• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 34 subscribers
  • Views 2698 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

12V RS485

Daniel Berces91912
Prodigy 75 points

Other Parts Discussed in Thread: SN65HVD08

Hi!

I have to install an RS485 bus on non-twisted, shielded (alarm) cable.

I plan to use SN65HVD08.

Because of  the lack of twisted pair, I would like to go for a 12V bus (instead of the usual 5V) - to get more reliable operation.

While supplying the SN65HVD08 with 3.3V (or 5V), can I pull up the bus to 12V with properly calculated bias resistors?

More or less I want to apply the method of "failsafe biasing" (however SN65HVD08 already contains this as far as I understand) - pulling up line A to 12V, and line B to GND - just for the purpose of bigger voltage range.

Do you think it is feasible?

Thank you very much in advance!

Best regards,

Daniel

  • 0
    TI__Expert 5400 points

    Daniel,

    I'm not sure I understand the motivation for such a configuration.  The extra resistors would actually present additional loading to the driver, in fact making the load heavier, and more difficult to drive.

    The primary challenge introduced by using untwisted wires is that of common mode noise, which adding these additional resistors would do little to remedy.  Aside from, of course, using twisted pair cable, you might consider adding common-mode filtering to your bus, as described on page 3 of this document.

  • 0 in reply to Bart Stiller
    Prodigy 75 points

    Hi Bart,

    thank you for your fast and helpful reply!

    So I will go with 5V bus, and put low-pass filters at the end of the line.

    Apparently the next challenges are:

    • Measuring the impedance of the untwisted pair
      • My plan is to send square-wave into the cable, while putting a potentiometer at the other end, and watch for the resistance-value where no reflections occur with an oscilloscope.
    • Selecting optimal value for the low-pass filter's capacitance.

    Could you please confirm that I'm on the right way (having no chance for installing twisted pair)?

    Thank you very much in advance!

    Best regards,

    Daniel

  • 0 in reply to Daniel Berces91912
    TI__Expert 5400 points

    Daniel,

    If you do not know the impedance of your cable from the manufacturer, that is one way to get a good estimate.  If you have access to a network analyzer you could achieve a similar result using the return loss, or a Smith chart.  Just more options.

    After looking a bit closer at the "common mode filter" I mentioned, I'm not sure it will add much noise immunity to your bus.  A simple way to think about it is that the midpoint between the two termination resistors is the node that will be filtered, not the two nodes actually going into your receiver.  It wouldn't hurt anything to leave it, but I don't believe it will offer much of an advantage.  Unfortunately I'm not aware of any other simple ways to recover the advantages lost by using untwisted cable.

  • 0 in reply to Bart Stiller
    Prodigy 75 points

    Hi Bart,

    thank you very much for taking your time again for my problem.

    I see.

    However, the only noise-sources I'm aware of are the 230V/50Hz (max. 16A) mains wires running parallel to the shielded-untwisted cable.

    My last hope is that on low baudrates (I don't need more than 9600) it will still work stable - expecting some protection from shielding.