Shop OBEX P1 Docs P2 Docs Learn Events
Point-to-point EIA-422 — Parallax Forums

Point-to-point EIA-422

InSilicoInSilico Posts: 52
edited 2009-04-17 22:16 in General Discussion
Hi,

I'm attempting to implement RS-422 in a point-to-point configuration to replace a RS-232 setup I have (for longer distances, higher noise immunity, etc.) I'm thinking of obtaining some SN75179BP ICs (http://www.mouser.com/Search/ProductDetail.aspx?qs=sGAEpiMZZMutXGli8Ay4kLJm31Vsi2pIdTq%252bzqwUdjY%3d)·to implement full-duplex over UTP·wire like Cat3 cable and·connect them with modular connectors.

The question is I don't know if I'm supposed to use any biasing or terminating resistors in a point-to-point setup. A National appnote (http://www.national.com/an/AN/AN-1031.pdf) shows a point-to-point setup with the driver and receiver connected together (+ to +, - to -, etc.) with no resistors, but with a suggestion to use TransZorbs.

Your second opinion·is much appreciated!

Comments

  • stamptrolstamptrol Posts: 1,731
    edited 2009-04-16 12:12
    I have used the 75176 chips with half-duplex over a CAT5 pair over 1000 feet or more. The recommended standard uses termination resistors. I expect the 75179 chips would as well. At short distances, the resistors probably don't make much difference. If you're going between buildings with separate electrical services, the Transorbs are a good idea.

    For practical application notes, B & B Electronics has numerous white papers free for downloading.

    Cheers,

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tom Sisk

    http://www.siskconsult.com
    ·
  • InSilicoInSilico Posts: 52
    edited 2009-04-16 22:48
    stamptrol, did you use terminating resistors in your applications?

    I·may use the fail-safe termination scheme (see image below, from http://www.maxim-ic.com/appnotes.cfm/an_pk/723/) at the receiver end because the receiver in the SN75179 has an undefined output if there isn't a voltage difference between them (which will happen if the cable is disconnected). The pullup/pulldown resistors should force the lines to a known state. I will test this out.

    723Fig07.gif

    Since some of the nodes will be in a remote location and are low-current consuming, I will obtain 6 conductor UTP, 4 for the full duplex differential lines, 1 for power (it will be higher than +5V, since cable losses must be taken into account), and 1 for ground.
  • stamptrolstamptrol Posts: 1,731
    edited 2009-04-17 12:21
    I used termination resistors at the extreme ends of the system. One at the master and then at the furthest slave.

    I used CAT5 cable: one pair for the communication (half-duplex, multi-drop), two pairs parallelled for the 12 VDC supply to the remote stations, the rest as spares.

    BTW, Jan Axelsons's book "Serial Port Complete" has some very practical pointers on serial communication in all its flavours.

    Cheers,

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tom Sisk

    http://www.siskconsult.com
    ·
  • InSilicoInSilico Posts: 52
    edited 2009-04-17 22:16
    Thanks for your response! Now onto placing the order for the parts.
Sign In or Register to comment.