Author Topic: RS485 termination resistor  (Read 3255 times)

0 Members and 1 Guest are viewing this topic.

Offline hamdi.tnTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: tn
RS485 termination resistor
« on: May 29, 2016, 06:54:54 pm »
Hi,

Am starting a project where i aim to connect 100 unit with an RS485 bus, and the length of this bus could be around 1KM, and i was reading this document about RS485 from TI
http://www.ti.com/lit/an/slla272b/slla272b.pdf

And looking at Bus loading (Page 4) they said
Quote
failsafe biasing contributes up to 20 unit loads of bus loading
i don't get how they calculated that !! any help please.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3893
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: RS485 termination resistor
« Reply #1 on: May 29, 2016, 07:17:39 pm »
Capacitance is the main issue.

I would highly recommend to split the cable run to 100m (maximum) sections and use sort of signal repeaters.

Termination resistors have to be used on both ends (of each cable section) and the stub length have to be minimized.

What baud rate are we talking about?


And by the way, the appnote talks about the bus load units and how the max number of stubs can be calculated.
« Last Edit: May 29, 2016, 07:23:03 pm by Yansi »
 

Offline NW27

  • Regular Contributor
  • *
  • Posts: 53
  • Country: au
Re: RS485 termination resistor
« Reply #2 on: May 29, 2016, 07:22:52 pm »
Using their figures, a standard input presents a 12k input impedance.
12k / 20 = 600. 600 is approximately the value of the Rb + (Rt/2) = (532+60) = 592.
 
The following users thanked this post: hamdi.tn

Offline hamdi.tnTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: tn
Re: RS485 termination resistor
« Reply #3 on: May 29, 2016, 07:38:52 pm »
Capacitance is the main issue.

I would highly recommend to split the cable run to 100m (maximum) sections and use sort of signal repeaters.

Termination resistors have to be used on both ends (of each cable section) and the stub length have to be minimized.

What baud rate are we talking about?


And by the way, the appnote talks about the bus load units and how the max number of stubs can be calculated.

for all those problem i will keep the baud rate to minimum, the amount of data to be transferred is not huge and timing is nor critical either so the lowest practical baud will be used.

And i will design a repeater too, to solve any problem on the field.

yeah i saw how max number of unit is calculated , it's just the detail that those termination resistors are equivalent to 20UL that i don't understand
 

Offline hamdi.tnTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: tn
Re: RS485 termination resistor
« Reply #4 on: May 29, 2016, 07:54:25 pm »
Using their figures, a standard input presents a 12k input impedance.
12k / 20 = 600. 600 is approximately the value of the Rb + (Rt/2) = (532+60) = 592.

yeah that's it !  :-+
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13987
  • Country: gb
    • Mike's Electric Stuff
Re: RS485 termination resistor
« Reply #5 on: May 29, 2016, 08:27:15 pm »
Capacitance is the main issue.

Cable resistance is also a significant limiting factor, so for long runs, cable selection is important, i.e. not CAT5!
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28013
  • Country: nl
    • NCT Developments
Re: RS485 termination resistor
« Reply #6 on: May 29, 2016, 08:57:18 pm »
Still you'll find it will be surprisingly difficult to find cable which has a low capacitance! At long distances it is necessary to use cable (or coax) with a well defined impedance (usually 100 Ohms) and terminate it properly so it at least looks like a resistive load to the driver. Nevertheless ane (copper) cable will have a significant attenuation (increasing with frequency) due to the resistance of copper and the capacitance. At 1km I think shielded twisted pair is about the best you can do before going to optical fiber.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2331
  • Country: 00
Re: RS485 termination resistor
« Reply #7 on: May 29, 2016, 09:16:21 pm »
The speed of the communication is a crucial point, if you plan to do at very low speeds even nothing will work, but take a note that if you plan to have 100 units on the same line will not be a easy task to do without using repeaters.

No general answer is possible, you need to know the exactly cable characteristic, a practical "way " is to put a trimpot near 500R and with a scope adjust to have a best "square wave" on full speed.
 

Offline hamdi.tnTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: tn
Re: RS485 termination resistor
« Reply #8 on: May 29, 2016, 09:56:57 pm »
At 1km I think shielded twisted pair is about the best you can do before going to optical fiber.

That's the plan , to use shielded twisted pair.
also what am planning to do is to use the units boards to link between those twisted pairs, each board will have two RJ45 connector that will link the previous unit to the next unit. am sure the pcb itself will only make it worst.

but take a note that if you plan to have 100 units on the same line will not be a easy task to do without using repeaters.

the most annoying thing for me is that even if this work in the lab , there is 1000 reasons not to work when they install the units. So yeah am not eliminating the need for repeaters.

it's the first time i have to use rs485 bus on such scale, so every datasheet and app note are a bit confusing about the limit of the system and the number of parameters involved. yeah this will not be easy.

 

Offline ebclr

  • Super Contributor
  • ***
  • Posts: 2331
  • Country: 00
Re: RS485 termination resistor
« Reply #9 on: May 29, 2016, 10:36:22 pm »
Don't forget a Hugh lightning protection
 

Offline hamdi.tnTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: tn
Re: RS485 termination resistor
« Reply #10 on: May 30, 2016, 02:23:12 pm »
Capacitance is the main issue.

Cable resistance is also a significant limiting factor, so for long runs, cable selection is important, i.e. not CAT5!

we will use CAT6
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13987
  • Country: gb
    • Mike's Electric Stuff
Re: RS485 termination resistor
« Reply #11 on: May 30, 2016, 04:19:30 pm »
Capacitance is the main issue.

Cable resistance is also a significant limiting factor, so for long runs, cable selection is important, i.e. not CAT5!

we will use CAT6
For 1km you'll need something more like 22AWG. The DC resistance of cat6 may be too high for reliable operation at any baudrate
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf