[willisodhiambo]

I am trying to figure out what could be happening with a faulty bus on an E90 before opening up things. This is my first foray into a bus system.

Looking at the wiring diagram of the K-bus on this car (only available options) we have

1. Junction box directly wired to IHKA, CID, KOMBI and MRS
2. Branch 1 has the following all in parallel configuration, Tilt alarm, Seat module driver, CIC, Radio
3. Branch 2 has CAS, FRM, Seat module Front passenger
4. Branch 3 has comfort access module

The fault is logged on the following branch wise
a) Junction box - Errors in JBE itself, IHKA, CID (MRS is OK!!)
b) Branch 1 - Errors occur in Radio but CIC and seat module driver seems ok
c) Branch 2 - Error logged in CAS, FRM, SMFA basically the whole branch
d) Branch 3 has comfort access, error logged

The first error by mileage occurred simultaneously on the IHKA and FRM. IHKA is wired direct to JBE, FRM to an external branch. So my questions are,

1. Wouldn't a short in the bus take out the whole bus or at least ECUs after the point that has shorted? In this case we seem to have an illogical order of faults, not what I would expect of an ECU had an internal fault or if the line had a break or short to ground.

2. Based on the above observation and purely from wiring diagram analysis, would it be safe to assume the JBE is most likely faulty? Thus the illogical faulty vs good communication of ECUs on the same circuitry?

I have looked under the carpets, dry. In any case none of these ECUs have wires away from the under dashboard area except the seat modules.

I have an oscilloscope, how would one go about looking for the problem here if it is not obvious its the JBE?

[xtac1080]

I am trying to figure out what could be happening with a faulty bus on an E90 before opening up things. This is my first foray into a bus system.

Looking at the wiring diagram of the K-bus on this car (only available options) we have

1. Junction box directly wired to IHKA, CID, KOMBI and MRS
2. Branch 1 has the following all in parallel configuration, Tilt alarm, Seat module driver, CIC, Radio
3. Branch 2 has CAS, FRM, Seat module Front passenger
4. Branch 3 has comfort access module

The fault is logged on the following branch wise
a) Junction box - Errors in JBE itself, IHKA, CID (MRS is OK!!)
b) Branch 1 - Errors occur in Radio but CIC and seat module driver seems ok
c) Branch 2 - Error logged in CAS, FRM, SMFA basically the whole branch
d) Branch 3 has comfort access, error logged

The first error by mileage occurred simultaneously on the IHKA and FRM. IHKA is wired direct to JBE, FRM to an external branch. So my questions are,

1. Wouldn't a short in the bus take out the whole bus or at least ECUs after the point that has shorted? In this case we seem to have an illogical order of faults, not what I would expect of an ECU had an internal fault or if the line had a break or short to ground.

2. Based on the above observation and purely from wiring diagram analysis, would it be safe to assume the JBE is most likely faulty? Thus the illogical faulty vs good communication of ECUs on the same circuitry?

I have looked under the carpets, dry. In any case none of these ECUs have wires away from the under dashboard area except the seat modules.

I have an oscilloscope, how would one go about looking for the problem here if it is not obvious its the JBE?

What problem are you trying to solve? What are the "errors logged"? A short to ground or 12v in either twisted pair wire would take out every module in that PDF.

[BaggieMatt]

First thing I would be checking is the PDC.

There is a common fault on some cars that the bay in which it sits gets wet. It will be I think behind the rear wheel well on the right hand side.

The unit shorts out and causes all sorts of bus issues.

If this is the case, take it out and open it up to dry it out. Once dry take a toothbrush and isopropyl alcohol and clean up the board. This isn't guaranteed to work but will give it a better chance if the contamination is removed.
If still dead then replace.
To cure the flooding there is a plastic plug at the bottom of the well. Just drill a small hole there.

[willisodhiambo]

I am trying to figure out what could be happening with a faulty bus on an E90 before opening up things. This is my first foray into a bus system.

