Do CEL Codes Confirm My ECU is Working? (Diagnosing Fuel Pump Power)

[marvin5881]

The connector that plugs into the fuel pump assembly isn't receiving any power, I've read online that this could be due to the DME because the DME receives crankshaft sensor data and sends it from the DME Relay and from there to the Fuel Pump Relay. I put in a jump wire and the fuel pump connector got power. I've already replaced the fuel pump relay. If I can rule out the DME, then I suppose that just leaves the crankshaft sensor (According to posts I've found online).

[ed323i]

DME relay is also important.

Ignition on -> DME wakes up -> DME sends ground signal to DME relay -> DME relay activates and sends +12V to all the injection electronics under the hood AND back to another 12V input of the DME -> DME now fully wakes up into run-mode and tries and find a good crankshaft position sensor signal -> Once it has a good signal and can at least find, say 100-200 rpm, then it will start firing the coils, injectors and it will turn on the fuel pump relay by sending a ground signal to the fuel pump relay..

To determine if DME relay is activated, turn on ignition and see if you can find +12V on the coil connectors or the ICV connector, using a multimeter. If you find +12V, then the DME relay is working. If not, then the DME relay is bad, or the DME is unable to drive the DME relay because it hasn't got a good ground input (or a defect in the DME).

[marvin5881]

Watched a video on checking the ICV connector seems like I'd have to go through a lot of work for that, I also checked videos on finding the coil connectors but couldn't find anything pertaining to the M42 4 cylinder. I assume by your answer, CEL codes are not enough to confirm that a DME relay is working?

[ed323i]

I'm not 100% certain about the CEL codes. It means the DME gets power, but I'm not sure if receives the +12V from the DME relay that activates the larger circuits in the DME, to allow the engine to run.

Coil pack connector is very easy on the M42, easier than on the M50/M52 where you have to remove a beauty cover of the engine to get to them.
The coil pack is located on the passenger side shock tower if I'm correct. Just unplug the big round connector and try and find +12V on one of the pins of the wiring harness connector. Use multimeter, one lead on one of the pins consecutively and one lead on one of the three shock tower bolts (great ground points). If one of the pins shows +12V, then you're certain the DME relay is supplying power to everything. If not, then the DME relay isn't activated, which can have several causes. A bad relay being the easiest and most probable cause. I think there is another relay in the relay box of the E36 with the same specs as the DME relay. So you can also just swap it to see what happens.

Good luck!

[marvin5881]

So I was only able to get 12V on a single pin in the wire harness connector, all of the rest of them showed nothing. But I noticed when I disconnected the DME Relay, I still was receiving 12V on it (I turned the car on and off again too). Pretty odd, not sure what to make of it.

[ed323i]

You're right.. I just checked Bentley manual. With 4-cylinder engines, the coils get power directly from the ignition switch. It doesn't go through the DME relay... Odd..
Hmm, now I'm double checking, also with the 6-cylinder, it's the same. The coils get power directly from the ignition switch.. So, I've been giving some bad advice lately :-( ..

Easier test, while looking at the wiring diagram:
The DME relay, when activated, sends +12V to the injectors.. But those aren't very easy to test. And it sends +12V to pin 86 of the oxygen sensor relay.
Also, it seems with the 4-cylinders there is no 2-phase DME activation...

Hmm, I'll be looking through the diagram some more and see if I can come up with something useful..

Update: Have looked it up.. There is a two-phase DME activation, also with the early M42 models.
The DME relay also sends power to the ICV. But, yeah, a bit difficult to get to, according to https://www.bimmerforums.com/forum/s...-ICV-clean-DIY .

So, that would be the easiest test. With ignition on, the DME relay should send power to pin 86 of the oxygen sensor relay (take relay out and 'mirror' it down to the socket), and see if you get +12V there.

Here is where the oxygen sensor relay is found:
http://bmwfans.info/parts-catalog/E3...n_sensor_k6309

Found another that would probably work: The Air flow meter, pin 1, measure on the wiring harness. Should show +5V, and comes directly from the DME. And I assume the DME can only send this 5V when it gets its power from the DME relay (but not 100% certain). So, to be 100% sure, best to use the oxygen sensor relay 12V to test if DME relay is activated and working properly.

Update 2:
Next steps.. If you can confirm the DME relay is working, then it's very probably the crankshaft position sensor.. Or a bad DME (not very likely), or bad wiring somewhere..
If the DME relay isn't working, then try and swap the DME relay first. Still no go, then it might be wiring or bad DME.

[marvin5881]

So I carried out the test, I put the positive end in pin 86 and grounded the other pin and sure enough I got +12V. So that confirms that my DME relay is working fine. I got another DME from the junkyard and the results don't change, my car was produced in October 1994 so it has EWSI so I just did plug and play with it. I'll try the crankshaft sensor next.

[marvin5881]

So I haven't gotten my hands on a new crankshaft sensor, but I should also note that this car hasn't started in 13 years. I wonder if that changes any troubleshooting steps? I also checked today and confirmed that the engine is getting spark, could that rule out the crankshaft sensor?

[ed323i]

That you've confirmed the working of the DME relay, AND that you have spark, is great news.
That means a lot of stuff is working just fine, and it should mean the DME is also trying to activate the fuel pump relay. And, indeed, it means the crankshaft (or camshaft) sensor is working sufficiently for the DME to interpret the RPM and decide on when to fire the coils (and probably injectors).

It also means, that if you would just supply +12V to the fuel pump, the engine should run.
Somehow the DME can't get the fuel pump relay to work.. If you've replaced both the DME relay and fuel pump relay, that means what's left is wiring (perhaps a simple fuse, #18), or a DME that has a bad solder internally and just can't supply a ground signal to the fuel pump relay.

Or, you can test with starting fluid in the rubber intake boot (you can remove and disconnect MAF sensor without any problem, DME can run just fine without MAF sensor), while depressing the gas pedal a little (or opening throttle body butterfly using your hands on the lever)..

I'll look up what voltages and ground signals you should have on your fuel pump relay socket. That will help diagnosing the problem in the wiring or DME.
You can go ahead and remove the fuel pump relay and measure on which pins you get +12V with ignition on and on which pin you can measure ground (multimeter in Ohm mode, measure between one of the pins and a good ground point like one of the 3 bolts on the shock tower; 0-1 Ohm means a good ground). If possible also re-measure while cranking. An extra ground pin should appear if I'm correct.

Update, OK, I looked it up: (page ELE-74 of Bentley manual)
1995 318i should have: (fuel pump relay socket)
+12V on pin 30 (hot at all times)
Ground signal on pin 85 (when DME sees it fit to activate fuel pump relay, as in when it receives say 100 rpm from crankshaft pos. sensor)
+12V on pin 86 (when DME relay is activated) (so this is another way to test DME relay)
+12V on pin 87, going to fuel pump (when fuel pump relay is activated, so can't be measured on the socket when relay is out)

So, three tests:
* Test if you see +12V on pin 30 always
* Test if you see +12V on pin 86 with ignition on (which turns DME relay on)
* Finally, test if you see a ground signal on pin 85 while cranking

[marvin5881]

So I'm getting +12v on pin 30 always, I'm also getting +12v on pin 86 with ignition on, and I wasn't able to get someone to look at the multi-meter to test for ground signal while cranking, but pin 85 has a good ground signal without key in ignition. I've noted that the ground signal on pin 85 disappears when I disconnect the DME. I've also hot wired pin 30 to 87 to get the fuel pump to work, but the engine still wouldn't turn over. I've tried my original DME, and a DME from the junkyard also. Junkyard DME seems to work just fine with my car, I think it's because it has EWSI (manufactured October '94)

[ed323i]

Odd that pin 85 has permanent ground. That would mean that the fuel pump would be running always once the DME relay is activated (ignition on).. Strange..

OK, when you hot-wired the fuel pump, can you hear the fuel pump buzzing? Can you measure +12V DC at the fuel pump connector?

You noticed that the coils are sparking correctly. This probably means most important sensors and the DME are okay.
To be sure, can you try and crank while spraying starter fluid into the rubber boot (between throttle body and MAF sensor; you can safely remove and disconnect MAF for this test) or directly in the throttle body, while opening the throttle butterfly valve a bit. You'd probably need another person to help though. Does it combust, does the engine run for a little while?

If so, you can be quite certain that the ignition is working fine (coils, crankshaft pos. sensor, probably also camshaft position sensor, DME, DME relay)..

I read the car hasn't run in 13 years (that's very long). Are you sure there is fresh gasoline in the tank? You'd have to discard all gasoline first and get fresh gasoline to have a chance of getting it to run. Also, I think you can expect quite some gunk in the injectors, possibly blocking the nozzles.

Good luck!

