fuel pumps runs constant when key in "on" position

L-Jetronic?
Your correct, the fuel pump should not run all the time.

From Bosch's Gasoline Fuel-Injection System L-Jetronic Technical Instruction booklet,

"The electric fuel pump starts immediately when the ignition and starting switch is operated and remains switched on continuously after the engine has started. A safety circuit prevents fuel from being delivered wehn the ignition is switched on, but when the engine is stationary (e.g. after and accident)."

this is accomplished via the Drive Relay, Tachymetric Relay, or Fuel Pump Relay. Pick your label, turn the key and the main relay turns on the Jetronic box which in turn turns on the tachymetric fuel pump relay for a moment to build pressure and give the system enough fuel to start. Anyway the tachymetric relay turns on and stays on once receiving a signal from the coil that the engine is running. Earlier S3s had an Inertia switch instead of or in addition to the Tachymetric relay so not so sure there but suspect the same problem(s).

Is this a long running problem, like since you bought the car or a new development?

Either way, under the rear shelf carpet on the passenger's side you will find the Jetronic controller, 2 relays (main - smaller & Drive - larger) and an altitude compensation device.

If a new problem, the relay is stuck on?

If an old problem, the relay could still be stuck on but there is a fair chance someone balked at the price of the relay or couldn't get one quick enough and jumpered the fuel pump on.

I guess the safety circuit in the L-Jet controller could have an issue.

the both relays are common and can be found from a few manufacturers.

tachy relay Bosch # 028230001, Beck-Arnley # 203-0100, Kaehler (KAE) # 3.304.500 (currently on sale at AutohauzAZ, stock # 13631276264 for <$25. You may want to confirm the #s and cross reference)

main relay Bosch # 0332014125

Rock Auto shows the Beck-Arnley part for a 1985 BMW 535i $42. That is what I bought a few years ago as I could get it the fastest. I think the brand was KAE? Worked fine until it went for a swim a couple months ago. Odd though, they dropped it from the Alfa listing?

You can google the Bosch # and should get several hits from BMW specialists.

Thanks for digging into this.

This is what I can tell you. The drive relay is as I mentioned 3.304.500 and the only markings on the main relay are Sipea 12v 0440 6G. There is also another main relay in there (unhooked - probably older-defective and not removed) thats marked R 601 047 and VW 311 906 061 C). I unplugged each and all the terminals are straight and not corroded.

I checked that loose ground and it does go through the floor plus with another pink/white wire that appears to go to the fuel pump. Not sure why its disconnected, why it melted and the effect on pumps operation if not grounded.

Here are a couple of pics. One showing the ECU area with the relays and the unhooked ground. Upper left corner shows were the ground goes under the car. The other pic shows the pink/white line and ground at the fuel pump.

Thanks.

Do I read the words right?

The wire with the ground loop is burnt and in the pic apparently shares the same insulation over~sleeve with the pink with white trace, which itself is not connected at the cargo area end?

If so, and IIRC the layout correctly, then its possible the burnt black one got into and shorted the pink/white inside the sleeve resulting in someone fiddling about to get an alternative power supply to the pumps by cutting and tapping into the pink/white downstream somewhere.

Absolute pure speculation on my part, but if you wanted to fiddle with a multimeter, you could check if there were continuity between the burnt black wire loop and end of disconnected pink/white in the cargo area, then check continuity between the pink/white in cargo area to the pink/white on the sending unit.

If I'm remotely close in my guess (and that's all it really is at this point) you would likely find continuity from burnt black to pink/white in the cargo area, but not pink/white in cargo area to pink/white at sending unit.

Ideally a papajam diagram would get you a lot better idea of what goes where and to what relay terminal, and I'm only currently basing any thoughs on the pix and info provided here.
(if I even understand it correctly, which apparently this morning I'm not doing such a good job of)

The drive relay can be rewired in such a fashion as to allow the pumps to run strictly with key power and no interaction form the coil tachymetric signal.

I even had it written down here somewhere in the past actually.

Even more reason to have a proper diagram beucase someone could shuffle the connectors in the relay harness plug ends and you've never know they'd been moved if they knew the (rather simple) means of releasing the connectors from inside the plug housing instead of playing cut~n~splice.

Edit to reply to post #10:
Wow.
Apparently I just need to like walk away for a few hours or something and get my brain in order as there seems to be a lot falling through the cracks at this moment.

(get your thread subscriptions going now folks before you miss the fun! This could be the onset of getting to watch someones mind go to complete and utter **** right before your very eyes)

I am a long way from my car and my wiring drawing but:

I don't think it would be the loose/disconnected ground but electrical/electronics can do some weird things when finding ground through another device or path.


white/pink should go to both fuel pumps

In your second photo, top of tank:

white/pink, should be in-tank fuel pump

violet, fuel gauge

white/black, low fuel light

black, ground

Once again, thanks for all your involvement with this one.

I will give that thread a read later when I have some time and will go looking for that ground.

As for the wiring at the pump, the pump works, the guage works and the low fuel light will illuminate when low. I suppose for all that to work, there needs a be ground somewhere.

I visited Vintre today and both his current 91 and pics from his 87 (sold) are both wired the same as mine. But what I did notice on mine is there are 2 black ground wires on the spade connector on the fuel pump. Weird. Earlier today I beleive Tifosi was saying something about the possibility of a second ground if the first shorted somewhere.

I will need to dig deeper into this and his thread and will report back.

Thanks

A couple days back I received the Cardisc for my car, and finally dug into it tonight to see what I could learn/find about this problem. We'll I wish I did sooner. So I've read the car has 2 fuel pumps; one in the tank, and one outside the tank. The pump that I've been refering to in this thread (which turns on in the key "on" position was the one in the tank which according to the manual is supposed to turn on in the "on" position - which it does. So it appears I don't have a problem, although I need find the out of tank pump (assume somewhere under the car) and confirm it doesn't turn on in the "on" position, as that one should only turn on during crank.

It just seemed weird that any pump would turn on in the on position. So if I'm somewhere outside and wanting to listen to the car radio, the in-tank pump will be also be on... I'm not a mechanic but seems weird. Are all cars like this?

Mike