Flooding protection system XJ12 III H.E

XJ
1

Dear all,

since this is my first contribution to the forum, I would like to say Hello! to everyone, and a big thanks to everybody for all the useful advice given here. The forum, in combination with the technical FAQ section on the webpage, is a hoard of treasure when it comes to fixing all those charming weaknesses we have to struggle with.

That being said, I today encountered a problem which is completely new to me (16 years XJ6 III, now XJ12 III HE since a few weeks). I am speaking of the following phenomenon (XJ12 III Sovereign H.E. 1983): When I turn the ignition swith to position 2, but without running or even cranking the engine, the fuel pump starts its duty (which is OK), but then won’t stop (which isn’t – it should stop after two seconds). In other words, the so-called Flooding Protection System is down. As this car is relatively new to me, I cannot say wheter this is a new issue or I just didn’t notice it until now.

I already checked the related relays, they are all fine and doing their job. However, while checking, I got puzzled by the following:

Without the pump relay installed, and with a multimeter in its place, the wires of which connected to the signal circuit of the relay, turning the ignition on pos. 2 will produce 12V for about two seconds in the signal circuit (which shoud switch the relay on and off). After these two seconds, the ground connection of the signal circuit, which directly enters the control unit, will be broken (while the 12V are kept – dead end). So, without the relay in place, everything works fine.

However, with the relay installed, the signal circuit will stay stable, that is to say: The control unit will keep the signal circuit grounded, and hence the pump will continue to run – forever.

I already studied the circuit diagrams in the workshop manual, but they won’t help much, as there seems to be something going on in the control unit, which is more or less a black box as seen from the perspective of the manual.

Anybody any clues on this weird behaviour of my fuel injection system?

Best
Florian

2

Strange.

I’ll do more mulling after a bit but, just for the moment, make sure the correct relays are installed.

Terminal 86 on the control side fuel pump relay is powered from the “main” relay. The main relay should have an internal diode. Sometimes they get replaced with a relay lacking the diode. I’m not at all sure this could cause your problem but, what the heck, couldn’t hurt to check. The diode will been seen on the mini-diagram on the relay itself.

Also, both the pump relay and the main relay should be the “double 87” type. That is, they should have terminals marked 30-85-86-87-87 as opposed to the much more common 30-85-86-87-87A. Again, I’m not sure if this could cause your problem.

Cheers
DD

3

That’s probably a 6CU ECU. Those suckers were famous for the fuel pump control circuit failing. You could ship it off to AJ6 Engineering to fix, or you could just jumper a wire from the fuel pump relay to ground and the pump would run whenever the ignition is on. If you ever crash the car and it’s upside down and dripping fuel, be sure to reach over and turn off the ignition.

4

Florian,
Welcome to the list.
I suspect that Doug may be onto something with the fuel relays. Attached is a picture of the Fuel Pump Relay (#4) and Main Relay (#5) mounted on the radiator top rail in my 1990 V12 Vanden Plas.

2019-01-11 10.03.10
2019-01-11 10.03.101885×1060 834 KB

The Fuel Pump Relay is Jaguar part number AGU1068 and Bosch part number 0332014113 and the Main Relay is Jaguar part number AGU1070 and Bosch part number 0332014112. They look alike from a distance but are stamped differently and perform differently electrically. Maybe someone working on your car didn’t realize this?
Please check again that you have the correct relays in the correct places. There were lots of changes to these XJ12s over the years and it is possible that your 1983 is wired differently than my 1990.

Paul

5

I’m pretty sure it’s the fuel pump relay that’s supposed to have a red line across it.

6

Hi everyone,

just let me say: You guys are great! Thank you so much for your quick replies!

I have just been down to my garage and double checked everything: The relays are working properly, they are all original parts, bearing the correct BOSCH part numbers. Even checked the diod in the main relay – works fine. I then disconnected the cold start relay to make sure there is no interference, but that makes no difference.

In the workshop manual I was able to find a passage (19-22) which fits my description of the pump relay’s signal circuit: The main relay steadily provides 12V at the pump relay’s contact no. 86 with the ignition on (no matter if the engine cranks/runs or not). Using the multimeter, the ECU provides ground connection at the pump relay’s contact no. 85 (wire directly enters ECU at contact no. 15), and this ground connection is broken by the ECU after roughly two seconds. However, with the pump relay in its place, the ECU keeps the ground connected (and has the pump running).

I was thinking: What’s the difference between these two situations? Obviously, the multimeter will allow for only a minimal current running between contacts 86 and 85. This current will be slightly higher with the relay in place: The resistance of the coil might be high, but not as high as the resistance of a multimeter in voltage testing mode. This difference between practically no current (multimeter) and little current (relay) triggers the error in a part of the ECU which is described in the manual as “electronic time delay” (relay?? the German manual has “elektronisches Zeitrelais”).

Theory 1: The current passing from 86 to 85 “tells” the ECU that ignition is now ON, and some other sensor fails in “telling” the ECU that the engine is not cranking/running, so that the electronic time delay is overridden by some other ECU-component.