[marvin5881]

So this time I opened the throttle butterfly sprayed the fluid in and it definitely combusted, the engine has never sounded so close to turning over. After trying 4 times I also heard a loud pop, and smoke came out of the throttle body so I decided to stop after that. There is fresh gasoline in the tank that I recently added, the tank was left sitting for 13 years with no gas in it btw. After jumping the fuel pump, I removed one of the hoses at the top of the fuel pump assembly and gas sprayed out so fuel is definitely moving. I can hear what sounds like fuel moving in the engine bay and that sound only happens when the fuel pump is jumped so it seems like the injectors are blocked. So I have two things to resolve before the car is operational, fuel pump only working when jumped and a fuel blockage. Also quick reminder that fuel pump relay is new.

[ed323i]

OK.. That's great news, that you got combustion.. It really means the DME, crankshaft/camshaft position sensors, coils and DME relay are working (at least mostly) fine.
And, in normal circumstances, the DME should activate the fuel pump relay.. So something is going wrong there. The DME has a proper ground signal to send to the fuel pump relay, otherwise it couldn't activate the DME relay.

I think your conclusion is correct: There are two problems:
1. The fuel pump relay doesn't get triggered, even when the DME relay is fine and the DME is definitely picking up a good signal from the crank/cam-shaft position sensor (otherwise no spark)
2. Somehow the fuel in the rail/injectors isn't good enough to combust

Next steps:
1. Can you bench-test the fuel pump relay. As in, supply +12V DC (battery plus) to pin 86 and a good ground (battery minus) to pin 85. You should hear a click, and you should then be able to measure 0-1 Ohm between pins 30 and 87.
2. I suspect there isn't enough fuel pressure in the fuel rail. This can be caused by (in order of probability, at least how I estimate it):
a. Fuel pressure regulator died
b. Fuel pump might have died in those 13 (not so lucky number ;-) ) years. It still pumps but doesn't build sufficient pressure.
c. Fuel filter severely clogged
Theoretically, the injectors could have clogged really badly, but I think chances of having this happen to all 4 injectors is very slim (if even 1 injector would flow OK, then you would hear combustion)

Best course of action would be to measure fuel pressure in the rail. The M42 needs 3.0 bar at WOT and 2.3 bar at idle (at idle there is a high intake manifold vacuum of +/- -0.7 bar, so FPR compensates for this to keep relative pressure between fuel rail and intake manifold at a stable 3.0 bar; M50/M52 use 0.5 bar higher pressure).

I think we're getting close to the solution now.. Good luck!

[marvin5881]

I started working on checking out the fuel rail system, and I decided to test the crankshaft sensor while I had the upper intake manifold off. The sensor is reading 550 ohms so that can be ruled out. I also checked my original fuel pump relay and the one I bought at the store and both of them click fine and get 0 ohms resistance when connected to a battery. I'm really starting to wonder what the culprit could be with the relays not activating..
As of right now I'm looking for videos on removing the fuel rail. I'm trying to figure out what that hissing noise is in the engine bay, maybe that can point me to the problem

[ed323i]

With regards to problem 2 (fuel not okay to combust), you could also disconnect the fuel return hose from the fuel rail, and connecting a small hose that runs to a clean glass jar. Then activate the fuel pump and see if clear fuel is running through. The fuel pump should pump a certain volume per minute of gasoline, the fuel pressure regulator makes sure the pressure doesn't go over 3.0 bar and releases the rest of the fuel (which is everything with engine not running, and after pressure has built up in a few seconds) through the return hose.

Sometimes (doesn't happen often, but not using a car for 13 years helps) the fuel pressure regulator dies in such a way that it doesn't let any fuel through. In that case you won't see fuel coming out of the return line of the fuel rail, and pressure will build up to what the fuel pump can supply, which in my case, in an E28 528i, was 7 bar..
If gasoline flows out, then you can do the fuel pump delivery volume test. Make sure you use a 2 quart jar/container to catch the gasoline. Bentley specifies 0.875 liter per 30 seconds.

Still, the best test, that will tell almost everything, is connecting a fuel pressure gauge. Should show 3.0 bar with engine not running, or 2.3 bar with engine idling, or, again 3.0 bar with engine at WOT.

With regards to problem 1, fuel pump relay not receiving proper +12V or ground on pins 86 and 85.
We can conclude that socket pins 30 and 87 work just fine, as when you connect those, the fuel pump starts running.
So the problem has to be in either pin 86 (+12V DC) or pin 85 (ground from the DME)..

Do you remember if you were getting a very good ground on pin 85, like in only 0-1 Ohm resistance when compared to a good ground point?
Theoretically there could also be a problem with the +12V DC on pin 86, in that the connection is really poor, and that when you put on any load, even the few mA a relay needs, the voltage drops to virtually zero..

