Re: can network establishing -- Can bit timing

And to add some more realism,

we have seen a 500k network with 10-20 nodes, some 20-20 connections 
(cable connectors) and bad wiring in PLC enclosures to fail occasionally 
around 50m.

As Heikki mentioned, isolations can cause additonal problems: A 1M network 
with app. 10 nodes and lenght of 30-40 failed due to extended ACK caused 
by slow galvanic isolation on one component.

isto




heikki saha




From:   heikki saha <[email protected]>
To:     [email protected]
Date:   20.10.2013 22:58
Subject:        Re: [CANLIST] can network establishing -- Can bit timing
Sent by:        [email protected]




Hello John,

Thanks. Main reason for nicely operating longer network can be found from 
transceiver
datasheets. E.g. for TJA1050, worst-case propagation delays are:
- 110ns for TX to network
- 155ns for network to RCX

=> total transceiver loop-delay is 265ns, not 255 as I accidentally wrote 
earlier.

But nominal values are:
- 60 ns for TX to network
- 95 ns for network to RX

Those values result max. 110ns difference, which equals 11m network! So, 
if we are very
lucky, our devices behave close to the nominal values or faster and if we 
are unlucky,
they are close to the worst-case.

Best regards,

-H


TK Engineering Oy
Heikki Saha, CTO
  M.Sc. Automation
  Dr.Tech. Electronics
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: "John Dammeyer" <[email protected]>
To: [email protected]
Sent: Friday, October 18, 2013 6:18:16 PM
Subject: RE: [CANLIST] can network establishing -- Can bit timing

Excellent description.  I'll add that from practical experience that 
although you can run as much as 35m at 1Mbps and have it work, it doesn't 
take much to disrupt the messaging when you are that close to the edge. 

