0
\$\begingroup\$

I'm working on a small custom PCB that uses the TI SN65VHD230 CAN Transceiver and the square waveforms of the CAN bus aren't as clean and crisp as it seems they should be.

I have a 10 μF MLCC on the VCC of the transceiver (0.1 μF passed lots of high frequency noise so I grabbed something I had on hand, 10 μF, and it helped), a 4.7 kΩ pull up resistor on the RX pin as suggested in the data sheet, as well as a 10 kΩ resistor to ground for helping with noise on the RS pin to help with noise. But I still can't seem to get the edges to be clean and the states to be stable.

Power is delivered to the transceiver from an internal MCU Vreg that takes it from 5 V down to 3.3 V. The MCU is fed 5 V from an LDO which is fed with 12 V. I did notice that I'm getting about 450 mV of ripple on the 12 V line that seems to be coming from the LDO. I'm not sure if that's affecting the CAN bus issue at all.. I wouldn't think so being that it's going through so many stages of voltage regulation but... maybe someone knows better than me.

enter image description here

\$\endgroup\$
4
  • \$\begingroup\$ What are the names of the shown waveforms ? Blue wave ? Schematic used? \$\endgroup\$
    – Antonio51
    Commented Aug 25, 2022 at 7:12
  • \$\begingroup\$ So what devices is on the bus? What kind of schematics your CAN device has? Why are you looking at CANH and CANL separately, as the difference of CANH-CANL might look just fine. So basically, what kind of waveforms you expect and why, assuming you know how CAN bus works and why it is different from other differential buses like RS-485? \$\endgroup\$
    – Justme
    Commented Aug 25, 2022 at 8:53
  • 1
    \$\begingroup\$ You should subtract CANH and CANL, instead of adding them. Configure the Math channel as DIFF(C1,C2). Then you'll notice that the rounded edges of the recessive bits aren't that great of a problem. \$\endgroup\$
    – Velvet
    Commented Aug 25, 2022 at 9:11
  • \$\begingroup\$ Thank you. Is tere anything I can do to improve this anyways? Of is this inherent in the chip im using st this point? \$\endgroup\$ Commented Aug 26, 2022 at 15:30

0