Help understanding the fuel pump relay operation

[Hupa]

Hi! My M535i won’t start. Can’t hear the fuel pump priming anymore when I power her up. The fuse was ok. Pulled the fuel pump relay and measured the socket: Pin 30, always hot 13V, correct I think. Power on, also pin 86 gets 13V. Pin 87, 0.1ohm to ground when power is off, 12-13ohm when power is on. I don’t get what happens here?? What does the 0.1ohm mean when power is off and 12ohm when power is on? I’m afraid to jump pins 30 and 87 as suggested to test. If 12 ohm is the pump resistance (sounds too high), then what happens when I switch off the power? Pin 30 stilll has 13V, but now jumped to 87, going throgh 0.1ohm to ground, this would be basically a shot circuit, right… or does that just leave the pump running?  0.1ohm sounds too low for pump resistance, or?

Then another mystery (to me, at least) , pin 85 doesn’t seem to go to ground, multimeter shows OverLimit resistance to ground. Am I wrong to assume, pin 85 should go to ground? How else would 13V at pin 86 close the relay switch if pin 85 doesn’t lead to ground? The resistance over the relay itself (pins 85-86) is 67ohm.

Is the relay just part of a more complex electronic system and I’m drawing wrong conclusions?

Educate me, please.

 

[Mike W.]

Assuming your M535i is wired like US spec cars, and I'm nearly certain it is, this is how it works.

30 gets 12V straight from the battery, unswitched.
87 is the load, the fuel pumps, plural. Now if your car is a "Green" M535i with a catalytic converter, it probably also has a heated oxygen sensor in there which would lower resistance.
86 gets power from the main relay when the ignition is turned on.
85. This is where it gets a little trickier. 85 is grounded thru the Motronic ECU, and only when cranking or when it sees that the engine is turning, so unless it's running or cranking you will not see a ground on 85.

Even with the low resistance, you should be able to safely connect 30 and 87 as it goes first to the fuse box, thru fuse 1 on US cars, then to the fuel pumps. Or get a fuseholder put a 10A or something in and connect that between 30 and 87.

And there is no prime function. Some cars like 520's had it, but not the motronic cars.

[Hupa]

Hi Mike, thanks for the tips. My car is european without catalytic converter. I tested the relay with external voltage source and it works fine, I get a click and continuity from 30 to 87. I get the pin 85 ecu-ground thing. The engine really has to be cranking to get ground? This makes the testing a bit difficult. You sure there is no pump priming? I have distinct memory of a buzzing sound when turning the power on (no cranking), and always assumed it was the fuel pump. I got the supposedly correct wiring diagram for the car, and there is nothing that should interfere with the circuit from pin 87 to ground, so the variable resistance (0.1ohm power off / 12ohm power on) is still a mystery. Maybe the fault lies there. All mentioned readings are measured with the ralay detached.
So, jumping pins 30-87 should be for testing purposes only? I mean that would send battery voltage through the fuel pump all the time, even with power off since 30 is always hot, right?
I think I’ll connect thin wires to the relay pins and tape them along the sides, so I can easily measure everything with the relay in place. A friend also suggested to get an amp-pen, as bad fuse/wire/connection may show good voltage/low resistance, but still restrict when practical current should pass through.
I fear, I still need to clear the way to the fuel pump(s)… well, I’ll cross that bridge when I get there.

[Hupa]

Update: cleaned all relay contacts, also in the socket, with sandpaper. While cranking without the relay, I get NO ground from pin 85 socket slot. Multimeter shows ”overlimit”. Did continuity test from pin 85 socket slot to ecu pin 20 slot, resustance 2ohm (my meter wire had a few connections, so I think this is continuous enough). But what to do next? Why would ecu deny ground?

[Hupa]

Update 2.0: This is getting interesting… was measuring the ecu connector grounds, found 4 of them, I think. I was following this: https://mofler.com/uploads/attachments/ ... 27f6c32e5f

and this:
http://e28-535i.com/upload/e28_Harness_connections.pdf

BTW, what does slot 5 ”SHIELD GROUND REF POINT” mean? This name was given in the second link.

