2
\$\begingroup\$

Just tried to simulate the communication line with the series resistance in it:

enter image description here

As you can see even at 1 MBPS the signal at the "receiver" is fairly good. I'm pretty sure that this will be OK to receive the signal.

The reason I tried to think about the series resistance is an intention to put the resistors at the reseiver with a parallel TVS (or Zener) diodes to protect the receiver (I will need to have a long transmition line). After I put the zeners the falls and rises become much slower (I beleive that it was due to diodes parasitic capacitance):

enter image description here

I think it would be a bad idea to put TVS without a resistors.

15 pF capacity I took from some RS-485 transceivers datasheet. Actually I was able to find the receiver input capacitance only in the TI's AM26C31ID datasheet (it was 6 pF). Probably there are other devices with this parameter specified but many from MAXIM and Analog Devices was not :(

So it looks that the wire series resistance is not an issue for communication under real circumstances. Are there any other points I am not aware of?

UPDATE:

I made a simulation with lossy transmition line (LTRA) and came to the conclusion that the series resistance is not influence much to the signal (it is 1 MBps, at 200 meters, which is pretty long for this speed):

enter image description here

However this resistance CAN influence the signal in some cases:

  • if the termination resistor placed AFTER there resistance
  • if the series resistance become close to the input impedance of the receiver which leads to significant amplitude reduction.
\$\endgroup\$
3
  • \$\begingroup\$ What is the output impedance of the source? What about the parasitic capacitances and inductances down the wires. Your model is way too simple. \$\endgroup\$
    – Trevor_G
    Jan 8, 2018 at 16:28
  • \$\begingroup\$ @Trevor, thank you for the correction and for the comment. I will do some clarification in the post to make it more clear. \$\endgroup\$
    – Roman Matveev
    Jan 8, 2018 at 16:34
  • \$\begingroup\$ Also: there are integrated dual-diode packages designed to clamp to RS485 voltages. See for instance the Bourns CDSOT23-SM712 or the Semtech SM712.TCT. \$\endgroup\$
    – Reinderien
    Jan 14, 2018 at 21:59

2 Answers 2

Reset to default
4
\$\begingroup\$

That is an oversimplified model. You should use a lossy transmission line with terminations to properly simulate the circuit.

In LTSPICE the model is LTRA. Expect it to take a rather long time to run.

The maximum recommended length is 4000 feet. Assuming AWG 24 wire, that's 200 ohms round-trip. Consider the extremes of common mode voltage and loading (32 unit loads).

enter image description here

Much more in the NS (now TI) application note [10 Ways to Bulletproof RS-485 Interfaces] (source of the above graph).

Probably if you are willing to relax the requirements from the standard, reduce data rate and use heavier gauge wire you can use longer wires. You would probably want to go to isolated drivers and/or receivers since it's difficult to guarantee the common-mode voltage over such distances.

\$\endgroup\$
3
  • \$\begingroup\$ could you review my last update in the question? \$\endgroup\$
    – Roman Matveev
    Jan 13, 2018 at 11:54
  • 1
    \$\begingroup\$ The termination resistor must be placed at the ends so the signal will be decreased. Otherwise you will get reflections at the unterminated end. \$\endgroup\$
    – Spehro Pefhany
    Jan 14, 2018 at 2:38
  • \$\begingroup\$ As I see from the simulation schematic 100 Ohms resistors are placet exactly at the ends of the line. 1K branches supposed to be short. \$\endgroup\$
    – Roman Matveev
    Jan 14, 2018 at 10:18
0
\$\begingroup\$

As this series resistance forming a resistive divider with a termination resistors, it is very impotant to keep this resistance as low as possible.

enter image description here

From the RS-485 specs:

  • transducer should make at leas 1.5V signal at its output
  • receiver should receive a signal at least 200 mV on its input

This leads to the conclusion that, 150 Ohms of series resistance (or greater) will not guarantee that the signal will be received.

So, 150 Ohms (for each end) can be a rought estimation on the maximum series resistance of the rs-485 network.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.