This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Capacitor recommendation for SN65HVD234 slope control line

Other Parts Discussed in Thread: SN65HVD234

I'm working on a product that uses SN65HVD234 for the CAN bus, but a number of devices in the current lot have very large numbers of CAN errors. We've had quite a few lots where we didn't notice this happening, so it seems like previous devices haven't had this problem, or it has at least been less severe. I've found that adding a small amount of capacitance on the Rs line fixes the problem, but I'm not sure that I have a good explanation of what's going on, so I'm hesitant to implement it until I know more. Our product is designed to use the slope control feature of the SN65HVD234 by pulling the Rs line down with 27.4 kOhms. I've experimented with stopping CAN errors from happening by adding 100 pF in parallel with the 27.4 kOhms.

The January 2015 version of the datasheet shows a capacitor, labeled C4 in the suggested layout, in parallel with the pull down resistor, but says that it's not needed when using slope control mode. Would there be any problems expected with adding 100 pF in parallel with the 27.4 kOhms, or is that in line with recommended use? Does it make sense that a capacitor on the Rs pin would be able to fix a problem with bad packets?

  • Hi Alex,

    It would be OK to put a capacitor in parallel with the slope control resistor. This is a DC control signal rather than a dynamic one, and so there isn't a risk of over-filtering it. That said, I haven't heard of this capacitor being required for proper operation and would be interested to get a better understanding of the issue you are seeing. Were you able to observe the CANH/CANL waveforms under both conditions (with and without the Rs capacitance) to see what specifically is going wrong? Or, is it possible to observe the Rs pin to see if there is some noise coupling onto it when a capacitor is not used?

    Regards,
    Max
  • I can't directly observe what's happening on the Rs pin because even the capacitance of the scope probe fixes the problem. I took some screenshots of CANL waveforms. The first screen shows CANL sampled with persistence on, and no capacitor. The second screenshot shows samples with persistence, and the capacitor added. As evidenced in the screenshots, the poorly formed signals seem to immediately and entirely disappear when I add the capacitance on Rs. The third screenshot shows a few poorly formed bits without the capacitor.

    One non-standard aspect of our bus is that it is made to have multiple devices plugged or unplugged from it, so the bus isn't one line. There is a 120 Ohm termination at the master transceiver, but not in any of the pluggable devices. The results that I've been looking at have been with adding the capacitor to the master transceiver. I haven't tried it on the transceivers in any of the pluggable devices yet because they're hard to open.

  • Alex,

    Thanks for providing the waveforms - that does look strange. We will set up the same test in our lab to see if the same behavior is observed. We ordered some new units for the testing, and they should be here in a few days. I'll let you know what we see.

    Best regards,
    Max
  • Alex,

    I wanted to follow up on this. We set up the devices in our lab with various pull-down resistances (and no capacitances) and observed the CANH and CANL lines. We just saw the expected waveforms, though, no noise like your scope shots show. We used a CAN evaluation board for this test: www.ti.com/.../SN65HVD255DEVM. I suspect the parasitic capacitance on the RS line on this PCB isn't much higher than what is on your board, but I wanted to double-check with you. If the capacitances are similar, maybe you could send us some of the units that showed this behavior so we can check them here?

    Max