So, what you could do, is (try to work safely, not causing shorts) working with the fuel pump relay, but not in the socket, but close to it. Connect a wire between socket pin 86 and relay pin 86 (supplying +12V DC to the relay) and then connecting a wire between a good ground point and relay pin 85. It should click then, unless there is something wrong with the 12V on pin 86 of the socket. And the 2nd test in reverse, connecting socket pin 85 with relay pin 85, and connecting a wire with a known good +12V DC source (can be directly from the battery) to the relay pin 86. Then you'll know some more. Again, work carefully, making sure not to cause a short.

[Thomas525]

I think you should jump the fuel pump relay, and see if you can start the engine normally. You have not done that. That's more critical . If fuel was the problem, the engine should have started and run for a few seconds with the starter fluid you used earlier.

Oftentimes, the problem is bore wash caused by overfueling due to a bad coolant temperature sensor. First, unplug this sensor - it will be on the cylinder head somewhere and might be blue tipped. Second, pull the dme relay and crank the engine 3 turns to vent all excess fuel out. Then, replace the relay, energise the fuel rail with the relay jump, then try to start the engine. If it doesn't catch hold down the throttle and start the engine. If this doesn't work too, then remove the fuel rail and spray wd40 down the port holes. Replace the rail and do a slow crank 3 turns without the dme relay to ensure the chambers are fully oiled. There will be zero bore wash after that. Then replace the relay, jump the fuel relay, and start the engine.

After the engine starts, let it run for 5 minutes and watch for anything abnormal. Assuming it is normal, while the engine is running, yank out the jump and replace the fuel relay quickly. The engine should continue running. Wait another 2 minutes, then shut down, and restart with the fuel relay in there.

[ed323i]

He already tried that, post 10: "I've also hot wired pin 30 to 87 to get the fuel pump to work, but the engine still wouldn't turn over"..

So, my conclusion is the ignition part is working great (as it almost starts with started fluid), but there is something wrong with the fuel. My guess is the FPR has deteriorated too much in those 13 years. Fuel pressure test is logical next step, I'd say.

But your test is interesting and might show something.

[Thomas525]

It is not supposed to almost start with starter fluid. It is supposed to start. You don't check fuel pressure when you have spark and the car still does not start with starter fluid. He has lost compression due to cylinder wall oil being washed off by overfuelling caused by a bad coolant temp sensor, which is a well known failure. That's why it does not start with spark and starter fluid. Op please follow my suggestions in sequence it will barely take you 15 minutes total.

[Thomas525]

Op, if you need to go the wd40 route, don't spray too little but don't worry about spraying too much oil. You will be slow cranking after that and it will be fine the excess oil after coating the walls will go out through an open intake or exhaust valve. And if it does not work with wd40 and a jumped fuel relay, then try starter fluid. If it doesn't work, hit it with wd40 again then starter fluid.

I'm not sure if you can easily access your fuel rail to lift it up and expose the ports easily. If you can't, then you'll need to remove the spark plugs and spray wd40 inside. This is called reoiling the chambers and you'll usually read how you need to put in 10ccs of engine or transmission oil into each chamber. However once I was in a jam and just tried with wd40 it worked perfectly. Much much easier.

[ed323i]

Seems like a good test, but the FPR cause is still also an option (and will cause the exact same symptoms if the pressure is too high: flooding).
While you're at it, you can read the coolant sensor resistance, the blue one that's easier to reach, more on the radiator side than on the firewall side.
Here is the chart: http://www.bosch-motorsport.de/conte...2782569739.pdf

The black one [6] is for the temp gauge, and the blue one [8] is for the DME.
See http://www.realoem.com/bmw/enUS/show...diagId=11_1505

Good luck!

[Thomas525]

Seems like a good test, but the FPR cause is still also an option (and will cause the exact same symptoms if the pressure is too high: flooding).

That is not how fprs fail 99.9% of the time. When they fail, they don't hold fuel pressure are the rail. The gas drains through a ruptured diaphragm, rail pressure drops, and is too low to support a quick startup. Supporting quick startups is one of the reasons why bmw moved the regulator to the fuel rail from the fuel pump.

I'm sorry Ed you do outstanding research and that car of yours is very drool worthy but you are mistaken about many basic things.