Theory 2: Sort of short circuit in the ECU at the electronic time delay (might be a slightly broken transistor).

Theory 2 seems much more plausible to me. First because of what Kirbert says on those ECUs, secondly because the engine runs fine – a dead or distorted sensor (temperature, air, pressure, rpm) would probably cause other, more severe effects. (I should admit that the idle is a little high, about 1,000 rpms in park/neutral, falling back on 800 rpm in drive, but that might be a different, and, as I suspect, not that puzzling, but well-known issue I just hadn’t had the time to look at thoroughly until now.)

So, what do you think? Seems like I should open the ECU next and dig into the circuit there …

Best
Florian

7

Kirby,
Sorry, but the Main Relay (AGU1070) is the one with the red stripe across it and the Fuel Pump Relay (AGU1068) does not have a red stripe across it . This is true in both our 1990 XJ-S convertible (5.3L V12 w/ Marelli ignition) and 1990 V12 Vanden Plas (5.3L V12 w/ Lucas ignition).
The attached picture show the two relays side by side with the AGU1070 Main Relay on the left and the AGU1068 Fuel Pump Relay on the right. The diode is shown on the Main Relay schematic.

20200924_121650 (3)_LI
20200924_121650 (3)_LI3229×2187 1.29 MB

Paul

8

Yeah, my memory acting up again. The diode in the Main relay prevents the entire system from powering up in the event some nimnul has connected the battery backwards. If they had put that diode in the harness or socket rather than in the relay, they could have used a standard relay there rather than the red stripe version.

What’s with the string?

9

Kirby,
Those strings hold cardboard identification tags onto the relays. I have hundreds of good spare parts for our Jaguars that I harvested from my parts cars or purchased from others. I usually put a tag on them with part numbers, where I got them, any tests that I have done on them, etc. This helps when I have a failure and need a replacement part or when I sell them on eBay. There are several Bosch relays that look the same from a distance but have different part numbers and performance characteristics. Having them tagged helps me to quickly sort them out.

Paul

10

Kirby,
Here is another picture of the spare Main and Fuel Pump Relays that I carry around in my concours 1990 V12 Vanden Plas as spares just in case I need them.

20200924_150847
20200924_1508473365×2601 2.07 MB

This shot also shows the cardboard tags held onto the relays by the strings shown in the other picture and some of the info I wrote on the tags.

Paul

11

Damn! You actually look organized!

Back in 1980’s, my FIL’s Citation developed a water pump leak. He was talking about taking it to the dealer and shelling out the big bucks and I said phooey, nuthin’ to it. The water pump on a Citation is right up front, easy to access. I volunteered to help him change it out and save a pile of money as well as know that it’s done right. First step, went to AutoZone and picked up a rebuilt water pump for about 50 cents.

We open the hood to get started, and he’s got newspapers laid out all along one wall of the garage. Every time I remove a bolt, he sets the bolt on the newspaper, draws a circle around it with a marker, and writes a description of where that bolt came from. Needless to say, I’m yanking bolts faster than he can write. He wants me to slow down, I want him to speed up. I remove five identical bolts and he wants to make sure each one goes back where it was, I say “Who cares? They’re all the same!”

Then we find one bolt with a spacer on it. Something about locating some bracket correctly. So I point out the spacer to him and tell him to remember where it goes. He goes to the newspaper to write it down.

Pump out, compare with the rebuilt. The impeller is wrong! It’s a mirror image of the original. Back to the store. Get another off the shelf, that one is correct.

Putting the car back together, I get to that one bolt and say “It’s time for that spacer!” and he breaks up laughing! Like, “How did he remember that!” While I’m like “It was only an hour ago! How could you forget?”

12

**
How did you connect the multimeter, Florian?

The ECU starts operating as the main relay engages, ign ‘on’. Which nominally starts the ‘2 second’ delay sequence within the ECU. The orange wire, fuel relay socket ‘85’, should be grounded for two seconds - then possibly switch to 12V, or indeed ‘break’. (In ‘voltage’ the multimeter should be connected between ‘85’ and ground). The ECU action is independent of external inputs

Frank
xj6 85 Sov Europe (UK/NZ)
**

13

Hi Frank!

I performed the following tests:

  1. Multimeter in voltage testing mode, 86 on +, engine block serving as ground for -. With ignition on (but engine not running), I have steady 12V.

  2. Multimeter in resistance mode, 85 on +, engine block serving as ground for -. Turning the ignition on, gives roughly 40 ohms for about 2 seconds, after that resistance goes up to infinite (the same with the ignition off).

  3. Multimeter in voltage testing mode, 85 on +, engine block serving as ground for -. Without ignition: 0V, turning ignition on: 0,06V for the famous 2 seconds, then 0,01V until ignition is turned off.

I mention test nr. 3 as I am not sure wheter this could point to some sort of current leak (transistor? capacitor?) inside the ECU?

