325i Auto: Cranks but Not Starting

[MunsterScot]

Hi E30ers,

I want to check something out with you and see what you think.

I've got a 1991 325i automatic vert.

The car would start and run fine but occasionally wouldn't start. Now it will crank over but will not start. I've worked through the 'Engines Troubleshoot' flow chart on the forum and done the following:

Does Engine Crank = Yes
Test bulb on coil, does it glow = Yes
Does the ICV buzz with IGN on = Yes
Check for power to fuel pump & spark = No to both

Fitted new CPS (Meat & Doria)

Power to fuel pump = Yes
Check for spark with lead removed from plug in head and spare plug on rocker cover = No Spark

Removed Distributor cap and checked rotor arm moving when cranked (in case timing belt broken) = Rotor moving (Note: New cap and rotor fitted last year)

Lead from king pin of coil to spark plug on top of rocker cover:
- When ignition key turned to position 2 = Single spark
- When trying to start car = no spark

From the 'Engines Troubleshoot' it's now pointing to the ECU is faulty.

Took ECU out and opened it. No obvious signs of corrosion/marks on the board. I've no way to actually check the electronics though.

Before I look into buying a replacement ECU is there anything else I should try/look at?

Any input/suggestions welcome.

Cheers.

[MunsterScot]

Update:
I've checked the continuity of the black wire from terminal 1 on the coil to pin 1 on the diagnostic connecter and it is good. Continuity to pin 1 of the plug that connects to the ECU is also good.

If I do the following I can make a spark plug spark:
1. Disconnect black wire from pin 1 on the coil
2. Connect lead to king pin on coil, spark plug on the other end and rest plug on rocker cover
3. Turn the key to position 2 to put 12V on the green wire/terminal 15 of the coil.
4. Use a piece of wire to quickly (touch and off) ground pin 1 of the coil

Conclusion:
The ECU is not 'switching' (timing) on and off the circuit of the black wire from terminal 1 of the coil. Therefore not creating a spark at the spark plug. So I need a replacement ECU.

Does this make sense?

I'm also thinking on changing from a 0 261 200 380 ECU to a 0 261 200 173 ECU as I've heard they are better. Any comments on this?

[Brianmoooore]

From what you have written, I tend to agree that you probably have a rare ECU failure. Probably the FET that drives the coil.
These can be damaged by a flashover in the coil, where the EHT from the king post on the coil flashes over, either internally or externally to the - terminal, and then, via the black wire, to the driver in the ECU. Flashovers can occur either because a coil is faulty, or sometimes if the coil is triggered when there's no spark plug connected to it to limit the peak voltage.
Fresh from encouraging HenryM3 to successfully repair his light delay module, it's not too difficult to open up the ECU and replace the FET if you can identify it and find a replacement.
The 173 ECU is no better or worse than a 380, except that it doesn't suffer from a fairly rare idle control fault that can afflict the 380.

[MunsterScot]

Thanks Brian.

[MunsterScot]

Update 2:

Replaced ECU with replacement working 173 and still wont start. Rechecked the following:

Does Engine Crank = Yes
Test bulb on coil, does it glow = Yes
Does the ICV buzz with IGN on = Yes
Check power to fuel pump = fuel at injector rail

Lead from king pin of coil to spark plug on top of rocker cover:
- When ignition key turned to position 2 = Single spark
- When trying to start car = no spark

Rechecked continuity of wire form pin 1 of coil to pin 1 of ECU plug = good.

Checked green wire above glovebox. Plug that connects the green wires to the green & green/red wires has been cut out and the two green wires connected together (links out immobiliser).

Still think it is electrical. Could it be one of the following:

1. New CPS is faulty? Does look slightly different to the original. Original all metal, new one plastid with a small metal sensor on the end (might just be new design)
2. Faulty Immobiliser. Has the key switch in the drivers door and a socket below the steering wheel where a fob would be put to deactivate ( have neither). But as said above the green wire above the glove box has been joined.
3. Faulty OBC: It a 13 button OBC but the screen is gone. I can see one small dash flashing on the left hand side of display.
4. Earthing issue some place?

Thanks.

