RE: reliable CAN connection

"John Dammeyer" <[email protected]> Sat, 4 Jun 2016 22:42:21 -0700
Newsgroups gmane.comp.hardware.bus.can
Message-ID <[email protected]>
Hi Peter,
Way back around 1992 the company I was working for used DB-9 connectors for
their CAN bus connections.  The manufacturer of the equipment had proposed a
pin assignment that had 6 of the pins dedicated to CAN, two for power/gnd
and one for shield.   

The connection in the module generally just connected the CAN_IN to the
CAN_OUT bus and to its' driver.  If the module was removed or not needed on
the bus, a DB-9 connector with the pins jumpered was plugged in its place.
Alas, this was not accepted as the CANopen DB-9 and instead a T type cabled
connector is required.

In either case, odds are your hardware just wires the CAN signals between
the two M12 connectors and the internal PC board.  Or perhaps both go to the
PC board and then are connected by traces and to the modules CAN driver.

So now it becomes a study on the probability of failure of components/solder
joints and how that impacts overall system failure likelihood.

Internally the module will not be configured as a bridge with a receiver and
transmitter like it might be for a DMX-512A lighting system with
uni-directional RS-485.  Not unless it's marketed and sold as a bridge or
repeater.  

CAN is a party line system where all nodes are all on the same bus and
everyone, including the original sender,  receives what everyone sends.  A
node that has a hard electrical failure can take down the bus.  Been there.
Seen that.  Also experience hard mechanical failures in the T connection
where there is one connector into the device from a T connection on the bus.


The first is easy to solve.  Remove each module and in your case bridge it
with a M12 to M12 cable.  If the system is alive again you've found the bad
module.

The second is harder.  We've used a TDR (Time Domain Reflectometer) to
approximate where the bus has a short or open circuit. Then mechanically
inspected the connections near the fault area.

If that doesn't answer your question perhaps you could explain the problem
differently?

John Dammeyer



> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On
> Behalf Of Peter Lauer
> Sent: June-04-16 1:01 PM
> To: [email protected]
> Subject: Re: [CANLIST] reliable CAN connection
> 
> 
> thanks for the quick response.
> I know its possible to build more redundancy into systems. What I wanted
> to see if projects use CAN as a single channel to communicate with only
one
> device, and them have  a CAN channel for each device.
> In our application each device has a CAN-In (M12 male ) and an a CAN-out
> (M12 female), if any of these connections fails, the CANbus is down.
> 
> Peter
> 

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