But then… I noticed that pin 18 slot should have ”continuous power”. Tried measuring the voltage against ground, but the probe went into slot 18 with no force at all, and sure enough, the slot was completely EMPTY, no connector, no nothing. What does this mean? Have I been trying to start the car basically without an ECU, or what? The ecu has been dangling on its wire bundle as I was working on the ABS connector, and I have driven it a bit around the parking lot. Could this have stressed the slot 18 connector to become loose and eventually off? I’d still think I should seethe connector back there, but there was nothing. This is starting to be too much disapointment for one day..

[Hupa]

More testing: even though multimeter doesn’t show continuity to ground from pin 85 when cranking, the amp-pen connected between 85-86 illuminates when cranking. This is strange… where does the current go if not ground?
Then, reattached the relay, connected the amp-pen from fuse 1 to ground, illuminates when cranking (tried both ends of the fuse holder, same thing). So it seems that the relay is connecting the voltage from pin 30 to 87 and then to fuse 1. However, the car won’t start.
Then, jumped pins 30 and 87. Distinct hissing sound somewhere in the engine started right away (what is this?) Still, the engine won’t start. I don’t hear the hissing sound when cranking the engine normally, but maybe it can’t be heard through the other noise. Strange… strange.

[Mike W.]

More testing: even though multimeter doesn’t show continuity to ground from pin 85 when cranking, the amp-pen connected between 85-86 illuminates when cranking. This is strange… where does the current go if not ground?

I would guess since it's going thru a transistor you don't get a good reading.

[Hupa]

Mike, transistor where? In the ecu?
Anyway, good news! I got the car running!
Bad news is, I don’t really understand why, and I’d really like to learn this.
Well, lets start with the test updates: So, believing the ecu provided ground through pin 85 (though multimeter tells otherwise), I postulated that the relay must work, and indeed, when I connected the amp-pen from either side of fuse 1 to ground, it illuminates while cranking. So the relay does relay (pun) and the pump(s) must be getting at least voltage, and also current since resistance from pin 87 to ground was small/moderate. But the engine wouldn’t start. I was out of other options, so I jumped pin 30 to pin 87. Instantly, the was quite a loud hissing sound from the top side of the engine (what is it?). Jumped in the car, cranked, it started, it died, started again, gave some throttle, let go of the pedal and engine died. Took out the jump wire, had a cigarrette and tried again. This time I just let the engine idle 5 minutes, shut it off, took out the jump wire and replaced the relay. Cranked, it started straight away, but would almost die out if I touched the gas pedal. Let it idle 5minutes, and it started redponding to gas pedal… run like in the good ole days. I let it run 25 minutes, idling, giving throttle… response feels normal… though I didnt rev it above 4000rpm.

My theory (please challenge this) is that the internal relay switch contacts 30 to 87 are corroded/oxidized, and would only start passing enough amps when the alternator kicked in with 14,x Volts.
The ecu with the empty ”constant power” pin 18 is still a mystery. I would be very happy to understand what it means, and how anything works like this.

[Mike W.]

Mike, transistor where? In the ecu?

Yes.

Anyway, good news! I got the car running!

Excellent!

My theory (please challenge this) is that the internal relay switch contacts 30 to 87 are corroded/oxidized, and would only start passing enough amps when the alternator kicked in with 14,x Volts.

I doubt it. I think it's all electronics inside the ECU, no discrete relay.

The ecu with the empty ”constant power” pin 18 is still a mystery. I would be very happy to understand what it means, and how anything works like this.

Someone else was asking about constant power recently and I don't see anything on the ETM, nor have I ever run across anything that indicated constant power.

[Hupa]

Mike, in response to my comment ”My theory (please challenge this) is that the internal relay switch contacts 30 to 87 are corroded/oxidized, and would only start passing enough amps when the alternator kicked in with 14,x Volts.”

