Re: Communications Cable

Hello,

It's true, there are also other things than line impedance to be taken into account. In order
to reduce the use of scope, it is recommended to use a cable intended to be used with CAN and
in specified environment. One should notice, that in market terms, both "CAN-cable" and
"DeviceNet-cable" are common wordings in datasheets.

Existence of failures does depend on edge slopes, not on the bit-rate. The faster edges, the
more significant network topology, line impedance, termination and other parameters become. At
lower bit rates, commonly at 125kbps and lower, bit timing margins are so large that mismatches
or topology violations need to be huge, to be able to cause error-frames. But, even if error-
frames do not exist, there may be significant threats in physical layer. And it is not only
theory -- everything need to be assembled efficiently, which typically means the use of patch
cables, terminator plugs and other kind of standard building blocks. Therefore the only safe
approach is to use a cable or twisted-pair specified for CAN.

Lars-Berno said it right, CAN is really forgiving a lot, but in some circumstances it may cause
negative surprises if one is trying to get it forgiving too much. Unfortunately I have seen all
too many cases and all to close, when too many compromises have been made in cabling. As L-B
briefly mentioned, cabling has close relationship with communication dependability -- and in
that way directly to functional safety performance! If extreme compromises are done during the
design, they have to be considered in details in safety analysis of the target system.

I'm not sure, for what L-B mentioned 1/10000. First it brought into my mind de-facto estimation
for bit-error probability for any kind of twisted-pair medium. Ferreira et.al. has proved in
article "An Experiment to Assess Bit Error Rate in CAN", that for CAN-networks the bit-error
probability is far less than 1/10000, even under extensive electromagnetic emissions.

Best regards,

-H

Heikki Saha, CTO
  M.Sc. Automation
  Dr.Tech. Electronics
TK Engineering Oy
Mail address:  P.O. box 810, FIN-65101 VAASA
Visit address: Yrittäjänkatu 15, FIN-65380 VAASA
+358 (0)50 588 6894
[email protected]
http://www.tke.fi/
http://www.canopen.fi/

----- Original Message -----
From: "Lars-Berno Fredriksson" <[email protected]>
To: [email protected]
Sent: Saturday, July 5, 2014 3:03:17 PM
Subject: Re: [CANLIST] Communications Cable

Cables are often a matter of compromises. Not only CAN has to be taken in account. Flexibility, temp. range, resistance to chemicals etc. has also to be taken into consideration. It is a matter of engineering. A mismatch in impedance will cause reflections wherever they occur and often the end points are not the worst spots. 100 ohm may work well if the rest of the network is done fairly well, the nodes have good oscillators and the bit rate is not too high. Use an oscilloscope and check that the signal quality is good at the sampling point. If so, the solution will work well. Monitor the rate of error frames over time with a logger. If the rate is less than 1/10.000, it is probably good enough. CAN is very forgiving. 
/Lars-Berno 



On 14/07/04 20:26, John Dammeyer wrote: 





I’ve got a client who really doesn’t like prefab cables or the larger diameter Belden DeviceNet thin cable that is the proper 120 Ohms impedance. They’d like to use this Alpha Cable which is rated at 100 Ohms +/- 10. Bit rates are 250kbps, total length of network isn’t an issue. Of course the hardware which has embedded termination is terminated at 120 Ohms. 



http://www.mouser.com/ds/2/14/6459%20OR005-213474.pdf 



I’d like to be able to tell them not to use it but for a small network with minimal noise producing peripherals I’m running out of arguments. 



Any suggestions? 



Thanks 

John Dammeyer 





"ELS! Nothing else works as well for your Lathe" 

Automation Artisans Inc. 

http://www.autoartisans.com/ELS/ 

Ph. 1 250 544 4950 


--
Archives and useful links: http://groups.yahoo.com/group/CANbus
Subscribe and unsubscribe at www.vector.com/canlist/
Report any problems to <[email protected]>