[Brianmoooore]

Engine is earthed to the body OK, otherwise it either wouldn't crank, or would produce smoke from places that shouldn't produce smoke when you try.
Door immobiliser and OBC immobiliser both connect were the green wire has been linked, so you should be able to discount them.
If you're sure there's no spark, then the CPS is suspect, but this would affect the fuel pump as well. Pull the plug off of the fuel pump and connect your test lamp to it. Should light when you crank the engine, and stay lit while it is turning.
Disconnect a wire from the coil primary and connect your test lamp in SERIES. Turn the ignition on. Does the lamp light?

[MunsterScot]

Hi Brian, Checked what you suggested:

Fuel pump connection: Connected a bulb to it. Ignition in position 2 = bulb off. Try to start car and bulb lights and stays lit while the engine is cranking.

Green wire to coil disconnected and bulb connected in series = no light when ignition in position 2 or trying to start engine.

[BristolE30]

If you're sure there's no spark, then the CPS is suspect, but this would affect the fuel pump as well. Pull the plug off of the fuel pump and connect your test lamp to it. Should light when you crank the engine, and stay lit while it is turning.

In my experience this wasn’t 100% true. Had a faulty CPS sensor and we were getting plenty of fuel but it just wouldn’t start. It was shorting out on the unsheathed earth inside the cable

For the OP, your situation sounds very similar. You’ve tested everything else and getting a very weak single/intermittent spark.

Maybe try replacing the CPS again if you haven’t already? Check the wire for damage from your pulleys? A brand new cable could be damaged in a few seconds if it rubs on the pulleys. This was our situation

[Brianmoooore]

Hi Brian, Checked what you suggested:

Fuel pump connection: Connected a bulb to it. Ignition in position 2 = bulb off. Try to start car and bulb lights and stays lit while the engine is cranking.

Green wire to coil disconnected and bulb connected in series = no light when ignition in position 2 or trying to start engine.

Unfortunately, that all seems as it should be. I say unfortunately, because that means we are still no nearer to finding the cause of the problem, now that everything has been checked.
This means it's time to start revisiting what's already been done, which brings us back (as the most recent post suggests) to the CPS. A duff CPS call still produce enough 'noise', under certain failure conditions, to provide enough signal to be integrated and activate the fuel pump, without producing the pulses to be amplified into sparks.
Although it's not at all unusual for a new component to be a faulty component, it would be unusual for a new CPS to be faulty in this way, but we've now moved into the area of low probabilities. Damage to the lead would be a possibility, but it would be a pretty poor mechanic that installs a new CPS with the lead passing millimeters from a rotating pulley and does not take steps to secure it away from the rim.

[MunsterScot]

Thanks Brian and Bristol.

I've also checked to ensure that the black wire from the coil to the ECU is not shorting to earth. Thinking being that a short to earth would allow the 1st spark when the key is turned to position 2 but then not spark after that. It checked out OK.

I'll take the CPS out and have a check of the cable just in case. I'll also see if I can pick up a 2nd one and try that. I'll update once I've done this and let you know how it goes.

Thanks for your input so far.

[MunsterScot]

Update:
I've taken the CPS out and the cable looks ok, no damage. When I measure the resistance across the 3 pins in the plug I get the following (pin 3 being the side the cable comes out the plug):
Between pins 1 & 2 = 500 Ohms
Between pins 1 & 3 = Open circuit
Between pins 2 & 3 = Open circuit

[MunsterScot]

Correction: CPS reading 507 Ohms.

When something metal put across the sensor the resistance changes.

Connected CPS back to socket on car and measured resistance at terminals 47 & 48 at ECU plug. Getting 507 ohms there as well (so continuity of the wire is good). Resistance varies when something metal put beside sensor.

With the resistance only being 507 Ohms (I've read it should be 540) could it be the CPS signal is strong enough for the fuel pump to run but not good enough for the timing/spark circuit to work (i.e. the square wave signal is not large enough to trigger the timing/spark circuit) ?

[Brianmoooore]

The CPS is simply a coil of fine wire wrapped around a bar magnetic. The OE sensor has a resistance of 540 ohms but there's no reason why a pattern part should be the same.
One of the failure modes is for some of the turns to short out, which has the effect of damping out the pulses, as well as reducing the overall resistance.
The other failure mode is for the wire to go open circuit which is easily measured

[steve617]

I would get the original DME out and send it off to ACtronics or whoever else u may use / know of. If the DME has a fault, u can go on testing stuff for weeks to no avail, I speak from experience. My 1990 325i convert would crank,, but not start - DME was faulty.

[Brianmoooore]

I would get the original DME out and send it off to ACtronics or whoever else u may use / know of. If the DME has a fault, u can go on testing stuff for weeks to no avail, I speak from experience. My 1990 325i convert would crank,, but not start - DME was faulty.

Look further back up the thread. ECU's already been swapped with a known working one.

[Blanca]

Got that the other morning, gave the pump in front of the rear wheel a thump and off it goes. I need to get down and under.

[MunsterScot]

Well. I've verified the resistance of the new CPS changes when it is placed near metal and it does. I re-fitted the new CPS to the car and measured the resistance of it while turning the car over to make sure it is detecting the toothed wheel (in case the new CPS wasn't lining up correctly). While turning over the car the resistance of the CPS was varying between 1.1 KOhm and 1.3 KOhm. Measured the resistance of the CPS when not turning car over and it measures 399 Ohms.

I realise my multi-meter doesn't react quick enough to pick up the real values of the CPS when the engine is turning over on the starter but does readings between 1.1 KOhm and 1.3Kohm sound about right? With the CPS only reading 399 Ohms on the bench Vs 540 Ohm expected (I realise Brian explained before a new CPS doesn't necessarily measure the same) what I'm wondering about the new CPS is - is the new CPS giving a good enough signal(resistance) for the fuel pump to run but the resistance isn't high enough / varying enough to trigger the timing electronics for spark (the ECU isn't seeing a big enough change)

Other than that I'm at a loss to the cause. Could it be related to the ignition switch on the steering column? It's about the only thing I haven't looked at yet but I don't see how it could be.

[MunsterScot]

I've done a bit more investigation. The new CPS is from Meat & Doria. I've looked up the OEM part number on RealOEM and it comes back with 12141720856. When I look on the Meat & Doria web site for OEM part number 12141720856 it shows a different looking sensor (all metal like the original Vs plastic with a metal tip on the end that I have) and the the sensor measures 550 Ohms Vs the one I have measuring 400 Ohms.

So, I think the motorfactors have sent me the wrong CPS even though the web sire said it was suitable. Next step - another new CPS with OEM part number 12141720856 so I know it is 100% correct.

Will let you know how it goes.

[BristolE30]

I've done a bit more investigation. The new CPS is from Meat & Doria. I've looked up the OEM part number on RealOEM and it comes back with 12141720856. When I look on the Meat & Doria web site for OEM part number 12141720856 it shows a different looking sensor (all metal like the original Vs plastic with a metal tip on the end that I have) and the the sensor measures 550 Ohms Vs the one I have measuring 400 Ohms.

So, I think the motorfactors have sent me the wrong CPS even though the web sire said it was suitable. Next step - another new CPS with OEM part number 12141720856 so I know it is 100% correct.

Will let you know how it goes.

Hopefully that sorts it!

[MunsterScot]

Latest update:

New CPS arrived from motor factor. Double checked OEM number from RealOEM and on Meat & Doria website to make sure I had the correct part = yes.

New CPS fitted and still doing the exact same thing. Turns over, won't start, no spark with original CPU fitted. Will try again in the morning with spare ECU.

Failing that it's back to the drawing board and start again with the fault finding. Arrrggghhhh.

[Brianmoooore]

Well. I've verified the resistance of the new CPS changes when it is placed near metal and it does.

It's not changing resistance as such. It's generating electricity which confuses your meter.

Could it be related to the ignition switch on the steering column? It's about the only thing I haven't looked at yet but I don't see how it could be.

Your test with a lamp between the + terminal of the coil and earth, with the ignition on, should also have included turning the switch to the start position.
It's not unknown for a faulty ignition switch to disconnect ignition power when turned to the start position.
Are the tips of the spark plugs wet with fuel after a cranking session?

[MunsterScot]

Thanks Brian.

Checked what you suggested. Lamp connected to + terminal of coil and to earth. Key in position 2 = lamp on. Try to start car = lamp still on.

[Brianmoooore]

Sorry to burst your euphoric bubble, but an open circuit blue temp. sensor is very unlikely to prevent your engine from starting, especially at this time of year. It would start as normal, but then start to misfire as it warms up, before stalling, and refusing to restart, until it had cooled right down again.
HOWEVER, you may have hit upon an important clue, although nothing that would affect sparks at the plugs. You may have also noticed I asked about wet spark plugs in my last post. This is because I was already suspecting what you might hopefully now find.
The signal from the blue temp. sensor passes through a plug and socket attached to the black metalwork under the inlet manifold, and this is by far the most likely place for the temp. sensor wiring to be open circuit.
The plug and socket has a known problem with water ingress which gets trapped inside the rubber boot of the lower half of the connector, which results in electrolytic corrosion of the copper of the wires. As well as both temp. sensor signals, the power and both signal feeds to the fuel injectors also pass through it, and failure of these wires will result in the engine not starting, leaving the spark plugs bone dry after trying.

