Author Topic: RS422 terms at both ends?  (Read 1125 times)

0 Members and 1 Guest are viewing this topic.

Offline bitbangerTopic starter

  • Regular Contributor
  • *
  • Posts: 164
RS422 terms at both ends?
« on: August 26, 2019, 02:24:27 pm »
Under what circumstances would one want to parallel terminate both the TX and RX ends of a 422 network with 120ohm? I see this referenced every once and a while. Is the hope to diminish reflections at both ends? Does this not inherently conflict with matching cabling characteristic impedance of ~100-120ohm?

One example:
http://www.interfacebus.com/Design_Connector_RS422.html
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: RS422 terms at both ends?
« Reply #1 on: August 26, 2019, 03:56:56 pm »
See Figure 4

https://www.maximintegrated.com/en/app-notes/index.mvp/id/723

From the point of view of the master, the transmit pair are terminated on the far end and the receive pair on the near end.

 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: RS422 terms at both ends?
« Reply #2 on: August 26, 2019, 04:43:56 pm »
For bidirectional operation, the transmission line is terminated at both ends.  The RS-422 transmitter only sees the near termination as an additional load.  The termination on the far end from the transmitter is the one which matters.

This is also why transmitters in the middle do not require any termination on their output at all.   They see two parallel transmission line loads and both end terminations absorb reflections.
« Last Edit: August 26, 2019, 04:46:42 pm by David Hess »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: RS422 terms at both ends?
« Reply #3 on: August 26, 2019, 11:04:41 pm »
422 is always-on so the transmitter impedance is already very low (typically less than Zo) and so you're only wasting power and range while doing nothing for reflections.

485 has an enable so there may be some value in terminating the transmitter -- namely if it's at the end of a line as mentioned above.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline bitbangerTopic starter

  • Regular Contributor
  • *
  • Posts: 164
Re: RS422 terms at both ends?
« Reply #4 on: August 27, 2019, 12:51:53 am »
Sorry perhaps I should have been a little more clear. In the included link, they provide calculations for fail-safe resistor network. The calculations seems to imply that on a single pair, from TX to RX, you would have 120ohm on each end of the pair. No matter multi-drop or not, each RS422 pair is still considered "one direction", there is no TX enable like 485, so I don't see the point in terminating both ends.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf