IET
Decrease font size
Increase font size
Topic Title: RS485 Resistance Measurement
Topic Summary: Inconsistent measurement of resistance over the RS485 network
Created On: 11 November 2013 10:51 AM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 11 November 2013 10:51 AM
User is offline View Users Profile Print this message



carlcheetham

Posts: 4
Joined: 29 March 2012

I recently was asked to fault find on a large RS485 network and experienced some odd results whilst trying to measure the resistance of the loop.

I had one of the termination resistors out of circuit and was expecting to measure the resistance of the wire and the other termination resistor. I was expecting approx. 160ohms for the kilometre of wire and the 120ohm resistor so 280 ohms ish in total.

I was reading values in the mega-ohm range which went up when I added the termination resistor at the other end of the wire which obviously should have pulled the resistance down?

The cables used are not shielded and do run within a few feet of fluorescent lighting luminaires hence there is quite a lot of noise on the signals. (when I work out how to post a picture I will put up the waveforms)

Any thoughts or help welcome...
 11 November 2013 12:09 PM
User is offline View Users Profile Print this message



jarathoon

Posts: 1043
Joined: 05 September 2004

There aren't many variables here...High noise suggests the two wires are open circuit at the other end and acting like an aerial.

Two Questions
1. Are the termination resistors still resistors? Have you extracted and checked the resistors at both ends to make sure they haven't been burnt out by a current surge, perhaps due to an electrical storm event? (After a current surge resistors can develop fine cracks, which can only be seen under close examination.)

2. Is there a wire break? If there is an earth line between the two locations you could use that to check each of the two RS485 wires individually for breaks, by attaching them to earth via two different resistances (to save on journeys).

-------------------------
James Arathoon
 11 November 2013 12:17 PM
User is offline View Users Profile Print this message



carlcheetham

Posts: 4
Joined: 29 March 2012

Both resistors were measuring around 117 ohms, and at the time of this testing the network was working so I don't think there was a wire break.

The reason I was there was due to a number of intermittent comms faults occurring for 2-3 days after an electrical storm. When I attended site there had been no failure for three days.

I thought I could measure the resistances to verify the network cabling but started to get odd readings on the meter even thought the network was working!?
 11 November 2013 12:19 PM
User is offline View Users Profile Print this message



carlcheetham

Posts: 4
Joined: 29 March 2012

Whilst measuring these resistances I had removed the master node from the network so there was no network activity but the 28 receiving nodes were still connected.

Can anyone think why this may have caused odd meter readings?
 11 November 2013 01:01 PM
User is offline View Users Profile Print this message



jarathoon

Posts: 1043
Joined: 05 September 2004

Is it a three wire bus (differential signal + earth) or two wire (differential signal only)?

If your system is a robust three wire RS485 system it may still work (just) with a wire break in one of the differential signal wires. Some of the slaves may work more reliably than others in this case (e.g. the ones nearer the master node).

If it is a two wire system then the slave nodes after the break would not work, and those before the break would have an increased sensitivity to external electrical interference.




-------------------------
James Arathoon
 11 November 2013 10:21 PM
User is offline View Users Profile Print this message



carlcheetham

Posts: 4
Joined: 29 March 2012

IT is a two wire bus cobbled together from existing site twisted pair cabling.

The faults when present were intermittent but everything failed then recovered...
 12 November 2013 11:06 AM
User is offline View Users Profile Print this message



jarathoon

Posts: 1043
Joined: 05 September 2004

With a two wire system you are relying on the fact that local earthing points stay roughly at the same potential. This won't necessarily remain true in an electrical storm event for slave nodes distributed over a km or so.

I am sorry your observations don't make sense to me. I think you are just going to systematically fault find until you get some added inspiration from your measurements on this one.

When you get the opportunity switch off and disconnect all the slaves and insert the termination resistor at the nearest slave point and see if this makes any difference to your cable resistance measurements.

If it does then do the same thing at the middle slave and so on...

If the wiring is fine then physically reconnect the slaves one by one and see what happens to your measurements each time.




-------------------------
James Arathoon
 17 November 2013 02:04 PM
User is offline View Users Profile Print this message



kengreen

Posts: 400
Joined: 15 April 2013

Carl,

Although I risk being drummed out of the Brownies the above posts have raised wry smiles in front of this computer. There are times when you have to disregard the rule-book and use your professional nonce to render a system safe - AND you have to take precautions that anyone who has to service the installation after your departure will be aware of the problems.

My property here sits on several hundred feet of slate - laid down under the sea - and divided into discrete layers by very-thin layers of sand and shale. The small stream at the bottom of the V-shaped valley, for which it is responsible, has left this property with a series of terraces the top layers of which is turf which varies in depth from 0 to a maximum of around 27 inches. If you want to know something about measuring Earth loops but all means come and try - particularly when you measure between different terraces!

I sunk a 4-foot earthing rod in a patch that was kept watered by a surface spring and connected ALL my RCD's to that - in 40 years I have not had one single accident nor one complaint although there was an occasion when a badly wired bedside clock shutdown the entire installation on average once per day - the clown had connected the 2-wire mains lead between L and E ?

I have a real problem today when "electricians" are called in by my daughter and they proceed to rip out my wiring of up to 40 years and revert to the IEE regulations; when I try to give explanations they simply walk away from the old fool! ! ! However I am assured by my legal advisers that responsibility now rests with them - which is a "great comfort"?

Ken Green
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.