My Lightbulb Moment

And how you can have them too

Ben Johnson - AllMech

As is often the case, I encountered an interesting challenge last Friday. On the surface, it seemed like a simple CAN fault, but as we all know, things in the automotive world are rarely straightforward.

For years, I've grown accustomed to CAN systems with a simple mainline. It typically consists of a twisted pair with various branch lines connecting to control units throughout the vehicle. Now, for those who are new to the world of CAN (and there are still newcomers despite its maturity), let's dive into the specifics of PT-CAN, the star of today's case study.

PT-CAN, demands a total circuit resistance of 60 Ohms. It achieves this value by connecting two 120 Ohm resistors in parallel at the ends of the main lines. In most cases, you'll find these resistors in the DSC unit and the DME/DDE (ABS unit and Engine ECU for non-BMW applications). These resistors in a parallel configuration, according to Kirchhoff's law, are divided. That's 120 divided by 2, giving us 60 Ohms total CAN resistance.

Now, for those of you well-versed in CAN bus resistances, you'll understand my point when I say that, in my examination of the F31, I measured 120 Ohms at various points in the circuit, including the GWS and EKPS, among others. This measurement tells us in an instant that we have an open circuit within one of the main lines of the bus. But this is where things got a little tricky. Here is an example of a basic CAN BUS with a twisted pair main line.

In the example above, we could easily spot an open circuit if we had one, and usually involves tracing the exact point where the wiring harness has become open due to damage or corrosion. While this task can be challenging in itself, imagine if the CAN layout were designed to save costs and reduce wiring complexity. That's where the concept of a 'hybrid' main line comes into play.

But what do I mean by 'hybrid'? Well, think of it as a main line that combines physical wires from the instrument cluster to the body module, with terminating resistors placed within the instrument cluster and the body module.

BMW adopted this innovative approach with the Kombi and the FEM, and in a car I recently worked on, I encountered yet another open circuit. However, this time, it was a bit of a twist. Just like the DSC/DME setup, the Kombi and FEM have terminating resistors, right? Well, not exactly!

The hybrid setup takes the majority of the PT-CAN main line and transforms it into a solid state. This means it's packed with a myriad of electronic components that add complexity to our troubleshooting process. Check out the diagram below to get a clearer picture of what I'm talking about.

So How did I start this troubleshooting job off? Well I observed the fault codes and the fact that on the vehicle tree there were no comms to the GWS and the EKPS. EKPS is the fuel pump control unit that is responsible for providing duty cycled voltage to the single phase in tank fuel pump and the GWS is the gear selector in this automatic variant equipped with the solid ZF 8HP gearbox.

So I was facing two nodes with no comms and 120 ohms at each of them but why did other units not lose comms? Well the devil is in the detail. If you can imagine (not shown) that from the wiring diagram that the circuit is almost split in two - quite complicated to get your head around but the wiring before pins 48 and 47 split off to other control units and also before pins 18 and 19 the wiring splits off to other units. Now the thing is the DME was actually 120 Ohms but you will notice that since the DME, EGS and GWS have a PT-CAN 2 backup line even though the DME and EGS had 120 ohms open on the PT-CAN they did not go yellow. Interestingly the GWS did go yellow despite the PT-CAN 2 backup. There may be logic there but I am damned If I know why only the EKPS and GWS were yellow. The lack of PT-CAN 2 on the EKPS would explain about the lack of comms of course but the GWS troubled me.

So I knew I was facing an open and had realised that it was in between the Kombi and the FEM. I though the best thing to do would be to remove the connector A173 7B at the FEM and measure the resistance back from pins 36 and 35 to the Kombi terminating resistor. Result was a solid 120 Ohms - no break in that line. Then I removed the connector A173 2B at the FEM and measured the FEM terminating resistor and had again a solid 120 Ohms - no fault with either resistors. It was then just a case of tracing it all back and coming to the realisation that the break was solid state in nature and was in the FEM so that there was simply no connection to both resistors and we all know - if you see 120 you are seeing only one resistor so therefore you see only one line so there's a break somewhere.

You must keep thinking outside the box. General training rarely gives one this unique ability. Be an innovative problem solver and be proud of it. Learn by yourself and create solutions to complex problems.

Now, you might be wondering, how in the world can I bypass this complex FEM setup and achieve that elusive 60 Ohms across the entire PT-CAN? Well, let me tell you, I didn't just sit around pondering this question during my waking hours. Nope, I even dreamt up a solution, or should I say, half-dreamt it.

Picture this: It's 3 AM on a Saturday morning. I'm in a semi-state of sleep when suddenly, a brilliant idea pops into my head. I couldn't contain my excitement, so I sprang out of bed, grabbed a piece of paper, and began sketching out this brainwave. Of course, this unexpected burst of inspiration might not have been ideal for my dear Wife Johnson's beauty sleep, but when a breakthrough strikes, who can resist? Below is the ugly yet totally functional wiring harness that I used to bypass the FEM completely. Luckily I had a 120 Ohm resistor lying about.

It may appear confusing at first glance so I will talk you all through it. From left to right we have:-

1. DME Low
2. Kombi Low
3. FEM main line output Low
4. FEM main line output High
5. DME High
6. Kombi High

With the FEM out of the picture and my own handy resistor standing in for the FEM terminating resistor I was able to see 60 ohms at all parts of the PT-CAN thus ruling out anything else. In fact on this F31 the only bit of real main line wiring is from the kombi to the FEM and that is super easy to access and test. Believe it or not the rest of the main line is really just inside the FEM in solid state format. All this saves money by using less wiring but it gives us fault finders a headache at times when you spend years used to one type of CAN BUS layout then have to get your head around a different type of CAN layout.

So there we have it - A used FEM will be fitted and cloned. If this does not cure the fault then I will make that the topic of next months newsletter. But I feel confident so expect some other topics for next month. Until next month keep learning and keep posting the cool stuff you have all been posting.