You replied (sorry, can’t use the quotes)

”I doubt it. I think it's all electronics inside the ECU, no discrete relay.”

Here, i was talking about that possibly the FUEL PUMP RELAY contacts that upon relay activation make continuity from relay pin 30 (12V) to relay pin 87 (going fuel pump(s)) inside the relay may have oxidized, and have poor contact under load. When I got the engine running with jump wire to build fuel pressure, the car then started also with the relay in its place (jump wire removed ofcourse), and the higher (than battery) alternator voltage was enough to maintain current to fuel pump (inspite the supposedly poor contact inside the relay). I mean, before the jump-wire start I could crank the engine endlessly and it would not fire up. And now it fires right up with the original relay like there never was a problem. So, something changed. I believe that now there is residual fuel pressure as the engine and pump have been running, and when I crank it, she starts with the fuel pressure readily available, and the relay gets the higher alternator voltage that is able to push enough amps through the semi-bad fuel pump relay. So, just a theory so far.

Well, I’ll get a new relay. Then I can test this theory ”on bench” by measuring the voltage drop accross the 30-87 connection under load, e.g. the amp-pen (with substantial current going through the contact). If my theory is correct, the new relay should have almost zero voltage drop accross pins 30-87, and the old one then something larger.. don’t know.. maybe 1V voltage drop could already prevent the fuel pump’s normal operation? I did measure this old relay’s resistance over the actuated contact 30-87. It was small, but that still doesn’t quarantee it stays small when substantial current is pushing through. We’ll see, I’ll report my findings.
Thanks again for the help, Mike!

[///M]

Hi Hupa 
With regard to your issues. What year is your M535i?
The DME for Motronic in Euro models does not have a pin on number 18. It is empty. 12v+ is from pin 87 on the main relay. Pin 87 also provides 12v+ to the fuel pump 86
You may have the incorrect main relay. It is specific to the Motronic. There are relays that are similar but not the same. They are cheaper to buy.

[Hupa]

Hi M,

my car is a standard euro market 1987 M535i as far as I know. It is for the swedes, so there is some shit in the exhaust manifold, don’t know if ecu is different? So, there might not be an issue with the Motronic  connector after all? I just googled m535i ecu pinout to see which ecu pin slot I should check for continuity from fuel pump relay pin 85 slot (I was looking for ground for relay pin 86 current to go to). Found the correct pin from ecu (20), and assumed the pinout was correct. Anyway, I checked the grounds from the ecu connector while I was crawling in there, and also this ”continuous power” which was supposed to be at pin 18 slot. But the connector had an empty slot there. I’m referring to ecu pinouts here:

https://mofler.com/uploads/attachments/ ... 27f6c32e5f

and here:

http://e28-535i.com/upload/e28_Harness_connections.pdf

So, is there another pinout different from these for 535i?

[Hupa]

..and, there’s no reason to suspect the relay is wrong, it has been there workin perfectly for the 5 years I’ve had the car and prolly decades before that. It looks old, ugly green with pins 85 86 85 30. Looks like the green ones in this post photo (can’t see the pin numbers on them, but they are visually similar). Anyway, would it matter if the relay is genuine? The operating principle is very simple, and identical if the same pin numbers are found, correct?

viewtopic.php?t=138596

[///M]

For a 1987 M535i Euro there is no pin in 18 as I said. Most of the electrical information comes from BMW North America as stated on the ETM. It is correct for US models 
There are some differences is Euro models /years. Use as a guide.

With regard to the relay I said "Main relay" or as BMW refer to  it relay 2 it has 5 pins 85,86 30 and two 87. Often called the DME relay on sites in the US. This is the specific relay I referred to. It is the 12v+ to the DME (ECU) pin 35 and the 12v+ to pin 86 of the fuel pump relay.

The fuel pump relay is a standard relay (The green one you referred to, BMW call it relay 1) same as some of the other relays in the fusebox which can be swapped for the fuel pump relay if needed.

