ECU swapped, coils swapped - STILL NO IGNITION ON 4

[ArtemLepilov]

Today i pulled everything off the engine - namely the harness will try to find a cause for the misfire problem. This really pisses me off... Does anyone know what else I can do?

-Artem

[granit_silber]

Today i pulled everything off the engine - namely the harness will try to find a cause for the misfire problem. This really pisses me off... Does anyone know what else I can do?

-Artem

Could the spark wire be bad?
-ashley

[ArtemLepilov]

Possibly - thats why I took the wiring harness off - im going to check and see... Unfortunately I need to do a hell of a lot of work in astronautics at the same time and ship a lot of items... This is a very very bad time for me... This week is going to be hell.

-Artem

[mholbrook]

Did you try moving coil from number 4 to another cyl to see if you just have a bad coil?

[winfred]

did you check that you were getting the signal to that fuel injector?, the power should be constant with the key on and the computer triggers it with a pulsed ground signal

[ArtemLepilov]

Are both the coil and the fuel injector work together with the same pulse at the same time?

-Artem

[winfred]

no but one can shut down the other


Are both the coil and the fuel injector work together with the same pulse at the same time?

-Artem

[ArtemLepilov]

so, when there is no continued circuit on the fuel injector, there is no spark? Is that how it works?

-Artem

[ArtemLepilov]

Note - some people suggested that a faulty crank position sensor can cause that problem. Is that even plausible? please let me know - I really need to get the car running asap!!!

-Artem

[Rustam]

Note - some people suggested that a faulty crank position sensor can cause that problem. Is that even plausible? please let me know - I really need to get the car running asap!!!

-Artem

I don't find this plausible. If the crank position sensor was faulty you would not be able to start the car. And then - why necessarily 4th cylinder? There must be something else.

[ArtemLepilov]

Thats what Im thinking - my problem seems to be really weird. Just checked the wiring harness thoroughly - it is in good shape, continuity test shows all good. ECU is brand new from a different car, which worked in it fine, and my ECU didnt seem to be bad anyway.

I got no idea what it can be - everything is fine and is there, it just isnt producing any spark on cylinder 4... People suggested crankshaft pos sensor and camshaft position sensor. Can that possibly cause it - i head if the voltage is different on the signal form the CPS it may misfire on that particular cylinder whenever the voltage is different at that particular time - its a longshot... but i have to test EVERYTHING POSSIBLE... Does anyone know for sure?? I am seriously lost at this point...

-Artem

[Rustam]

Today i pulled everything off the engine - namely the harness will try to find a cause for the misfire problem. This really pisses me off... Does anyone know what else I can do?

-Artem

Have you actually seen no spark produced in the spark plug connected to the coil but out of the engine during cranking?

[Rustam]

Thats what Im thinking - my problem seems to be really weird. Just checked the wiring harness thoroughly - it is in good shape, continuity test shows all good. ECU is brand new from a different car, which worked in it fine, and my ECU didnt seem to be bad anyway.

I got no idea what it can be - everything is fine and is there, it just isnt producing any spark on cylinder 4... People suggested crankshaft pos sensor and camshaft position sensor. Can that possibly cause it - i head if the voltage is different on the signal form the CPS it may misfire on that particular cylinder whenever the voltage is different at that particular time - its a longshot... but i have to test EVERYTHING POSSIBLE... Does anyone know for sure?? I am seriously lost at this point...

-Artem

take a look at the camshaft position sensor. does it look like it can provide variable signal at any point during its turning or only within a limit if few degrees? if it does not sense all along the circumference then it cannot provide cylinder specific information.

[ArtemLepilov]

Have you actually seen no spark produced in the spark plug connected to the coil but out of the engine during cranking?


Yes - i tested every coil separately at ignition mode, every other coil fired the spark plug - on the same spark plug that i tested other coils on, except #4, swapped the coils around to make sure they worked, the same coil that didnt fire on #4 fired on a different connector. I grounded them on the engine case, and it produced a spark on others.

-Artem

[ArtemLepilov]

take a look at the camshaft position sensor. does it look like it can provide variable signal at any point during its turning or only within a limit if few degrees? if it does not sense all along the circumference then it cannot provide cylinder specific information.