Generally speaking, tests nr. 1 and 2 indicate a functional system. However, with the relay in place (the coil connecting 86 and 85 has a resistance of about 80 ohms), ground connection at 85 will last.

I’m still puzzled by this Lucas feature, but up to now didn’t have the time to look into the ECU. Unfortunately, there seem to be no schematics floating around the WWW, which would make it much more easier.

Best,
Florian

14

Correct.
IIRC, the ECU is switching the ground.

Multimeter should be on continuity mode.

Forget the multimeter for a moment and connect a 2w test lamp at terminals 85 and 86 (+). You could also use a LED with a 800 Ohm resistance.
With both the test lamp and the relay connected you can see first hand what is happening when you start, and more important, what happens when the engine dies.

I’ve made this a permanent modification on all the important relays in my car and it’s great for troubleshooting.

Relay 005

15

**
#2 Principally, Florian; resistance measurements on powered circuits is unreliable - as it is on connected circuits…

#3 Theory dictates that with the orange wire grounded in the ECU; the voltage on ‘85’ should be 0V - with ECU ‘break’; it should be 12V; no current flowing. Ie, the ECU doesn’t make a clean ‘break’…

Current is restricted by the resistance in the relay coil - what resistance do you have on the relay coil itself? (Relay pin ‘85’ to pin ‘86’). This is the nominal current to operate the relay, but there is no specs for the minimum current required to operate the relay - or indeed, in this case, to hold the relay engaged once engaged…so…

As an aside; the ‘2-second’ pump run is not related to flooding protection - it simply pressurize the fuel rail prior to cranking. The main input telling the engine is cranking/running is the signal from the ign amplifier.

That said; there is no actual harm done with the pump running continuously - it’s sole purpose is to deliver fuel to the fuel rail. There is no injection unless the engine is turning - though at a finer level; leaking injectors may deliver fuel to standing engine with pump running.

However; pump not running with ign ‘on’ is one part of ‘collision’ safety protection; in a crash, when fuel lines may be disconnected/broken - a running pump represent a fire danger if the driver is unable to, or forget, to turn the ign ‘off’. The second part of the safety is the inertia switch, which, when tripped in a collision, will disable the main relay - which in turn unpower the ECU and fuel pump…

Unless tests of the relay reveals that the relay ‘hangs up’ when engaged; I ‘think’ the problem is in the ECU - failing to completely ‘break’ the fuel relay connection in the ECU. Whether this fault is confined to the fuel pump relay, or a symptom of something more is sort of guesswork - I ‘think’ not…

Frank
xj6 85 Sov Europe (UK/NZ)
**

1 Like
16

Hi everybody!

Thanks again for the replies. I like the LED mod of the relays. :slight_smile:

As for “flooding protection”, that’s just the heading for the 2-sec-mechanism in the workshop manual. I completely agree with you, Frank, that safety comes first.

The relay is fine (resistance of the coil is about 80 ohms), I also tested two other relays with identical specs and they acted just the same. So, yes, ECU is to blame. Bad news …

Best,
Florian

17

**
A back-up test is to measure orange wire current, Florian…

Using meter’s ampere setting in series with the ‘85’ relay prong - feeding 12V to relay ‘86’. Nominally, with the relay grounded in the ECU, the current should be 0,15A - but, as said before, I have no idea of the current required to keep an engaged relay latched…

The fault, in itself, should have no influence on engine performance or fuel economy - drivable…

Frank
xj6 85 Sov Europe (UK/NZ)
**

18

Frank: Sorry for the delay, I had no access to the car for two days. Today I did the test as advised: With the multimeter in line, turning on the ignition results in a current of 0,12A. After two seconds the current is slightly reduced to 0,11A, but that obviously is still a sufficient current to keep the relay switched on. I did this test with two different relays (of the same type) – same results. However, if I start the ignition, wait for two seconds and connect relay and multimeter after these famous two seconds, there is no current flowing (0,0A). All this points, I think, again to the ECU: While the ECU breaks the ground wire without anything connected, it doesn’t succeed in doing so when current is already flowing.

Best
Florian

19

**
Seems pretty conclusive, Florian - well tested, though ‘why’ is elusive…

The functions of the ECU in detail is murky, even with diagrams available. Rumor has it that even the constructor wasn’t too sure - with an analogue, or partly analogue computer, some tweaking was the order of the day to get obedience when they made it. The fault is likely very simple one, and may be obvious on inspection - successful tinkering has been done…

However, the fault is a minor one, with no running downsides…so…?

Frank
xj6 85 Sov Europe (UK/NZ)
**

20

Yes, it is a minor fault, that’s for sure. On the other hand, I feel a little uncomfortable about the security issue. The other day I was playing around with the idle adjustment and the engine died because of me tightening the AAV screw too much – fuel pump keeping running. I would not want to have that in case of a severe accident, so, I’ll see what I can do about it. I bet it’s just a small current leak inside a transistor or thelike. Just another item on an endless todo-list…

Thanks again for all the help, I’ll keep you up-to-date if I find something out.