Mini MPI - K line fault

Great, many thanks for your quick reply!

I've used the wire on pin 33 to switch off and on the ignition and the result is the same as before, same messages.

I'm now going to check both k line cable and the one for the airbag ecu.

Thanks!

Sorry, forgot to ask, do you want me to check for those 5 V in the Pscan connector pins?

Cheers,

No, I mean that there should be 5v going to sensors such as temperature sensor

Hi! I've got the results from the last test and would like to revisit the previous results if you don't mind.

I've measured continuity between pin 7 of the OBD socket and pin 31 of the red ECU connector and it is fine. I've measured continuity between both of the above and some ground points in the car and there is no continuity, so apparently it is not connected to the ground.

As for the connector on the airbag module, I haven't even been able to remove that connector, do you know if I need to do something with those kind of tabs in the connector?

And the questions about the previous results in order to try to understand the whole picture:

Test 1. Measure Voltage and Intensity on the OBD socket when ignition is on

Did the measured voltage and intensity figures match the expected ones?

Voltage pin 7 to pin 4 = 12.25 V
Voltage pin 13 to pin 4 = 0 V

Intensity pin 7 to pin 4 = 0 Amps
Intensity pin 13 to pin 4 = 0 Amps

Voltage pin 16 to pin 4 = 12.25 V

Test 2. Check that the Pscan interface is not damaged by connecting it to another car

When the Pscan was connected to a VW Golf the message 'Possible K line short to 12 V - wait for 3 sec' was shown.
.
Test 3. Remove MEMS ECU from the car and connect it to the Pscan

When K line cable is connected
- K line fault message is not shown anymore
- Message shown: Wait for line for settle
- Message shown: Attempting fast init - When connected to the VW Golf it tried Slow init instead of Fast init, does this make any difference?
- Message Switch off ignition (by removing wire on pin 33)
- Message Switch on ignition (by connecting wire on pin 33)
- Message Init fail - Is this the expected behaviour?

When K line cable is removed
- K line fault message is not shown anymore - Is this the expected behaviour when the K line cable is not connected?
- Message shown: Wait for line for settle
- Message shown: Attempting fast init
- Message Switch off ignition (by removing wire on pin 33)
- Message Switch on ignition (by connecting wire on pin 33)
- Message Init fail - Why is the Pscan behaving the same way regardless of the K line cable being connected or not? Is this the expected behaviour?

Based on the above, can the following be stated?

- K line cable might be shorted to 12 V (Test 2)- Message 'Possible K line short to 12 V - wait for 3 sec' is shown
- K line cable connected or not connected (Test 3) - No messages about K line are shown
- K line cable might be shorted to earth - or somewhere else? - Message 'K line fault' is shown

And last question, is there any difference between Fast init and Slow init? can that make any difference?

Many thanks for your support!
David

The K line fault detection in EOBD/OBD2 mode (as of version 420) is not the same as the K line fault detection with rover proprietary protocols.
The reason that we have K line fault detection is this:-
The line is supposed to be held up to battery voltage by a resistance of about 510 ohms. An ECU or a diagnostic tool can then short the wire to 0V with an open collector transistor which the other end can detect.
Suppose that the K line is accidentally shorted to 0V. In this case current flows through the 510 ohms (approx 24mA) but it is not enough current to damage anything. The diagnostic protocol will not work, but nothing is damaged.
Suppose that the K line is accidentally shorted to 12V. In this case if either the diagnostic tool or the ECU attempt to communication and switch on their open collector transistor then a high current can flow which could damage something. This used to happen with Volkswagens a lot because they extended the K line to the plug that goes into the back of the radio so that diagnostic tools could talk to the radio. If you put in an aftermarket radio it would short the K line and damage diagnostic tools.

According to the OBD2/EOBD2 standards there is a solution to this. The engine ECU is supposed to have a weak resistance to 0V, whilst the diagnostic tool has 510 ohms or so to 12V. The tool can make this a switched 12V source. The tool initially sets the switch off and so the K line should be at 0V. The tool tests that this is the case. Then the tool switches on the 510 ohm pullup. Now the K line should be at 12V (or near it) and the receive side should be able to detect this. Now the tool knows that the K line is neither shorted to 0V nor to 12V and it has not yet risked using its open collector transistor on the K line.

Unfortunately not all cars (including Rovers and apparently Golf diesels) are compliant with the standards and actually have a resistor to 12V in the ECU. This means that there is no way for a diagnostic tool to test that the K line isn't shorted to +12V. For this reason pscan has to just give a warning in EOBD mode but not actually refuse to work.

This warning became annoying and useless on Rover ECUs and so what we do in Rover proprietary mode is that we transmit a short pulse onto the K line and test with our receive circuit can pick up the pulse. If it cannot then the next stage of proper communication does not take place, the idea being that this short pulse is not long enough to overheat and damage the chips but long enough to test if the K line is shorted.

For this reason the tests are different in different modes and the results may or may not mean the same thing.