I somehow doubt that the sensor provides a clinder specific information. I will double check tomorrow though..

-Artem

[Rustam]

I somehow doubt that the sensor provides a clinder specific information. I will double check tomorrow though..

-Artem

I doubt that too. The wheel that stimulates the sensor has a small extension that covers no more than 15 degrees. And the sensor itself is not radial:

http://img.autopartswarehouse.com/live/F601027370BOS.JPG

So this tells that the system is unable to provide camshaft position with respect to a cylinder.
_____

If you checked the coils in that fashion and another ECU did not provide a change then the only culprit is the wiring...

[632 Regal]

go back and read Winfreds post.

[Jon K]

Like I said last night in IMs - test the fuel injectors.

[Bill R.]

things that may or may not help... first off the dme won't shut off the spark to an individual cylinder if there's no continued circuit on the fuel injector.

It has a feature in the DME known as the catalytic converter protection on your car which is dme version 3.1.
This feature will shut off the injector to any cylinder that a fault is detected in the coil primary circuit. The later cars which have 3.3 and higher monitor the primary and secondary ignition circuit and shut down the fuel injector if needed,but yours only monitors the primary. This is to keep from dumping raw fuel into the cat.
If you have a injector fault it won't shut down the coil since spark without fuel won't cause a cat to overheat

Next since you have changed the dme and the coils both, have you checked for voltage at terminal 3 of the coil connector with the key on?

All of the coils have 12v at them when the key is on and the dme makes and breaks the ground to trigger the spark. If you look at the harness you'll see that all of the coils have a junction point in the harness where they all meet for the positive supply line. If you don't have 12v at terminal 3 with the key on then back trace it and find out at what point you do.




so, when there is no continued circuit on the fuel injector, there is no spark? Is that how it works?

-Artem

[ArtemLepilov]

things that may or may not help... first off the dme won't shut off the spark to an individual cylinder if there's no continued circuit on the fuel injector.

It has a feature in the DME known as the catalytic converter protection on your car which is dme version 3.1.
This feature will shut off the injector to any cylinder that a fault is detected in the coil primary circuit. The later cars which have 3.3 and higher monitor the primary and secondary ignition circuit and shut down the fuel injector if needed,but yours only monitors the primary. This is to keep from dumping raw fuel into the cat.
If you have a injector fault it won't shut down the coil since spark without fuel won't cause a cat to overheat

Next since you have changed the dme and the coils both, have you checked for voltage at terminal 3 of the coil connector with the key on?

All of the coils have 12v at them when the key is on and the dme makes and breaks the ground to trigger the spark. If you look at the harness you'll see that all of the coils have a junction point in the harness where they all meet for the positive supply line. If you don't have 12v at terminal 3 with the key on then back trace it and find out at what point you do.


Hi Bill. :) I checked all the coils a while ago, as well as checking for continuity - all of the coils were getting power 12V - 13.46 to be exact. The Coils, the DME wiring is fine and the DME is fine. I can test fuel injectors while the DME is on the car.

Meantime Im putting the harness back on the car.

-Artem

[joshua43214]

Since you have verified key power to the coils, I would check for proper ground signal from the DME next while the engine is running by backprobing at the coil, then at the DME if no signal at the coil, this will verify the wiring inbetween. I am not a big fan of continuity checking if its possible to dynamicaly check things, I have been burned on a few jobs where wires passed a continuity check, but failed under load.I believe on your car the DME recieves a crank angle signal from the harmonic balancer, and the big gap is the #1TDC signal, the cam sensor tells the DME whether it is at compression or exhaust stroke. If you have proper injector firing to #4, then the inputs to the DME "should" be ok, since it uses the same inputs for both spark and fuel. The DME counts the teeth on the balancer to know crank angle, the other 5 cylinder work and 4 is in the middle of the firing order, so its unlikely the DME is losing count due to a faulty sensor or broken teeth.

Also, use a multimeters frequency function to verify you are getting similar signals at #4 and another coil, a digital lab scope would be the pefect tool here if you have access to one.

My bet is that you have a bad ground from the DME to the chasis. My memory is that each coil has its own ground from the DME to the chassis rather than a common ground, but take that with a grain of salt, it too hard to remember that kind of detail.