[ed323i]

Thomas,

Do we know each other? It seems like it, because I get the impression you're trying to show the world and yourself that I don't understand nothing about cars, and you do. Projection? Doubting your own knowledge and trying to prove it by down-talking others? I don't know what's going on, but I'll leave you to it to help Marvin out, and I'll leave the forum for a couple of weeks. I used to have fun helping people and learning in the process, but with your contributions, which almost seem like personal attacks, you've taken the fun out of it. You're talking in 'absolutes', like you know it all and every other opinion is nonsense, like: "I can say that this is completely wrong." and "This is, again, completely wrong" (other topic; I agree to disagree) and "That is not how fprs fail 99.9% of the time" (have you done statistical analysis on 1000's of cases?).

All I wrote is that if the FPR fails in the way that the pressure gets too high, it will also flood the engine, just like a bad coolant sensor.
If the FPR fails in a low-pressure way, the engine won't combust because the gasoline doesn't get atomized properly and there just isn't enough fuel injected.
So, all I'm saying is that the FPR can still be causing Marvin's problems, just like a bad coolant sensor COULD (no absolutes) be causing his symptoms, or a worn-out fuel pump.

I'm quite certain that with all your experience you'll help Marvin resolve the problem (I mean this, no cynicism). I acknowledge your experience and knowledge. But please stop trying to prove it by talking in absolutes and by down-talking others who are sincerely trying to help using their own experience and knowledge.

[Thomas525]

Ed,

You probably did not realise that I was being polite by saying earlier that 99.9% of fuel pressure regulators fail in the way that you suggested they do. No fuel pressure regulator fails in such a way that it has HIGHER rail pressure than their rating.

No we don't know each other. I'm sorry that the threads with headlines that caught my attention happened to involve you giving generally questionable advice or at least questionable probabilities of occurrence of the various possibilities. What am I supposed to do, not respond to avoid offending you? Should I have pmed the poster to avoid hurting your feelings?

I speak in absolutes, or near absolutes, because I know what I'm talking about when I talk about it. I've read, I've studied, and struggled with that issue on my car most likely more than once in the past 10 years.

If you drive a bmw, even an old bmw, you need to have the self confidence to handle robust criticism. You also need to have the self confidence to accept that you're wrong on many issues you thought you were right on. It takes self confidence to handle that situation without going to pieces - or leaving a forum in a huff.

How do you get that, or recreate that, core visceral, manic, trumpian self confidence ? It is surprisingly simple. You go for a WOT run in your ride on the highway when there's low traffic , roads are dry, and your tyres are good. Works every single time.

[Thomas525]

Marvin, there are 4 things simultaneously needed for successful combustion. Fuel. Air. Spark. Compression. (With carby engines, there was a fifth one - timing).

We always forget the last one - compression. Until that becomes the reason for your shutdown lol.

That's the most likely issue for you here. I have already detailed how you can check that and fix that.

It is of course possible that you have an issue with your fuel system wiring too. But get the engine started first. Then see if your wiring is still the problem. If it is, then attack that. Especially when you're dealing with a car that has not been started in 13 years, you need to aggressively compartmentalise and proceed in order of importance. That's because, the longer a car has not been used, the higher the chances that more than one problem is at play in a shutdown. And whenever multiple problems could be at play, the best approach is to compartmentalise and attack in the order of most important to least.

p.s. Have you looked at your fuel pump fuse in the fusebox ?

[marvin5881]

About a month ago (when I originally planned on this repair project taking a week) I put a tablespoon of oil down each cylinder wall, I put it straight down as opposed to letting it coat the walls of the spark plug well because I thought that might effect the spark plugs, not sure if that mattered. From my understanding with the way I did it the oil would fall directly onto the pistons and I assume the oil would make its way to the cylinder walls. Also would the cylinder walls dry in a month? I'm going to go ahead and wait for a response about the cylinder walls since last time I took out the spark plugs I had to borrow a spark plug socket but this time I'll have to buy one.

Also, just so anyone helping is 100% certain of how trying to start with starting fluid went, I opened the throttle butterfly and gave it two spritzes of starting fluid. What would happen is I hear the engine coming to life as the fluid combusts then I hear a big bang and I see a bright red flash of combustion coming out of the throttle body followed by lots of smoke.

I tested the Coolant Temperature Sensor, I got good resistance about 2500 ohms which is about the engine when it's off. Also I've looked at the Fuel Pump Fuse and it looked fine to me. I put it in the horn fuse slot and it worked, also I when I would take the fuse out, jumping the fuel pump would no longer work so I think the fuse's good.

This afternoon I'll check the fuel hoses, rail, and the injectors to see if everything is fine since I currently have the upper intake manifold off (because I only hear the engine about to turn over with starting fluid), then I'll reinstall the upper manifold and crank the engine with the coolant temp sensor disconnected also doing the procedure to remove any bore wash.