[MunsterScot]

Thanks again Brian.

I've checked out the blue sensor to ECU and it is actually OK.

On the connector under the inlet manifold. Approx. 18 months ago I cut it out and soldered the wires together together. It's ran fine for approx. the 18 months. Couple of times it wouldn't start or cut out. The last time I took it out it started fine. Approx. 15 - 20 minutes into the run it gave a couple of tiny hic-ups where it felt like it might cut out (but didn't).

About 5 min later I stopped and turned car off. Approx. 15 min later went to start and it wouldn't. Key out, Quick look under bonnet, Made sure connections to coil tight, tried to start again and she fired up. 25 min run home all ok. Went to put it in the garage a few hours later and wouldn't start and hasn't since.

I'll keep going with the fault finding and report back.

Only other thing to mention was that on the last run it was raining and roads were wet but I don't see how that would make a difference.

[MunsterScot]

Well after being in the garage from 07:30 to 12:00 I've (re)checked the following:

1. New CPS fitted = No start

2. Resistance at ECU plug for CPS reads the same as the CPS

3. Tried 2nd ECU = No start

4. Test bulb on Coil:
- Key in position 2 = Bulb lights
- Try to start car = Bulb stays on

5. Key in position 2. Bridges connections 11 & 14 in C101 = engine turns over but does not start

6. Bridges DME Relay = No start (but didn't check for wet plugs at this point),

7. Disconnected fuel pump and attached bulb to connector = Bulb lights when trying to start car.

8. Fuel line disconnected from fuel rail and placed in bottle. Tried to start = fuel in bottle.

9. Blue temperature sensor reading 4.79KOhm. I had thought the sensor was reading open circuit on ECU plug pins 24 & 45 but it's not. It is reading 4.79KOhm.

10. King lead from coil with spark plug resting on rocker cover. Turn key to position 2 = single spark some times. Try to start car = no sparks.

11. Key in position 2. Wire to Pin 1 of coil disconnected. King lead from coil with spark plug resting on rocker cover. Tap earth to pin 1 of coil = spark on plug.

12. Took the front 4 plugs out. Thought they looked wet but not 100% sure. Dried them with a rag and put them back in. Tried starting 5 times for 5 seconds. Took plugs out again. Centre electrode ground electrodes looked dry.

So, Looks like I have no spark or injectors firing.

13.: Connector under inlet manifold:
- I'd cut it out approx. 18 months ago and soldered wires in it's place when I had a similar problem (problem starting and erratic running). Fix last time was I replaced the blue temperature sensor. Checked the wires where the plug was and the soldering and wires look fine.

[Blanca]

Have you tried a second coils, just to eliminate that?
Disconnect negative side of coil and with ignition on and using a bit of wire, flask negative to earth a few times quickly, should spark, I think.(depends on model)
Check with voltmeter across an injector connector to see if you are getting a signal? Should hear them clicking anyway.
You say you disconnected fuel line from fuel rail, I assume that was at the rear, what about a faulty regulator at the front so there is no pressure, although from what I read the problem might be spark.

[HenryM3]

Just as a bit of a story, my son many years ago would call, needed towed home mk3 golf wouldn’t start. Got home car started, happened loads of times. Checked everything, all reading correct. Some old bloke (in the Brian mould) said change the coil. Explained it checked out correct. He said just change the f*%king coil. Problem never happened again. Now I’m not saying that’s your problem but just because something checked out while static doesn’t mean it will when under load/pressure. Just saying.

[Blanca]

Just as a bit of a story, my son many years ago would call, needed towed home mk3 golf wouldn’t start. Got home car started, happened loads of times. Checked everything, all reading correct. Some old bloke (in the Brian mould) said change the coil. Explained it checked out correct. He said just change the f*%king coil. Problem never happened again. Now I’m not saying that’s your problem but just because something checked out while static doesn’t mean it will when under load/pressure. Just saying.

Changed so many you would not believe it, especially on the off road buggys, They still carry a spare on tour.

