Other Parts Discussed in Thread: TCAN4550
Hello,
We are current developing a new device with 3 CAN channels. 2 CAN channels are controlled directly from the MCU. The 3rd channel is controlled via TCAN.
During testing we have noticed that talking to products we have developed in the past with the TCAN causes them to fault due to numerus CAN frame errors. This doesn't occur with the native CAN channels from our MCU.
Furthermore we noticed it is part by part (Same product but different serial numbers) dependent if it will fault on the bus. The weird thing is it tends to only be TX errors form the module. Not RX errors. As in, the product will always act upon the commands received but can't transmit status frames back. Its almost as if TCAN is holding the buss in a recessive state, which I can't see how it could.
We also noticed this interesting waveform from TCAN (attached). Looks like its having issues with regulating the +5V rail. Scope shots have proved otherwise. Any thoughts on this? Or what other things to look for?
Currently the test setup is two of our products on the test bench with 6 M of twisted pare cable. 2 120 om resistors on ether end with the attached modules connected within 2' of the ends. (one on each end)
Here is how it is in our schematic:
