[Ian M Davis]

Your car has three main computers:
(i) engine ECU
(ii) BSM (top part of the engine fuse box)
(ii) BSI (cabin fuse box)
These are connected by computer buses.

When you try to start the car the BSI sends a command to the BSM to operate the starter motor. The engine ECU is not involved in the actual cranking of the starter motor.

You mentioned two important things:
(i) Both keys are losing synch with the car
(ii) The car has an intermittent starting problem
This makes me suspect the BSI is losing power. The BSI gets its power from the the maxi fuse box which is located underneath the BSM. It could be a bad connection to the maxi fuse box or a bad earth. One simple thing you can check is the BSI earths (MC32 and MC31A) which are located on the sill in the left front footwell-see attachment.

 

[Ian M Davis]

Thanks for your reply and explanation. Keys lost sync only once, after that didn't occur anymore. I also suspect weak power as windows sometimes don't close fully whilst holding the button as if the power is weak. I'll try to check it out tomorrow and update it here.

Two things that are puzzling me:
(I) What do you mean "the electrician turned off the battery"? How did he turn off the battery - disconnect it?
(ii) You said you had to break into car - Is there a problem with using the key in the door lock to manually unlock the door?

 

[Ian M Davis]

(I) Disconnected the battery with closed windows and doors making us unable to enter, so we needed to break in. This happened moments after he successfully retrieved code and programmed the new key (second one). He didn’t do any BSI reset and I noticed that none of them follow battery disconnection procedure (if it’s important at all in this context)
(II) Yes, the problem is usual - last owner never used the manual lock so it’s stuck. It’s on my list of todos…

I’ll try BSI reset tomorrow morning and check the grounds, let you know how it goes. The whole car is a rubbish but it’s not something I would sell at this point, not until making everything work.

Okay, Thanks for the explanation.

The three minute rule is very important on all cars. Not observing the 3 minute rule is the same as unplugging a computer at the mains. In many cases it will do no damage but disconnecting battery while the BSI is awake can corrupt it. After turning everything off and shutting the doors it can take about 3 minutes for the BSI to shut down. One indication that the BSI is awake is the odometer display is on.

If you follow the shut down procedure there should be no need to do the BSI reset.

 

[Ian M Davis]

I have attached a wiring diagram for the starting system on your car. The part I am interested in block PSF1. This is the engine fuse box. It includes the BSM (main fuse box) and the maxi-fuse box. There are two things of interest:
Connector 5V JN
Fuse 8
5V JN means 5 pin yellow connector but on your car the plug is grey blue in colour (on later models it's yellow).

You need to check if power is getting to the starter motor solenoid. To do this you will need someone to turn the ignition key while you measure a voltage at the fuse box:
The yellow wire on connector 5V JN goes to the starter motor solenoid. Unfortunately the back of that plug is sealed so you cannot back probe the connector. What you need to do disconnect the plug and measure the voltage on the fuse box itself at pin 2 (the pin that connects to the yellow wire) while someone turns the key to the start position.

If there is no voltage at pin 2 of the connector you should also check the voltage at fuse 8. The fuse has two points on the top where there is bare metal you can put the meter probe on one of these points.

You may be right about the ignitions switch but I think it's more likely that a relay in the BSM is faulty.

The other thing to check when the car will not crank is get someone to turn the key and listen for click/clunk from the starter motor.

EDIT
If you want to check the ignition switch it connects to plug and socket 6V BA -see attachment for location. When the ignition key is in the start position pin 5 should be 12V. Note to identify pin 5 it is next to pin 6 which is not used.

 

[Ian M Davis]

I think you either have a faulty engine fuse box or just a bad earth. I have attached the location of the fuse box earths MC12 and MC30. Can you check both especially MC12. Make sure they are tight, clean and have no sign of corrosion.

 

[Ian M Davis]

I just want to check a couple of things:
(i) Can you confirm that when you measured 11.34V that it was at pin 2 of the fuse box? The voltage at fuse 8 is correct. This fuse connects to pin 2 via a relay. The relay only closes when the ignition is turned to the start position so there should be 0v at pin 2. Could there be some water in the fuse box?

(ii) When I listen to the sound in your video I can hear a fan but it sounds like the heater/air-conditioner blower running but you mentioned the engine ventilator are you saying the radiator fan is running as well or is the sound from the radiator fan so loud you can hear it inside the car.

EDIT
The 6V BA connector is probably inside the cover that surrounds the steering column and the ignition switch. To access it you will need to remove the cover.

Are you sure the 11.34V wasn't 11.34 mV? Meters will often read a few millivolts when connected to an open circuit.

 

[Ian M Davis]

The only doubt I'm having is that this started happening after alternator reparation, and previously it never happened...

Have you checked if the alternator is working by measuring the battery voltage with the engine running. There should be a black two pin plug on the alternator. I think it should only have one wire can you check if that is connected properly.

 

[Ian M Davis]

When you check MC12 can you also check MC11 and MC13. They are shown on the same diagram as MC12.

EDIT
The reason for checking more earth connections is the engine cooling fan is totally independent of the starting system. The fan is controlled by the engine ECU whereas the starter motor is controlled by the engine fuse box.

 

[Ian M Davis]

Thanks, Ian so much. I'll clean these three earth points and then give it a try as it's giving me 0L on a multimeter, while MC30 goes from 0.05 ohms toward 0.00. Still, I think I've done great work with your help considering I'm so illiterate when it comes to the cars and its electrical systems

multimeters are not very good at measuring low resistances. You will get more reliable results if you measure the voltage from the battery negative terminal to the earth wire. For example if you draw 10 amps through a 0.1 Ohms resistor you will get a 1V drop.

 

[Ian M Davis]

I don't think this problem is linked the thermostat or temperature sensor. It is quite common for electronics to develop heat sensitive problems. It can be as simple as a dry solder joint that makes and breaks contact depending on temperature.

The interesting thing about your video is when the ignition switch is turned to the starting position the reading on the meter suddenly changes but then goes back to its original reading. Digital multimeter have a very high input impedance so they don't load the circuit. While you should have been reading 0V at pin 2 of 5V JN if there was a small leakage the meter may have read 11V. I suspect there is a bad contact or dry joint probably inside the fuse box which breaks contact as soon as there is a load applied.

One question: The engine fuse box relay R5 controls a lot of things on the car including the headlights and the starting motor relay (R6). Relay R5 should close when the ignition is turned on. Do you know if headlights work when the car fails to start?

Note: Relay R6 also provides power to the BSI.

 

[Ian M Davis]

Headlights work when car fails to start.

Okay thanks for checking. It turns out fuse 9 which is for the headlights is only used on the discharge type of headlights. I think it is used to power the leveling system so they don't dazzle other divers. The actual lights don't have fuses.

 

[Ian M Davis]

My view is something is wrong with fuse box. You should not measuring 11V at pin 2 of 5V JN. You should have measured the full battery voltage when the ignition switch was turned to the start position.

The starter motor is at the back of the engine so I am not sure you will be able to access it from above.

 

[Ian M Davis]

if you can't access the starter motor to tap it. You can test the starter motor by providing power (12V) to the starter motor solenoid by connecting 12V to the yellow wire in connector 5V JN. To do that safely you should use a wire with a fuse holder and 15A fuse. Also make sure the car is not in gear.

 

[Ian M Davis]

I think you found there was full battery voltage at fuse 8. You could try taping a wire on the same point you measured the voltage at fuse 8. That way fuse 8 will provide the protection.

 

[Ian M Davis]

You need to make a connection to the yellow wire (not the fuse box). Some people take the wire out of the connector (but I think yours is sealed in) and just hold on the 12v rail in the fuse box for long enough to hear the starter motor turn. I prefer to have a fuse in case something goes wrong. All I was thinking is you may be able to touch a wire on the metal tip on top of the fuse.

 

[Ian M Davis]

I am sorry to hear that. There are not many things left to check. I think you need to do the test where you apply battery voltage to the starter motor.

 

[Ian M Davis]

None of those issues will prevent the car from cranking however. If engine ECU is faulty, it could cause other electrical faults. Hence hold off on DPF regen / refill.

If you repeat the test on engine ECU with hairdryer, it will confirm it further because replacing engine ECU is expensive (and a little complicated with immobiliser / stored data). The best test you can do. If when the ECU is too hot, the engine doesn't crank, no amount of cleaning anything else, elsewhere in the engine bay can fix anything

The only other thing you can do is test the wiring, under load, between the pin which controls the starter motor relay and the starter motor relay.

Like said previously, most likely it's the ECU though, as wiring looks fine.

I think your hairdryer test has confirmed the ECU is faulty. I suspect the last electrician knew that was the case but was hoping it wasn't.

There were some earlier PMs where is was confirmed there was full battery voltage at the solenoid connection when the car was cranking. The 11 V measurement was just some leakage which was only seen because there was no load on the pin.

I suspect changing the coolant temperature sensor wasn't necessary. At one stage we had it disconnected and I think this generated an historical fault code.

 

[Ian M Davis]

As a sidenote, be careful when testing only on voltage. As load increases, resistance increases, which drops voltage. Normally, this is accounted for by engineers, however when the wiring is damaged in any way, the voltage test will show full voltage as resistance is low (voltage test is very low load) but raises under load, to the point where you can get undervoltage situations.

There was something more interesting than that with the fuse box. One of the first things I learned in my elec. eng. degree is that digital multimeters have a very high input impedance so I would have prefered to have kept the load on the circuit but the back of the connector couldn't be accessed. The fuse box wiring diagram was only showing some relay contacts between the 12V rail and the connector pin. I was therefore surprise that there was always 11V at the connector pin. I had estimated there was about 130k of leakage across the relay contacts so I suspected some water may have leaked into the fuse box. However, the new fuse box is the same so clearly the design includes something across the relay contact possibly to stop them burning. I would normally expect a diode to be used in that situation but clearly not.

 

[Ian M Davis]

The camshaft belt breaking is quite serious. If it breaks it can damage the valves. Are you sure it wasn't just the accessories belt. The camshaft belt is the timing belt and it is a notched belt.

 

[Ian M Davis]

That's the accessory drive belt, some people call it the serpentine belt. In the old days when the fan was on the front of the engine it was the fan belt.

I suspect the electrician must have noticed either the camshaft sensor or crankshaft sensor must have got so hot its case melted. I think this what he replaced.

I was thinking about your problem last night and I think a previous owner must have had the ECU remapped. When the diesel particle filter additive tank ran empty you should have got a warning on the display. I suspect rather than fix the issue a previous owner had the DPF remapped out of the ECU.

All you tests indicate the ECU is inhibiting the starter motor from cranking. The engine ECU is not in the direct cranking circuit and the only way it can inhibit the cranking is by sending a command to the BSI over the CAN bus. If this is the case there should have been a fault code stored. I am beginning to suspect the remapping has corrupted some of the diagnostics.

I think I mentioned before that the ECU has two memories: EEPROM and Flash. The EEPROM contains data specific to the car such as the PIN. The Flash memory is the same on all identical models. It is the flash memory they remap. I suspect your EEPROM is okay so if you get a new identical ECU one way to match it to your car is to swap the EEPROM.