[Federico]

 


 


 

 

[///M]

Well done Federico. There it is in Black and white and German. "Relais 1 and 2" some German words are easy to translate.

[Hupa]

Thanks, M

very well, I need to be more careful that my references are for the specific car. I’m still happy to get the car running for summer. I’ll go through the fuel pump connections and main grounds, as they have been the root for many problems in my other cars as well.

BTW, does it make any difference if these ecu boxes are bolted to the frame (ground) or not to function correctly? Like the other (wrong) ecu pinout link has this ”pin 5 - SHIELD GROUND REF POINT”? Don’t know what this really means, but gives an idea it might be the ecu casing. Or?
 

[///M]

If you look at the wiring diagram you can see the grounds. The word Masse in German if you check  in a translator  you may be puzzled as I was when I first came on the word. There are primary  results which seemed meaningless. Masse is ground.
B is Masse electronik. The electronics have their own ground. The crankshaft speed and reference sensors  have a metal shielding around the wires to reduce interference as they are AC signals. The shielding is connected to the ground point which is B Masse electronik. You can check the diagram.

Having good clean grounds is essential in older cars especially in damp climates, as corrosion causes high resistance. If you look at the relays you can see a line of contact on the pins, taking the relay out of the socket and putting it back in can make it work because it makes a cleaner contact.

[Hupa]

Federico, 

suddenly everything makes perfect sense, thank you! 
I swear my car doesn’t have that many wires in it, I swear!  

I really have to take my hat off, if you guys can make something out of that wire drawing. Well, perhaps there is something more to map where everything is physically in the car. I remember seeing an ”electical trouble shooting” manual or something like that. It had wiring diagrams for dummies that even I could understand. I’m sure I downloaded it but can’t find it anymore… maybe it’s in my old phone.

///M, yes, my bad, you talked about the main relay. Don’t see why that would be wrong either. Everything looks old, stock and original in/around the fuse box… well everything is possible, but the car worked perfectly until these fuel pump issues. And, yes, I noticed the oxidized colour on the fp relay pins. Cleaned them, as well as the slots. Good idea to repeat that for all of them. Well, it is always nice to learn (though Federico reminded me how little I know), and it’s great that you guys are here to help. Thank you very much!

[Federico]

 Well, perhaps there is something more to map where everything is physically in the car. I remember seeing an ”electical trouble shooting” manual or something like that. It had wiring diagrams for dummies that even I could understand. I’m sure I downloaded it but can’t find it anymore… maybe it’s in my old phone.
 

 

 

The ETM's and the German Schaltpläne can be found here: E28 Wiki
There are indeed main wiring diagrams in the Schaltpläne, like the one below for the 528i.
The path for the instrument cluster lights is so ridiculous, I had to trace it on:


 

[Hupa]

Thanks Federico. It almost feels it would have been  easier for you to just follow the wires.

My starting issue is not quite fixed yet. I need to idle the engine 5-10minutes, before it accepts throttle. Once the engine is warm, it works normally. This somehow feels the problem could still be elsewhere. But since the car fired right up with the jump wire over relay, I still believe it’s related to the f.pump. I measured voltage drops with running engine, starting from battery+: Relay out 40mV, pump fuse in 170mV, pump fuse out 180mV. These feel like normal voltage drops, so the problem may be at the pump connectors and/or ground. I’ll back her up on ramps tomorrow and crawl under her, assuming that’s where the pump is. I do like lady being on top.

[Mike W.]

Just put a fuel pressure gauge on it, that will tell you more than anything else. You don't have to get a fancy automotive one, I made one years ago now from a basic water pressure gauge from the hardware store, a barbed fitting, a Tee and a little bit of hose. Cost me less than $10 at the time.

If you're fuel pressure is good and stays where it should be, your fuel pump circuit is good.

[JohnH]

Just put a fuel pressure gauge on it, that will tell you more than anything else. You don't have to get a fancy automotive one, I made one years ago now from a basic water pressure gauge from the hardware store, a barbed fitting, a Tee and a little bit of hose. Cost me less than $10 at the time.

If you're fuel pressure is good and stays where it should be, your fuel pump circuit is good.

 

 

Me too. Here's mine.

 

 

[Hupa]

Mike & John, good tips! Thank you. So, effectively you’re just tapping a pressure gauge to the side of the fuel line, correct? This sounds something within my set of skills. What kind of pressures should I expect at idle/eleveted RPM, if everything is ok?

I’m a bit worried the problem could be somewhere else, because the cold motor won’t take any throttle, almost dies out when touching the pedal. But as the engine warms up, throttle response becomes normal again. Does this hint to sonething specific? I don’t see how engine cold/warm would affect anything in the fuel system. Well, speculation is just a sorry excuse for not getting my hands dirty.

[JohnH]

Detach the fuel inlet pipe where it connects to the the injector rail. #14 below.
 
I think you are looking for 42.5 psi.

With regard to your hard start, it sound like the coolant temperature sensor (959 in the motronic diagram) has failed and is saying the engine has reached its operational temperature when its actually cold. Try disconnecting it before starting. If it starts ok and then the engine gets a bit rough/rich as it warms up, this would confirm that the sensor has failed. 
 

[Hupa]

John, thanks for the tips! 

Today, I cleared my small garage like 1.5 hours just to get enough room to drive the car forward, so I could back the car on the ramps, while keeping clearance to close the garage door. A larger workspace would be nice, but this is €27/month, so almost free.

BTW, someone said the fuel pump(s) don’t prime in e28 535i with power on, but I have a distinct memory before these problems of a humming/buzzing sound when turning the key to ”on” position to get power for the stereo (always took it as the fuel pump). Now this sound is gone. What could that sound be if not the fuel pump(s)? I heard quite a loud buzzing sound from the engine bay that could sound similar inside the cabin, when I jumpwired the relay. Another strange thing, when measuring the fuel pump relay slots for continuity, the resistance from pin 87 slot (to the pumps) was 12ohms with power off, and something very small with power on (the relay was off when I measured). The car doesn’t have cat and the wire diagram shows nothing ”extra” besides the pumps in circuit from pin 87 to battery minus terminal 

I’ll do tests you suggested. But now the wife keeps calling me every 5 minutes for some other pressing chores.

[Hupa]

John, guys…

I checkd the pump contacts under the car & fuel pressure with a gauge. Nothing off there. I then took out the coolant temp sensor (blue one) connector (a motherf**er to get out). Now the car starts right away as I turn the key, but sputters and wont take any throttle. Before it was very difficult to get it running and would take throttle only when engine was fully warm.

Measured the sensor resistance, 1200ohm @10C (~50F). With power on, voltage over the sensor connector is a bit over 4V? Do these values sound good/bad? Googling, the resistance at current temp sounds a bit low. I’m finding resistance should be 2000-3000ohm @68 ”degrees” (I assume this is in F, as it is 20C… so a ”standard” temp.) and 300-400ohm @176(F?), which equals 80C. In this light, the resistance seems to be on the low side, I expect it should be even higher than 3000ohm as it is now colder than 68F or 20C. I just install a new sensor and stop speculating?

[Hupa]

Update:

SUCCESS!  I took out the sensor, wasn’t too filthy but still I wiped it a bit with a rag. Re-measured resistane, 2500ohm… right in the spec for cold (if my values above go for this car). I was like what the hell, screwed the sensor back in, put the connector back, and now it fires straight up cold with normal throttle respone. So it was the temp-sensor. But I can’t get my head around why taking the sensor off or cleaning it corrected anything I mean, if the car hasn’t been running for 24hours, then every part of the car, including the sensor, is at ambient temp., and any dirt/slime on the sensor has nothing to do with its temp. So what difference did cleaning it have? Maybe nothing, and it was something that happened when removing it?! But I quess it can’t be any grounding thing from sensor to housing, since it has 2 wires and one must be ground, correct. Mysterious… mysterious.