steady 45 (?) above 1000 rpm.
1250 is good when cold: especially because it means that it is running. :slightly_smiling_face:

(Frank Andersen) #27

Two separate pressure tests should be carried out, David; ‘raw’ pressure to verify the pump by direct connection. And ‘regulated’ pressure to verify the fuel pressure regulator - which can be done by a ‘T’ anywhere between the pump and the rail. Or more conventionally by attaching the gauge to the CSI hose…

But a crude test is indeed watching the fuel return. No return; insufficient pressure to overcome the pressure regulator - or a faulty regulator…

xj6 85 Sov Europe (UK/NZ)

(Beau LeBlanc) #28

Another long lapse in updates, but I have made some progress.

I took it to a specialist in a neighboring city for evaluation of the fuel tanks. After the left tank was removed, I took a look inside, and it appears that my efforts at flushing it had little to no effect: it was so rust-laden that, for peace of mind, I opted to replace it. The right tank is also in bad shape, but I’ll replace it later, once I’ve resolved other issues.

Now, the left tank is new and all fuel lines have been flushed. The pump works fine, and fuel pressure at the rail is as it should be.

However, I’m still having an intermittent problem similar to what I faced a few months ago: it will run for some time, but then idle will drop to a sputter, and the slightest touch of the accelerator will kill the engine. Sometimes it will start again immediately; other times it will not start until it has rested for a day or so.

I’ve discovered that, when it gets to the point that it won’t start, if I unplug the air flow meter connection, it will start and idle. Then, if I plug the air flow meter connection back in after starting the car, doing so will immediately kill the engine.

I would assume that the air flow meter is the problem, but this scenario has now occurred with two different air flow meters: the first one was present in the car when I acquired it, and the second is (what I believed to be, per information from the PO) a new air flow meter.

It is entirely possible that both air flow meters are bad, but as an alternative explanation, might the connection be the problem?

(Paul M. Novak) #29

What does the the fuel pressure at the fuel rail do when the engine sputters and dies?


(Frank Andersen) #30

When you unplug the AFM, Beau; the ECU fuels from ‘break’ data…

Basically the ECU sends current/voltage through the AFM, and reads return signal - for some reason, the engine is satisfied with the result. So a faulty AFM is not necessarily the problem; a sensor fault, or something else, may cause fuel errors. Disconnecting the AFM simply counteract another fault - which may be related to an intermittent air leak…

A sometimes bad connection is the CTS plug. Disconnect plug and measure resistance over the two wires, it should read some 2 Kohm. But also; while wriggling the plug/wire - there should be no variations in resistance…

Since there seems to be no detectable consistency reported in the problem’s appearance, like engine tempo, exhaust smoke or whatever; an open mind is required…

Which includes supervising fuel pressure as Paul suggest, fuel return and indeed that the pump is indeed running. And it is advisable to supervise/check ignition with the fault present by connecting a spare spark plug to any plug lead. This will not in itself prevent the engine from running…

You don’t specifically state how the engine runs on the road - power and response. Nor your hot idle set…?

xj6 85 Sov Europe (UK/NZ)