Author Topic: DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION  (Read 14887 times)

jcaf

  • Member
  • ***
  • Posts: 3
DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION
« on: February 06, 2014, 10:43:30 pm »
Hi everybody,

It's true that the impedance of 100ohms of a UTP
CAT5E twisted-pair cable is its DIFFERENTIAL IMPEDANCE?

I am routing a PCB, a RS485 half duplex conexion, and Should I route
the tracks A and B  with a differential impedance equal to 100 ohm to
the conector?

(A and B are microstrip because my PCB is double-layer)

Using Saturn PCB Design (www.saturnpcb.com) - PCB Toolkit V6.6

Width of traces A = B = 40 mils
Conductor spacing = 12 mils
Conductor height = 62 mils

Bse Copper weight = 18um
Er=4.5

Results:
Z diff = 102.714 ohms
Zo = 85.397 ohm

It's OK my calculation for routing my PCB?

Thank you very much for your attention

Juan
« Last Edit: February 07, 2014, 07:44:18 pm by jcaf »

Jan Axelson

  • Administrator
  • Frequent Contributor
  • *****
  • Posts: 3033
    • Lakeview Research
Re: DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION
« Reply #1 on: February 07, 2014, 09:44:58 am »
Yes, CAT5 cable has 100 ohm differential impedance and should be fine for 1-way RS-485 with a single 100-ohm termination across the lines.

If the link is 2-way, 120-ohm cable is better because two 100-ohm terminations brings the parallel combination to 50 ohms, which is less that TIA-485's specified minimum. With that said, most RS-485 drivers can source and sink 60 mA and thus will work with 100-ohm cable.

Depending on the drivers and cable length, you might not need terminating resistors, and in that case, the current isn't an issue.

As for the PCB routing, yes, I would say that matching the traces to the cable's differential impedance is a good thing.

jcaf

  • Member
  • ***
  • Posts: 3
Re: DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION
« Reply #2 on: February 07, 2014, 11:28:00 am »
Ms. Jan Axelson, I am grateful for your words.
I am an admirer and follower of their work.

I congratulate you with all my heart.

Greetings from Lima - Perú   :D

Jan Axelson

  • Administrator
  • Frequent Contributor
  • *****
  • Posts: 3033
    • Lakeview Research
Re: DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION
« Reply #3 on: February 07, 2014, 11:30:22 am »
Glad to help!

jcaf

  • Member
  • ***
  • Posts: 3
Re: DIFFERENTIAL IMPEDANCE IN UTP-CAT5 FOR RS485 - CALCULATION
« Reply #4 on: February 07, 2014, 07:39:42 pm »
Additional:

1.-
http://cp.literature.agilent.com/litweb/pdf/5989-5763EN.pdf

Agilent- Signal Integrity Analysis Series
Part 1: Single-Port TDR, TDR/TDT, and 2-Port TDR Application Note
   4.4 Measuring the differential impedance of a twisted pair cable.................Page 65


2.-
http://www.highfrequencyelectronics.com/Archives/Nov02/HFE1102_Lao.pdf

The Twisted-Pair Telephone Transmission Line
By Richard LAO
Sumida America Technologies
Page  26

"The balun transformer is necessary because we wish to measure differential impedance, not common mode impedance".
It is DIFFERENTIAL IMPEDANCE that cable manufacturers specify in their data sheets.