Looking at the wiring diagram of the K-bus on this car (only available options) we have

1. Junction box directly wired to IHKA, CID, KOMBI and MRS
2. Branch 1 has the following all in parallel configuration, Tilt alarm, Seat module driver, CIC, Radio
3. Branch 2 has CAS, FRM, Seat module Front passenger
4. Branch 3 has comfort access module

The fault is logged on the following branch wise
a) Junction box - Errors in JBE itself, IHKA, CID (MRS is OK!!)
b) Branch 1 - Errors occur in Radio but CIC and seat module driver seems ok
c) Branch 2 - Error logged in CAS, FRM, SMFA basically the whole branch
d) Branch 3 has comfort access, error logged

The first error by mileage occurred simultaneously on the IHKA and FRM. IHKA is wired direct to JBE, FRM to an external branch. So my questions are,

1. Wouldn't a short in the bus take out the whole bus or at least ECUs after the point that has shorted? In this case we seem to have an illogical order of faults, not what I would expect of an ECU had an internal fault or if the line had a break or short to ground.

2. Based on the above observation and purely from wiring diagram analysis, would it be safe to assume the JBE is most likely faulty? Thus the illogical faulty vs good communication of ECUs on the same circuitry?

I have looked under the carpets, dry. In any case none of these ECUs have wires away from the under dashboard area except the seat modules.

I have an oscilloscope, how would one go about looking for the problem here if it is not obvious its the JBE?

What problem are you trying to solve? What are the "errors logged"? A short to ground or 12v in either twisted pair wire would take out every module in that PDF.

I am trying to resolve the bus error you can see in ISTA. Everything works ok but I want the ECU tree to go green.

---

First thing I would be checking is the PDC.

There is a common fault on some cars that the bay in which it sits gets wet. It will be I think behind the rear wheel well on the right hand side.

The unit shorts out and causes all sorts of bus issues.

If this is the case, take it out and open it up to dry it out. Once dry take a toothbrush and isopropyl alcohol and clean up the board. This isn't guaranteed to work but will give it a better chance if the contamination is removed.
If still dead then replace.
To cure the flooding there is a plastic plug at the bottom of the well. Just drill a small hole there.

This car has no PDC. From my list above I have listed all the relevant ECUs. All ECUs with errors are in the dash area except passenger/driver seat module.

Shouldnt there be logic in the ECU or junction to go for first? I believe in a circuit of this type surely a fault must manifest in a logical electrical way. This one does not going by the fact that some ECUs are OK, others not within the same bus but I guess things don't always work that way

[Node]

I would really suggest you first give us a clue what the error is exactly. And not from one but from all units. Because if errors are different even slightly, this can give us a clue on whats going on.

[Enabled]

Best way to troubleshoot CANBUS is to get a multimeter out and measure resistance between high and low lines, as well as Voltage.

[xtac1080]

You should post the diagnostic codes that each module with faults is outputting. I strongly suggest you read up on how a CAN network functions. It's not complicated and will help you diagnose these issues quicker in the future. CAN Bus Wiki

[willisodhiambo]

I attached the faults as per ISTA. Everything seems to work physically but I log "K-Line fault" with ISTA.

---

Best way to troubleshoot CANBUS is to get a multimeter out and measure resistance between high and low lines, as well as Voltage.

So the circuit tree and fault code from ISTA can not play a part is giving a clue where in the tree to start the search? I hoped there would be a logical place to start by examining the tree and list of ECUs with fault.