[MunsterScot]

Blanca & HenryM3:

I've tried a 2nd coil and made no difference.

Blanca: On "Disconnect negative side of coil and with ignition on and using a bit of wire, flask negative to earth a few times quickly, should spark, I think.(depends on model)" this is what I did at point 11 that I wrote. Doing this I able to get spark so I know the coil and lead from the king pin of the coil are good.

When I checked for fuel I removed the fuel hose from the fuel rail under the bonnet (beside where the Fuel Pressure Regulator is).

Checking the wiring to the injectors is on the to-do list for tomorrow.

[Tineca]

Sorry I don't have the depth of knowledge to specifically describe which one it might be but the intermittent fault you had before suggests a relay might have been sticking and has now given up.

[Brianmoooore]

Doesn't fit the symptoms as described originally, but checking relevant relays was going to be my next suggestion - I see I've just been beaten to it!
First of all, are the correct relays in the correct place?
There are three relay sockets under the cover by the AFM. One is normally empty in the UK, and the other two should have a white five pin relay and an orange four pin relay. Check that the five pin relay is plugged into the socket which has a wire to the centre pin, and that the four pin relay is plugged into the socket that has a green/purple wire going to one pin. The individual bases are easily removed to check the wires underneath by bending the plastic tab near where they mount, after the relay has been removed.
If the correct relays are in the correct sockets, the best way to test them is by replacing them with wire links. Connect pin 30 to 87 on he orange fuel pump relay, which should make the pump run continuously. You should easily hear the pump whining at the back, and creaking noises from the fuel rail/FPR.
Connect pin 30 of the base of the five pin DME relay to both pins marked 87, with two wire links.
The pin numbers of the base are marked on the side of the relays.
Try to start the car with these links in place.
Remove the links after testing, or the battery will be drained.

[MunsterScot]

I didn't think it would be relay related as there is power to the fuel pump (fuel pump relay ok) when trying to start the car (tested with test lamp) and the ICV buzzes (power from main relay) when key turned to position 2. Not to worry, I've checked the following:

There are 3 relays under the black plastic cover beside the AFM. One white (on the left sits lower down) and two orange. Removed all three and tested on the bench using 12V supply to energise the coil and using a multi-meter to verify contacts operate. All ok. Linked out the fuel pump relay in the socket and verified power to fuel pump = OK. Linked out the three sockets for pins 30, 87 & 87 of the white relay. ICV buzzes. I also heard a click from what sounded like just behind the headlight beside the air intake it. I think this will be ABS related as the main relay also supplies power to the ABS. Tried to start car with links in place for white relay and no joy.

While checking this I also found the following and don't know if it is normal or not or could be related to the non-starting issue.

With the ECU disconnected and the key in the off position the green wire connected to the coil is shorted to ground. I broke the link in the green wire above the glovebox (where the immobiliser would be connected) and the green wire shorted to earth is on the ignition switch side rather than the coil side.

Question is: Is the green wire being earthed when the key is in the off position normal (e.g. to ensure all power is removed from the ignition circuit) or could the ignition switch be failing in some way or could there be a earthing issue causing the car not to start?

I had a look at the electrical wiring diagrams and could not see where the ignition switch would earth the green wire in the off position.

[Brianmoooore]

The green wire most certainly shouldn't be earthed, and I very much doubt that it is. The ignition switch feeds more than the engine ignition circuits via position two, and what you are measuring is the few ohms to earth of these other circuits. There's a joint in the wiring loom a few inches from the plug on the steering column where the green wires split in various directions.
You say you have three relays, so I presume you have the rare option of a cat. and lambda sensor. The third relay controls the heater in the lambda sensor.
I now you've said that the fuel pump runs when you crank the engine, but the sum of what you've done describes an engine that should run. Only problem is, it doesn't!
We've been through everything and more.
Has a different coil been tried?

[BenHar]

You say you have three relays, so I presume you have the rare option of a cat. and lambda sensor. The third relay controls the heater in the lambda sensor.
I now you've said that the fuel pump runs when you crank the engine, but the sum of what you've done describes an engine that should run. Only problem is, it doesn't!

Another reason for a third relay might be the cold start bodge on the 1988 cars.

<edit> OP says 1991, so not that.

Ben

[Brianmoooore]

Another reason for a third relay might be the cold start bodge on the 1988 cars.

That relay was mounted on the side of the fusebox, and, IIRC only applies to the 2.0 litre.