Re: Minimum stub length

Hello,

We have two cases, which are commonly mixed, when different cabling approaches are
used:

Stub-length

Stub-length has maximum length, depending on the bit-rate, where the main criteria
comes from the transmission line properties. The end of the stub is unterminated,
causing reflections. The longer stub, the closer to the sample point reflections
appear. So, the shorter stub, the better. The optimum length is zero, but in real
cabling/wiring, we have always non-zero stub length, typically varying some
millimeters into couple of centimeters. If a stub length is some dozens of
centimeters, the cabling/wiring approach should be re-designed.

Device-to-device distance

Device-to-device distance or length of daisy-chaining cables has a minimum value,
as proved by Steve Corrigan. The main thing is line impedance, which shall remain
within the specified tolerances. It is hidden into Steve's analysis, that the
minimum daisy-chaining cable length is relative to the length of the entire network.
As long as the devices are evenly distributed along the network, the net impedance
is approximately constant over the network. I have confirmed this in many customer
cases, having very different dimensions. In some applications, device-to-device
distance may be some millimeters and in one case we had to increase device-to-device
distance from ~0.5m into ~1.5m in order to get rid of the problems.

Summary

One professor, lecturing us years ago introduction into electronics, always used to
emphasize, that "there are not conductors, there is only impedance". That fits well
also for solving CAN physical layer problems.

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 17, FIN-65380 VAASA
+358 (0)50 588 6894
[email protected]
http://www.tke.fi/
http://www.canopen.fi/

----- Original Message -----
From: "YAP" <[email protected]>
To: [email protected]
Sent: Thursday, May 7, 2015 12:33:32 PM
Subject: Re: [CANLIST] Minimum stub length


Yes Steve is dearly missed. He always had time to answer questions from 
everyone. And did so in a very nice way. 

Thanks for your replies. No it is not run on clothes hangers ( :)))) ). At least not yet. The nodes I have sit on a DIN rail like this 

https://groups.google.com/group/vscp/attach/9cab5492e7e16710/006.png?part=0.1.1&authuser=0 or here http://grodansparadis.com/ 


and can be connected together either over a RJ-45 or a pin list. The 
shorted cable that can be used here the better for users of course. 

/Ake 


2015-05-06 17:28 GMT+02:00 Bram Kerkhof < [email protected] > : 






If you daisy chain CAN nodes, you effectively have a very short stub so I don’t see why it wouldn’t work. I think the quality and quantity of the connectors to your nodes will have a far bigger impact on signal integrity than a small length of cable. Unless you’re using rusty clothes hangers for cable (although I do recall someone trying and succeeding in running CAN on clothes hangers somewhere…) 



cheers, 

Bram 



From: [email protected] [mailto: [email protected] ] On Behalf Of YAP 
Sent: woensdag 6 mei 2015 16:40 
To: [email protected] 
Subject: [CANLIST] Minimum stub length 







Hi all, 



I have a system that runs on 125Kbps and I have always heard that one should try to keep stub length at a maximum of 0.3 meters (probably originate from late Steve Corrigan once in a time). But is there a minimum? And how critical is this at 125Kbps. If one goes down to 0,2 meters for instance. Does it mater at all? 

Cheers 



/Ake 







-- 



--- 
Ake Hedman 
Grodans Paradis AB, http://www.grodansparadis.com 


-- 


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