When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Have you checked the resistance of the two CAN wires to ground to make sure they aren't shorted to ground? I don't necessarily think that's the case here, but it just crossed my mind.
I suspect you already know this, but my understanding is that improper termination of either of the two ends of the CAN bus will affect the waveform since it's a transmission line, which needs to be impedance-matched. To me, all of this seems to be pointing to improper termination, but I don't really have enough experience with CAN to say anything definitive.
Have you checked the resistance of the two CAN wires to ground to make sure they aren't shorted to ground? I don't necessarily think that's the case here, but it just crossed my mind.
I suspect you already know this, but my understanding is that improper termination of either of the two ends of the CAN bus will affect the waveform since it's a transmission line, which needs to be impedance-matched. To me, all of this seems to be pointing to improper termination, but I don't really have enough experience with CAN to say anything definitive.
We tested both CAN lines to ground and they look to be OK.
In the document it mentions if one end of the bus isn’t terminated with the resistor it could have signal “reflections”.
Yeah, basically as your signal frequency gets high enough that the wavelength is on the order of the length or your line, you start to see wave phenomena on your signal line. If you put a voltage transient on the input, you'll see a propagating wavefront travel down the line - in this case, a twisted pair. Transmission lines have a characteristic impedance that is dependent on the conductor geometry (like 50 or 75 Ohm coax), and if the load isn't impedance matched to the line, the signal will bounce off and send some power back down the line instead of being fully transmitted into the load. If memory serves me correctly, matching the load impedance to the characteristic line impedance gives you no reflection. Twisted pair has a characteristic impedance of around 100 Ohms, hence the 120 Ohm resistors (I assume).
TL; DR transmission lines have to be terminated correctly or weirdness happens.
I was just tracing out some stuff, wondering if there could be any issue related to the PS work. The PS fan control module looks like it grounds to the same place as the DME, but I don't see why that would be an issue.
I also saw x6011 in a few places in the Bentley:
ELE-247
PS Pump Pin X6975 to diag bus somewhere?
ELE-128
ELE-121
ELE-111
ELE-106
x6011 to pin 43 and pin 42 of x60000 (DME)
pin 42 to leak diagnosis pump
pin 43 to O2 sensor.
ELE-98
PS fan relay to PS fan control module
Apparently this is what happens when the pollen is terrible, and I don't want to go outside and work on my own cars. Not sure if any of this helps, but I figured I would toss it in the thread for reference just in case.
I know that’s a different time scale than before, but CAN low doesn’t look so terrible to me there. The signal is pulling too high for some reason, but that could just be the same ringing (reflections) that you’re also seeing on the CAN high. There could also be a minor grounding issue giving you some trouble, but I haven’t fully wrapped my head around how that would work.
I have a sneaking suspicion that you could unplug the DME and put a 120 Ohm resistor across the CAN lines at that connector to suppress the ringing you’re seeing. I’m not sure if that’s totally safe or not though - I’m not really a digital electronics guy.
I know that’s a different time scale than before, but CAN low doesn’t look so terrible to me there. The signal is pulling too high for some reason, but that could just be the same ringing (reflections) that you’re also seeing on the CAN high. There could also be a minor grounding issue giving you some trouble, but I haven’t fully wrapped my head around how that would work.
I have a sneaking suspicion that you could unplug the DME and put a 120 Ohm resistor across the CAN lines at that connector to suppress the ringing you’re seeing. I’m not sure if that’s totally safe or not though - I’m not really a digital electronics guy.
That’s actually a good idea. I could add the resistor and then at least test resistance at the tach to see if it goes to 60 Ohms.
I should be able to scope 2 channels on Tuesday night.
I ordered a more advanced Foxwell scan tool that should arrive tomorrow.
I may take the battery in for a load test tomorrow if they can do it.
I know that’s a different time scale than before, but CAN low doesn’t look so terrible to me there. The signal is pulling too high for some reason, but that could just be the same ringing (reflections) that you’re also seeing on the CAN high. There could also be a minor grounding issue giving you some trouble, but I haven’t fully wrapped my head around how that would work.
I have a sneaking suspicion that you could unplug the DME and put a 120 Ohm resistor across the CAN lines at that connector to suppress the ringing you’re seeing. I’m not sure if that’s totally safe or not though - I’m not really a digital electronics guy.
Haha. I wasn't sure if you would want to take the chance with terminating with a resistor. I couldn't come up with a reason why it would be an issue since it should just be absorbing the power being transmitted into it, but I didn't want to claim it was totally safe either since I've never done it.