I have the fault list from ista, I have the bus circuit diagram, the faults do not make electrical sense given the order and randomness of the ECUs showing "K-Line fault"

Opening up ALL modules and testing every wire can be very time consuming for a novice and none technician like me but if it is the only way then I have little option

[Node]

Well, first you need to understand that even CAN bus uses two wires, it can remain operational (although with higher probability of errors) even if one wire is faulty. If the faulty wire is L one, then its almost totally operational. So based on this and your error list a couple of ideas for you to start:
1. Measure voltages on both H and L wires. Multimeter is fine, if you have oscilloscope its even better, because then you can see if there is any signal on the wire.
2. Based on error list I will guess that L wire got a fault. Why? Because JBE and MRS do not report the fault. Looking at schematics for E90 you can see that all control units which report a fault are connected to two different ports of JBE. Then you will notice that actually both of those ports share the same pin for L wire! Other units which do not report a fault are connected to other pins.

Now I have never opened a JBE before, so I do not know if those wires are connected within JBE directly or via some buffering electronics so I would try to see whats wrong with pin 10 on JBE connector. I would try to disconnect it from JBE and see if signal is OK on JBE side. If it is then JBE is ok and the short is somewhere in a bus. Else JBE is faulty.

If taking pin out of JBE connector is too hard then you can try disconnecting everything from X15002, X15004 and X15006 (if those are easily disconnectable, don't know how its done on E90) as those are distribution points of that signal.

If not two choises:
1. If you have quality oscilloscope (not those handheld POCs up to 2MHz) then you should reconnect pin 10 to JBE and try looking at the signal on various modules. Given your oscilloscope is sensitive enough you may notice that near one unit signal is totally (as an example) shorted to ground, while you can still see small signal near other units.
2. Disconnect units one by one and see if signal gets back again. You will know which unit to blame if the fault is not in a wire itself.

P.S. please note that some aftermarket alarm systems also connect to that bus. If such is present in that car it may be pain in the ... to find where it was spliced into cable, but my experience shows that such POCs are 99% of time responsible for malfunctions.

[willisodhiambo]

Well, first you need to understand that even CAN bus uses two wires, it can remain operational (although with higher probability of errors) even if one wire is faulty. If the faulty wire is L one, then its almost totally operational. So based on this and your error list a couple of ideas for you to start:
1. Measure voltages on both H and L wires. Multimeter is fine, if you have oscilloscope its even better, because then you can see if there is any signal on the wire.
2. Based on error list I will guess that L wire got a fault. Why? Because JBE and MRS do not report the fault. Looking at schematics for E90 you can see that all control units which report a fault are connected to two different ports of JBE. Then you will notice that actually both of those ports share the same pin for L wire! Other units which do not report a fault are connected to other pins.

Now I have never opened a JBE before, so I do not know if those wires are connected within JBE directly or via some buffering electronics so I would try to see whats wrong with pin 10 on JBE connector. I would try to disconnect it from JBE and see if signal is OK on JBE side. If it is then JBE is ok and the short is somewhere in a bus. Else JBE is faulty.

If taking pin out of JBE connector is too hard then you can try disconnecting everything from X15002, X15004 and X15006 (if those are easily disconnectable, don't know how its done on E90) as those are distribution points of that signal.

If not two choises:
1. If you have quality oscilloscope (not those handheld POCs up to 2MHz) then you should reconnect pin 10 to JBE and try looking at the signal on various modules. Given your oscilloscope is sensitive enough you may notice that near one unit signal is totally (as an example) shorted to ground, while you can still see small signal near other units.
2. Disconnect units one by one and see if signal gets back again. You will know which unit to blame if the fault is not in a wire itself.

P.S. please note that some aftermarket alarm systems also connect to that bus. If such is present in that car it may be pain in the ... to find where it was spliced into cable, but my experience shows that such POCs are 99% of time responsible for malfunctions.

Now this is very enlightening!! Really appreciated. I will give it a go. The splicings at x15002, 4 e.t.c may not be easily accessible. I wonder is there is a unit that the BMW plugs can go onto then out into the ECU so that opne can test from such a break out box!

Based on your very detailed explanation, I will first take out module by module and see effect, then remove all modules and test terminal at JBE with oscilloscope. I believe if the reading is not consistent I can then chuck out this JBE. Will report back once I get opportunity to carry through this plan.