John Dammeyer

> -----Original Message-----
> From: [email protected] [mailto:canlist-
> [email protected]] On Behalf Of heikki saha
> Sent: October-18-13 3:49 AM
> To: [email protected]
> Subject: Re: [CANLIST] can network establishing -- Can bit timing
> 
> Hello Eric,
> 
> Propagation delay consists of:
> - Propagation delays of transceiver: TX-to-CAN _and_ CAN-to-RX
> - Propagation delay of logic, if additional logic exist between 
controller and
> transceiver
> - Propagation delays of isolation, if such is used
> - Signal propagation delay required for signal to go from end to end of 
the
> network _and_
>   back. Propagation delay needs to be taken as twice, because during
> arbitration and ACK,
>   network state may be driven by multiple devices and there shall be 
enough
> time to enable
>   network state to stabilize before taking a sample.
> 
> Some example values:
> - TJA1050: TX-to-CAN worst case 110ns, CAN-to-RX worst case 155ns
> - Logic depends on the implementation, but fast CPLDs such as Lattice 
ispM4A
> provide <10ns
> - Best option is to use digital isolators, e.g. for ADuM1100 worst-case 
18ns
> - Signal propagation delay is 5ns/m in standard 120ohm CAN-cable
> 
> => Basic implementation without isolation and logic, we get:
>    1: At 1Mbps CANopen we have 4TQ=500ns reserved for propagation delay 
in
> a worst case
>    2: 110ns + 155ns = 255ns for transceiver -- 195ns is left for signal
> propagation
>    3: We can achieve maximum cable length of 195ns/(2*5ns/m)=19.5m
> 
>    4: If we have the best possible 5TQ=625ns propagation segment
>    5: Maximum cable length of 320ns/(2*5ns/m)=32m may be achieved
> 
> => The basic timing is only one part of the truth. Additional 
constraints are
> required for
>    the topology to keep the transmission line solid:
>    - Few group of devices far away from each other should be avoided, 
because
> the groups
>      modify the line-impedance in the area of the group => impedance 
junctions
> introduce
>      reflections
>    - In addition to previous, residual error probability decreases 
significantly
> when the
>      number of devices are increased -- use numerous devices, but try to 
install
> them evenly
>      along the network.
>    - Too long drop-lines are unterminated ends, causing reflections => 
the longer
> drop-line,
>      the closer to sample point the reflection can appear. The more 
there are
> long drop-lines,
>      the higher power reflections appear
> 
> That was timing in brief. I hope you got some help.
> 
> Best regards,
> 
> -H
> 
> 
> TK Engineering Oy
> Heikki Saha, CTO
>   M.Sc. Automation
>   Dr.Tech. Electronics
> Mail address:  P.O. box 810, FIN-65101 VAASA Visit address: Yritt j 
nkatu 15,
> FIN-65380 VAASA
> +358 (0)50 588 6894
> [email protected]
> http://www.tke.fi/
> http://www.canopen.fi/
> 
> ----- Original Message -----
> From: "Eric clawson" <[email protected]>
> To: [email protected]
> Sent: Friday, October 18, 2013 1:21:57 PM
> Subject: Re: [CANLIST] can network establishing -- Can bit timing
> 
> 
> Thanks Bertil,
> 
> 
> For this nice calculator. But i need to estimate exact propagation 
segment
> delay.
> 
> 
> 1> But my question is related to estimating the length of the cable to
> 1> find exact propagation delay . Theoretically length of the cable will
> 1> be : drop down length(combined for all nodes) + actual length
> 
> 
> 2> Mostly delay will also because of drop down connectors & also trans-
> receivers .
> 
> 
> 3> So when estimating total length of network then it will be :
> drop down length(combined for all nodes) + actual length ?
> 
> 
> 
> 4> Also how much maximum can be drop down length(combined for all nodes) 
?
> 
> 
> Any suggestion..!!
> 
> 
> 
> 
> 
> On Fri, Oct 18, 2013 at 11:32 AM, bertil b ck < [email protected] > 
wrote:
> 
> 
> Hi,
> have a look here:
> http://www.bittiming.can-wiki.info/
> 
> For Freescale msCAN (HCS12) use SJA1000
> 
> Br,
> Bertil B ck
> R&D Manager, Hardware
> TK Engineering
> 
> 
> 
> ----- Original Message -----
> From: "Eric clawson" < [email protected] >
> To: [email protected]
> Sent: Thursday, October 17, 2013 8:18:40 PM
> Subject: [CANLIST] can network establishing -- Can bit timing
> 
> 
> 
> Hi,
> 
> I am working on a CAN network where i have 20 nodes attached to my 
network.
> 
> Speed = 125 Kbps
> Clock = 8 MHZ (I can increase it further if required) Nodes = 20 Cable = 
120
> ohm characteristics impedance.
> Drop Length per node = 2 meter
> Total length of cable network = 80 Meter
> 
> 1> Now how can i set the Bit timing for my CAN network so that with this
> topology it works at 87.5% sample point ?
> One experience person told me at 125 KBPS we can cover :--
> 
> * 500m length total.
> * Double the length of drop down cable.
> * Maximum length of drop should be 120M
> * Also for every connector consider 6 meter loss in length.
> 
> 
> 
> 2> Please suggest is above calculation right?
> 3> If not then what eactly should i keep the value of propagaton segment 
of my
> network?
> 4> Also what value to be kept for phase segment 1 & 2 ?
> 
> I am also refering an1798.pdf for it. But i need some to share his 
experience in
> setting CAN network .
> 
> //Eric
> --
> Archives and useful links: http://groups.yahoo.com/group/CANbus
> Subscribe and unsubscribe at www.vector.com/canlist/
> Report any problems to < [email protected] >
> 
> --
> Archives and useful links: http://groups.yahoo.com/group/CANbus
> Subscribe and unsubscribe at www.vector.com/canlist/
> Report any problems to <[email protected]>

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




This e-mail is confidential and it is intended only for the addressees. Any review, dissemination, distribution, or copying of this message by persons or entities other than the intended recipient is prohibited. If you have received this e-mail in error, kindly notify us immediately by telephone or e-mail and delete the message from your system. The sender does not accept liability for any errors or omissions in the contents of this message which may arise as a result of the e-mail transmission.