So, when you connected pscan to your Mini and selected MEMS2J it would have put a short pulse onto the K line, but this short pulse was not seen by the receive circuit and so to protect itself the tool will not send any more data on the K line.

If you connect pscan to just power, and connect pin 7 to nothing then pscan will send a pulse to pin 7 and it will receive the pulse back and so there is no warning displayed about the K line.

So, if you have pscan in EOBD mode and it gives you a warning about the K line then it might be that there is something wrong, or it might be that the ECU is not strictly compliant with the ISO standard and this triggers the warning.

If you have pscan in MEMS2J mode and you get the K line warning then it means that something is wrong with the wire from pin 7 to the ECU, or with the ECU itself and pscan is not even receiving its own test pulse. You would then need to figure out what is wrong because there is no way that the diagnostic protocol is ever going to work.

If have pscan in RC5EC5 mode and you get the K line warning then it means that something is wrong with the wire from pin 13 to the ECU, or the SRS ECU.

Whether you use fast init or slow init makes no change to the K line testing.

David106 wrote: Message Init fail - Is this the expected behaviour?

This is consistent with three scenarios.
1. MEMS2J ECU not powered and so not communicating.
2. MEMS2J ECU powered but damaged, so not communicating.
3. pscan damaged. However this is not likely. pscan has sent a pulse onto the K line and detected the pulse (if it did not, then it would say K line fault). Also pscan working on the Golf. Therefore pscan is not damaged.

David106 wrote: - Message Init fail - Why is the Pscan behaving the same way regardless of the K line cable being connected or not? Is this the expected behaviour?

Because pscan is sending a pulse to pin 7 and receiving it back in both cases.

David106 wrote: Based on the above, can the following be stated?

- K line cable might be shorted to 12 V (Test 2)- Message 'Possible K line short to 12 V - wait for 3 sec' is shown

The important word here is "possible". If it says that there is a "possible" short, but then the tool can talk to the car then you can ignore it. If it says "possible" but then the tool cannot talk to the car then you have a reason that is worth investigating. This warning is not relevent to Rover ECUs.

David106 wrote: K line cable might be shorted to earth - or somewhere else? - Message 'K line fault' is shown

In MEMS2J mode it means that the tool transmitted a pulse and something is interfering with the pulse. This could be anything that is connected to pin 7 of the diagnostic tool.

There is one strange thing with your car.

You say that with the MEMS2J on the table that you do not get the K line fault, but on the car you do. There must be a difference. On the table the wire goes straight from OBD2 port pin 7 to Red socket pin 31. If the car has a wire straight from OBD2 port pin 7 straight to Red socket pin 31 and only that place then I would expect the same behaviour, but you do not get the same behaviour so something is different.

Many thanks for the comprehensive reply!

I've given it another try, setting up the same bench arrangement by the car but using the existing cable in the car to connect pin 31 on the Ecu to pin 7 on the Obd connector. The result is the same, Init fail message is shown.

For the next step, I'll try to connect the Pscan to a working Mini MPI and, if that were possible, to install my Ecu in order to see if it works or not.

Any advice on swapping an existing ECU by the one from my car? I guess that I'd need to at least reprogram the codes of the immobiliser, right? Anything else that I should bear in mind? Any precautions to be taken?

Many thanks again!

You'll just need the code for the immobilizer - the rest ought to be plug and play

Rob Bell wrote: You'll just need the code for the immobilizer - the rest ought to be plug and play

You have two options.

Fit the borrowed MEMS2J to your car and use the "learn EMS" feature to match it to your 5AS. When you return it to the car that you have borrowed it from, you will have to "learn EMS" again to set it back to 5AS from the donor vehicle.

Or there is a better option:-
Log into the 5AS on the donor vehicle and "read EMS".
Then go to your car and log into the 5AS and also do "read EMS" and write it down.
Then do "write EMS" and set the EMS to be the same as the donor vehicle.
Now you can fit the MEMS2J and it will be receiving the EMS code that it expects and so it will not be necessary to "learn EMS". By this method you can borrow a MEMS2J and not make any changes to it.

You only need to worry about this if you want to start the engine. The diagnostic protocol will work even if the EMS code is wrong.

Today I checked my MEMS2J.

This is how it is wired:-


It works. Ignore the green wire, I wasn't using it.

I found that when the ECU is powered that there is 5V on pin 18 of the black connector.

If your ECU does not have 5V on pin 18 then I would say that it is dead.

If it is dead then it is probably worth sending it to BBA Reman for repair.

Thanks both! I'll check for those 5V while I'm looking for the ECU to borrow and will get in contact with the guys that could repair it if needed.

Thanks for the tip and for the comprehensive explanations! 😊

Cheers

Hi again! Just to double check, those 5V are to be measured between the pin number 18 and the ground, right?

Cheers,

David106 wrote: Hi again! Just to double check, those 5V are to be measured between the pin number 18 and the ground, right?

Cheers,

Yes, exactly.