Well, the 60 Ohms seems to a step in the right direction. I think this jives with the paragraph in the TIS about checking with a possibly iffy module disconnected, doesn't it? I wonder if the CAN signals look clean with the right termination on the BUS.
Haha. I wasn't sure if you would want to take the chance with terminating with a resistor. I couldn't come up with a reason why it would be an issue since it should just be absorbing the power being transmitted into it, but I didn't want to claim it was totally safe either since I've never done it.
Well, the 60 Ohms seems to a step in the right direction. I think this jives with the paragraph in the TIS about checking with a possibly iffy module disconnected, doesn't it? I wonder if the CAN signals look clean with the right termination on the BUS.
Well, I think CAN bus signal looked bad on scope with just ECU disconnected. I kept wondering about the TIS and if you disconnect the ECU, you lose that termination resistor but I couldn’t find any info about that online.
Well, I think CAN bus signal looked bad on scope with just ECU disconnected. I kept wondering about the TIS and if you disconnect the ECU, you lose that termination resistor but I couldn’t find any info about that online.
Oh, did you have the ECU disconnected before? Yeah, you should lose the proper termination on the bus if you disconnect either the ECU or the cluster since it seems BMW mounted the resistors internally.
Or I guess I should be more specific and ask if post #104 had the ECU connected for the measurements. I assumed everything but the tach was connected for those.
EDIT: I measure this resistance again down below and get a measurement in the Mega Ohms.
I checked resistance between CAN Hi and CAN Lo pins 29, 30 on ECU. It should be 120 Ohms (I assume) since it’s a termination point. If it wasn’t a termination point I think it should be around 28K Ohms from what I saw online for CAN bus modules.
But I’m getting 126K Ohms which seems out of spec even for a non-terminating module. Fishy?
Last edited by audihere; Apr 27, 2023 at 10:37 PM.
I'm reading the document the same way you are - that you should have a terminating resistor across those pins inside the ECU. If that's the case, you should see 120 Ohms looking into the ECU, so, yes, I agree that 126 kOhms is fishy.
IIRC, you had checked this earlier and not gotten 120 Ohms then either. That's why I suspected that terminating the bus with the 120 Ohm resistor in the connector like you've done would clean up the CAN waveform. If the termination inside the ECU is bad, and it's a high impedance, it will just bounce the signal off and send it back down the line. In fact, now that I think about it, the reflection coefficient of an open circuit transmission line is -1, so that might explain the waveform you were seeing on CAN low.
If the CAN bus system were fully functioning, you would read 60 Ohms with both terminating resistors in place because they're in parallel. If you unplug either module with the terminating resistor, you'd read 120 Ohms because you would just see the one resistor. If you unplugged both terminating modules, you should just see a very high impedance.
Edit: Dunno why I explained this. I know that you know that.
Last edited by deepgrey; Apr 24, 2023 at 04:41 PM.
I’m in contact with Precision ECU and I rechecked resistance from CAN Hi to CAN Lo.
He said it should 100% be 120 Ohms between pins 29 and 30!!
New measurement pic on pins 29 and 30 at 13 Mega Ohms. Don’t know why it’s different unless last pic was wrong. I thought it was Mega Ohms when I took that pic and was surprised to see Kilo Ohms.
Well, that confirms it then. I'm not super well-versed in failure modes, but I would guess the resistance is changing because of some damage inside the ECU. Hopefully it's something really simple like a cold solder joint on that resistor.
I was really hoping it wasn't the ECU, but I was afraid it was, which was partially why I was still looking through connections to the ECU through X6011 the other day. There has to be some explanation for the failure, and if it was caused by something external, I wouldn't want it to happen again.
This also probably explains your transmission issues. IIRC from when I took a gander through the auto trans docs, the transmission control unit needs to communicate with the DME for some critical functions.
We took the DME and the EWS out of the car. Overnighting those and a key to Precision ECU tomorrow. They will read the ECU (with mods) and program a refurbished ECU to my VIN and overnight back.
Just wanted to add these pics. We had some confusion finding pins 29 and 30 on the X60004 connection of the ECU.
The molded in pin numbers don’t match the wiring diagrams.
Going by the molded pin numbers, pins 29 and 30 would be pins 110 and 111.
They continue over from the numbering in the larger connector. Maybe there is a difference between the numbering molded into the EMS2000 and the facelift MS5150 ECU’s?