[ArtemLepilov]

Just put it together, started it, injector connector was bad, replaced that - still no spark, coil tests fine, resistance in the coil is good, it fires on other connectors.

I will try to see the DME ground thing - will double check that. I have been told that by a few people but I am not sure where to look for a bround from the DME to the chassis. Please let me know - everything seems to work, wiring harness works fine.

I tested the coil voltage and signal voltage - the coil voltage came out to be 12.22V. the signal voltage while every other coil was running was about .67V, while the signal to the other coils on the same multmeter showed up as .62 alternating with .64V... DME swap doesnt do any good, voltage stays constant in either case.

-Artem

[ArtemLepilov]

Does anybody have absolutel any idea what it can be? I disconnected the fuel injector, i run about 22MPG which is ok for now and so on and so forth... But its still annoying to run without a cylinder :(

-Artem

[winfred]

it's sorta hard to diagnose a oddball problem without getting our hands on it, i've done battle with this **** before and it can drive you nuts, have you tried another computer (or your computer in another car) since you fixed the bad injector plug? it's possible the injector driver for that cylinder got popped, seen that happen a few times as a matter of fact my non vanos m50 test computer has a bad injector driver on #1


Does anybody have absolutel any idea what it can be? I disconnected the fuel injector, i run about 22MPG which is ok for now and so on and so forth... But its still annoying to run without a cylinder :(

-Artem

[Rustam]

Just put it together, started it, injector connector was bad, replaced that - still no spark, coil tests fine, resistance in the coil is good, it fires on other connectors.

I will try to see the DME ground thing - will double check that. I have been told that by a few people but I am not sure where to look for a bround from the DME to the chassis. Please let me know - everything seems to work, wiring harness works fine.

I tested the coil voltage and signal voltage - the coil voltage came out to be 12.22V. the signal voltage while every other coil was running was about .67V, while the signal to the other coils on the same multmeter showed up as .62 alternating with .64V... DME swap doesnt do any good, voltage stays constant in either case.

-Artem

break the harness apart wire by wire and examine them. or get new harness...

[ArtemLepilov]

I think I just broke the connector today.. So the problem was there just now - since the "check engine" light came on only today - that was an indicator of a broken injector/connection, etc...

I tried both computers in the car, same problem throughout... Does anyone know a good shop in botson area I can take it to without asking too many questions?

-Artem

[ArtemLepilov]

gotta bump it... tested the wiring harness = it was good... still same problem. There is no degree to explain how lost I am...

-Artem

[E34 530]

gotta bump it... tested the wiring harness = it was good... still same problem. There is no degree to explain how lost I am...

-Artem

Sorry to say, but maybe it's time to bring it to a mechanic??

[ArtemLepilov]

Thinking the same actually.. just need some cash to get it done and pay the fines :(

-Artem

[E34 530]

Thinking the same actually.. just need some cash to get it done and pay the fines :(

-Artem
True, I'm in the same situation with painting the trunk lid lol.

[ArtemLepilov]

haha.. yea projects.. projects...

[Paul in NZ]

can i suggest something simple?Have you ever swapped the spark plugs around to see if maybe #4 is out,or maybe a fine carbon filament bridging the gap?

[ArtemLepilov]

Do you mean spark plug #4 itself is bad? - If thats the case I have tested all the coils with a same extra spark plug that I had lying around.

Also - the spark plug when I pull it from the cylinder head... It looks very clean, like brand new from the store kind of.

Thank you for all your suggestions :) they are valuable, and one of them might have solved the problem had it not been this damn complicated :(

-Artem

[Javier]

A.- You said you have no spark at 4.
You verified you have spark at all the rest under same testing method?

B.- You said you swapped DME.
After replacing coils?

C.- You said swapping DME won't fix it.
Swapped with a known good one?

D.- You said you have good battery voltage at pin 3 of coil connector.
Did you also checked you have good ground at pin 2?

E.- You said Cyl. 4 coil sparks OK at other positions.
Did you also checked other coils don't spark at Cyl. 4?

F.- Did you swapped Spark Plugs and problem stick to Cyl 4?

