CAN port question
-
I have read the various dozuki write-ups but I am still unclear on what the lines in the CAN cable are for. Specifically, why are there CAN0 and CAN1 lines. Are there two separate CAN buses on one RJ11 connector. Why are there 4 lines going from the tool distribution board to the tool boards. Am I opening myself up to more possibility of interference by only using two wires rather than 4 wires?
-
There is a type of CAN bus setup that in addition to the two wires needed to carry the data there are two additional wires that power a terminating bias circuit.
-
@jens55 the MB6HC has two CAN ports. Current firmware uses just one of them, but in firmware 3.4 (under development) we have added support for plain CAN (not FD) on the second port, so that it can be used to communicate with certain devices to configure them e.g. the ODrive.
CAN is a linear bus system, so apart from the devices at the ends (one of which is the Duet main board) each device needs CAN wires for the previous device and CAN wires to the next device. You can have short stubs attached to the CAN bus. The maximum recommended stub length for the 1Mbps signalling rates used by Duet is 300mm.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42. thanks for chiming in. I am still confused about why there are 4 lines in the CAN FD connection but we only use two of these. Your comment on the 300 mm stub brings up another question as well. Long range plans call for a 500mm x 500mm Jubilee build but what I hear you saying now is that I can't run a tool distribution board since the stub length would probably be in the 600mm range .... am I understanding this correctly? If this is so, does that mean I can't run tool boards since they appear to only have a CAN input but no way to daisy chain it to the next tool board ?
-
@jens55 the tool distribution board provides two pairs of wires to and from each tool board, so there are no stubs.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42, ah, so in effect the tool boards are daisy chained then ....
So does that mean I need to run 2 pairs of wires from the tool distribution board to the tool boards but I only need one pair from the 6HC board to the tool distribution board?
This would explain the 4 wires to the tool board .... Do I understand correctly then that ver 3.4 (or later) will have one CAN-FD bus and one standard CAN bus on each connector and that is why there are 4 lines on the RJ11 connectors ?Sorry to be slow on the uptake ....
-
@jens55 said in CAN port question:
@dc42, ah, so in effect the tool boards are daisy chained then ....
So does that mean I need to run 2 pairs of wires from the tool distribution board to the tool boards but I only need one pair from the 6HC board to the tool distribution board?Yes. In practice I think you would be unlikely to have issues if you connected the tool boards as stubs using a single CAN twisted pair.
This would explain the 4 wires to the tool board .... Do I understand correctly then that ver 3.4 (or later) will have one CAN-FD bus and one standard CAN bus on each connector and that is why there are 4 lines on the RJ11 connectors ?
No, it's just a software change, to enable the second CAN bus that the MB6HC already supports in hardware.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42 said in CAN port question:
@jens55 said in CAN port question:
Yes. In practice I think you would be unlikely to have issues if you connected the tool boards as stubs using a single CAN twisted pair.
But if the tool distribution board daisy chains the lines to the tool boards then I can't possibly get away with a single pair of wires going to the tool board because the daisy chain would end at the tool board since nothing is returned to the tool distribution board to carry on the daisy chain.
No, it's just a software change, to enable the second CAN bus that the MB6HC already supports in hardware.
That second CAN bus you are talking about is on the same RJ11 connector as the first CAN bus ... correct ?
-
OK, I think I got the tool board thing figured out. When you leave the jumpers on the tool distribution board as if the particular port wasn't used and if you then connect a single pair of wires from the tool distribution board to the tool board, you in effect put in a stub that isn't in the daisy chain.
Is the 300 mm stub length limit the limit of a single stub or all stubs added together? -
@jens55 said in CAN port question:
OK, I think I got the tool board thing figured out. When you leave the jumpers on the tool distribution board as if the particular port wasn't used and if you then connect a single pair of wires from the tool distribution board to the tool board, you in effect put in a stub that isn't in the daisy chain.
Correct.
Is the 300 mm stub length limit the limit of a single stub or all stubs added together?
It's per stub. There is also a total limit on the length of all stubs, which is about five times the limit per stub.
In practice, if the toral bus length s no more than 5m then stubs 1m long should not be a problem (with the total length of all stube being no more than 5m).
All this will change if in future we need to enable bit rate switching in order to improve throughput.
