PicoScope 7 Automotive
Available for Windows, Mac, and Linux, the next evolution of our diagnostic scope software is now available.
Canbus issues
Canbus issues
Hi,
The attached waveform is from a Ford Kuga suffering can communication issues between engine ecu, abs module and instrument cluster, Can anyone explain the anomaly in the middle of the waveform, I have disconnected all ecu's and it still appears, going around in circle's with this one, also doesn't seam to decode?
The attached waveform is from a Ford Kuga suffering can communication issues between engine ecu, abs module and instrument cluster, Can anyone explain the anomaly in the middle of the waveform, I have disconnected all ecu's and it still appears, going around in circle's with this one, also doesn't seam to decode?
- Attachments
-
- Ford_Kuga_2008_20230808-0001.psdata
- (65.63 KiB) Downloaded 110 times
-
ben.martins
- Pico Staff Member
- Posts: 545
- Joined: Tue May 16, 2017 1:02 pm
Re: Canbus issues
Hello Burt_100,
Thank you for your post.
Lets tackle one thing at a time and to start with that it seems to not decode.
This was due to a slight oversight in the decoder settings where the threshold was set to 3.6V and so outside of where the packet actually is. The decoder will try to fit the signal with an automatic threshold but it's more often that not incorrect especially if the acknowledgement bit is high. Instead, drop your threshold to the middle point of CAN H which if we went between 2.5 and 3.5, 3 V would be somewhere in middle. Alternatively use the math channel A-B to display the differential of CAN H and CAN L allowing you to see what the ECU is seeing. As this is 0-2V signal, 1V can be used for our crossing point. This is useful when there is a lot of interference affecting both CAN H and CAN L that will be cancelled out in the ECU's.
When you complete the decoder you will now see that the data decodes successfully.
I've also included another math channel, A+B where we add CAN H to CAN L and so should end up with a fairly flat 5V line. In your capture that isn't so as seen in the above image. The only time it is at 5V is when the packet with ID 190 is communicating which appears to be our anomaly.
I think in order to see if this an issue with just this ID or something else it would worthwhile getting more time on the screen to capture more data. This can easily be done by using the built-in guided test for serial decoding which will set the scope and sample rate up for you, along with the math channel and the serial decoder.
I would also draw up the topology of the network along with noting the fault codes. This can then start to give you some direction to see where the ECU's are pointing towards.
I hope this helps.
Kind regards
Ben
Thank you for your post.
Lets tackle one thing at a time and to start with that it seems to not decode.
This was due to a slight oversight in the decoder settings where the threshold was set to 3.6V and so outside of where the packet actually is. The decoder will try to fit the signal with an automatic threshold but it's more often that not incorrect especially if the acknowledgement bit is high. Instead, drop your threshold to the middle point of CAN H which if we went between 2.5 and 3.5, 3 V would be somewhere in middle. Alternatively use the math channel A-B to display the differential of CAN H and CAN L allowing you to see what the ECU is seeing. As this is 0-2V signal, 1V can be used for our crossing point. This is useful when there is a lot of interference affecting both CAN H and CAN L that will be cancelled out in the ECU's.
I think in order to see if this an issue with just this ID or something else it would worthwhile getting more time on the screen to capture more data. This can easily be done by using the built-in guided test for serial decoding which will set the scope and sample rate up for you, along with the math channel and the serial decoder.
I would also draw up the topology of the network along with noting the fault codes. This can then start to give you some direction to see where the ECU's are pointing towards.
I hope this helps.
Kind regards
Ben