A = Problem should be ignition, not injection.
B = New DME output transistors didn't suffer from an old shorted coil.
C = DME is not the problem
D = Power source wiring is OK
E = Coil pack is OK
F = Spark plug is OK

As you may figure out, all the sentences can't be True all together, so have to find out which one isn't true.

If you install a toggle switch from coil pack connector pin 1 to ground and exercise it, you should have spark if coil is OK and get Battery voltage at pin 3 and Ground at pin 2. For that, in order to keep maximum scope on the test, I would remove DME, connect the toggle switch pin 1 to DME connector pin 50 (is the wire to cyl. 4 coil pack connector pin 1), connect Toggle switch pin 2 to good ground, connect the coil pack in its connector, plug the spark plug as usual, and ground it properly. Then go switching the toggle switch and check for spark. This operation has nothing to do with esoteric (CPS, Cam PS, DME, etc.) should discard coil pack, spark plug, and wiring issues. If spark is there, then D, E, and F hold true, and you should go and verify accuracy of B and C. Though I would first check for A, I mean, Is the spark really missing in 4?

For DME verification, you can assemble a simple tester to verify DME transistor is pulsating to ground. For that, get a High intensity red LED (Digi-Key MV8141-ND i.e.), wire the cathode (short lead) to a 560 ohm 1/2 watt resistor (wire wound preferably). The you have a two lead tester, (+) lead is the LED anode (long lead), and (-) lead is the free resistor lead.

Plug back in the DME, remove Cyl. 4 coil connector, remove fuel pump fuse, Stick the LED tester (+) lead to pin 3 of coil connector and tester (-) lead to pin 1 of coil connector, crank the engine and observe the LED (if alone, just turn ignition key to run and connect battery voltage to Diagnostic connector (http://www.unofficialbmw.com/e36/electrical/e36_reset_service_lights.html) pin 11 to crank engine), it should pulse light, one pulse for every two engine revolutions. Plug back fuel pump fuse, and start the engine. LED should light on (dimmed), and light intensity should increase slightly with engine RPM. Stick the tester to any other coil and verify light intensity is consistent with what you find for Cyl. 4.

If LED light OK, then DME is sending proper ground pulses to the coil pack, so B and C hold true. Leaving A as not possible.

You can use the same LED tester to verify injector driver operation, in this case (+) lead is to be connected to injector connector pin 1 (Battery Voltage) and (-) lead is to be connected to injector connector pin 2 (DME Driver). If LED pulses are there, you may still have a blocked injector. If Injector is OK, then you should start considering some cylinder damage. (Have you tested compression in it?

I've tried to put together some testing procedure to help you out of this hole. Would love to see comments and mods. from more experienced forum members.

Javier

[ArtemLepilov]

WOW - thats a hell of a procedure. I will definitely run it step by step next week when I have time away from school. Most likely wednesday. That procedure will definitely isolate the culprit - if that doesnt tell me what I need... Nothing will.

THANK YOU SO MUCH :)

-Artem

[Rustam]

For DME verification, you can assemble a simple tester to verify DME transistor is pulsating to ground.



why not simply use a multimeter between "positive" and the output looking for AC?

[ArtemLepilov]

why not simply use a multimeter between "positive" and the output looking for AC?


I did that - the result is in the previous post.. I dont remember it from the top of my head.

-Artem

[Rustam]

I did that - the result is in the previous post.. I dont remember it from the top of my head.

-Artem

actually i was asking Javier...

if you did this, how do you think the test with a LED is different from it? I really don't see a difference - both ways the presence of same voltage is checked...

[Javier]

of Vac readings in low duty cycle square waves. Also have to consider some load (or keep an under suspicious coil) in the open collector DME output circuit, as meter may be a very high ohm load for an accurate (noise free) reading. If you have a proper meter, you can also read frequency instead, and confirm it is consistent with the engine rpm.

Simple instruments helps a lot in handling and understanding what is going on, and ease neat conclusions.

From Artem posts, it seems that DME transistor is doing OK (as every other component in the system), though sentence is somehow confusing


I tested the coil voltage and signal voltage - the coil voltage came out to be 12.22V. the signal voltage while every other coil was running was about .67V, while the signal to the other coils on the same multmeter showed up as .62 alternating with .64V... DME swap doesnt do any good, voltage stays constant in either case.

he seem to report some Vac in the coil #4.

Javier

[joshua43214]

actually i was asking Javier...

if you did this, how do you think the test with a LED is different from it? I really don't see a difference - both ways the presence of same voltage is checked...

they are very different wave patterns, Measureing AC voltage would deliver an inacurate reading, since you would be measuring intensity rather than frequency of AC voltage on a DC circuit. A meter set on AC is looking for a syne wave sweep from positive to negative. The coil is a square wave from positive to zero voltage. Measuring DC voltage at the negative terminal would provide better results than measuring AC, but it would still be meaningless at this point since the car is past the fast and dirty method of diagnosis, and you would still be measuring intensity rather than frequency.

Javier's home made "noid" light is a simple, safe, and elgegant solution to bypassing the need for a lab scope, since it tests for frequency rather than intensity. It gives a clear and simple results, light on, light off. In one simple motion, the entire primary ignition circuit and function relating to the DME is verified. the only thing not tested in this is pin 2 of the coil. I have an old analog RPM meter I use for this sort of thing that I know is safe on these circuits, but hooking up some analog test equipment can damage many computor circuits.

Over 15 years ago, I got my first real experience in electrical bugs. The car was a Fiat X19 that had weird problems with the lights all coming on dimly at idle, stalling when you hit the break pedal, rough idle, and some other weird things. I spent several days tearing deeper and deeper into the car looking for problems. I wound up with 1/2 the interior pulled out and every attempt to check the wiring harness was good. All continuity tests where good, I raged and cried and called people asking for hints. Customers marveled at all the wires and decided I must be knowledgeable since I was so deep into the wires(little did they know).

A friend stoped by the shop to shake his head, as we where standing the talking over the trunk, I leaned on the bumper and suddenly the idle smoothed out, I steped back and it got rough again, carefuly leaning again, I noticed that the idle smoothed out and the running lights went off went I leaned on the tail light assembly. To make a long story short, the problem was a bad ground at the left tail light fixture that involved a shared ground with the ignition(don't ask, it a Fiat after all). I spent coutnless hours for no pay simply because I made 3 very big mistakes; 1, I assumed that something was good when it was bad. 2, I ignored every ones statements that most electrical problems almost always have very simple causes. 3, I ignored the customers statement that the problem started after someone rear ended her and cracked the tail light fixture.

To sum up, never make any assumptions, listen to what the experienced people are saying, most of them learned by some very hard lessons, and always look at the initial causes. I suggest you back up a bit and start over from fresh, cross every T and dot every I without making any assumptions at all about what is good and is bad. Isolate large areas as good or bad, then continue to isolate in smaller circles until you are left with nothing but the problem. As I said in an earlier post. my money is on a bad ground rather than a bad part, don't rule out bad terminals inside a connector, backprobe whenever possible, probing the connecter is a good way to damage the contacts and cause even more head ache.

[Javier]

when switching the coil manually to confirm system connections, remember the Toggle switch position to open the circuit, as it is the resting position ALWAYS. You should not keep the coil primary winding energized for long time, as high currents will stay long causing unnecessary heating.

The test should make quick connect-disconnect cycles, that is what the DME transistor would do. Coil operation is based on storing magnetic energy in the coil, and releasing it trough the plug.

If you don't get spark when switching at pin 50 of DME socket, please check for proper ground at pin 2 of coil pack. Guess it is your pending verification as suggested on D (Actually, I would do it first thing).

Javier

[ArtemLepilov]

I know I have proper ground on pin 2 - I ran a connectivity test, but as someone said - you never know when a wire under load will not work properly.

I will try to run those tests tomorrow, will report results.

-Artem

[Rustam]

Javier's home made "noid" light is a simple, safe, and elgegant solution to bypassing the need for a lab scope, since it tests for frequency rather than intensity.


Javier's proposal was concerned with observation of connection to ground that the transistor is supposed to provide. For that, there is no need for a lab scope, not for the LED tester. Since some voltage will be there if the transistor does work a multimeter suffices neatly (regardless of whether its set to "ac" or "dc" - which in cheap cases is technically same thing). To see whether the transistor switches to ground is enough to use a multimeter set to same mode on all the coils looking for differece at cyl#4 coil...