RS Networks for DeviceNet

From daba's post #15:

If and when you add any new nodes to the scanlist, you just have to remember to turn off that "Automap on add" option, as it is on by default. I can't find a registry value that configures it off by default.

Do not forget this!!
 
@ daba so once i went into the module i unchecked the "Automap on Add" but before i moved all the devices/nodes into the scan list i went into the IO tabs, their i selected options tabs and i specified the data alignment as "DWord Align" by default it was set for "Byte Align". Once i did that and populated the mapping all devices take up an even amount of space, and although their are devices that don't need that much space in memory at least its well organized in proper order to easy data manipulation. I was worried that 8btyes of space wouldn't be enough for some of the devices but it worked out great. Thanks on your help.
 
The "options" that i referred to will not be applied if you turn off "Automap on add". So it is irrelevant what you choose here if you are going to map the devices manually.

With "Automap" turned off, you have the freedom to place the data where you want it.

Mapping at 2 x node number was good enough for GM, it is their world-wide standard, and has to be strictly adhered to by all of their system integrators. But they are not unique, I have seen this mapping implemented in a few places.
 
Last edited:
Guys,

Just wanted to let you know the commissioning was a success, i tell you the hard part was the configuration and making sure all devices were on the same baud, terminating resistors, power supply type etc, but once it worked that was it.

I have another question related to this. I want to eventually tie in the current devices that i have configured to another MCC that has its own device net rated power supply running its own set of devices over device net. The goal is to have both MCC's one one device net network. I have read some where that you can run two device net rated power supply's for one device net network i just can't remember the requirements to do that, possibly grounding only one power supply?? So let me know. Thanks.
 
To tie two powered DeviceNet networks together, just don't connect the positive wire.

Connect the negative wire, and connect the grounds. Connect the white (CAN_H) and blue (CAN_L) signal wires.

MCC #1 gets powered by Power Supply #1, MCC #2 gets powered by Power Supply #2, but all the CAN_H and CAN_L signals are referenced to the same DC Common.
 
Ken,

I currently have a device net cable run from MCC#1 and not terminated in MCC#2, the cable plugs into a Device net terminal block there are slots for these on the MCC think of them as ports you can plug into. The device net cable is a 4 wire multiconductor cable Red is your positive wire, black is your negative, White is your CAN_H and Blue is your CAN_L. So if i connected this cable thats coming from MCC1 and plug it into MCC2 without plugging in the positive wire should it work? what ground wires are you referring to?

**EDIT
Technically speaking i still don't know how not wiring the positive wire makes all the CAN_H and CAN_L reference the same common, if you can elaborate on that i would appreciate it.
 
Last edited:
Connecting the DC Common (the black wire) makes the CAN_H and CAN_L signals both relative to the same point.

It's a very common misconception in DeviceNet and in RS-485 that the signal is "differential" between the two signal wires only. It is not. Each signal voltage is measured by the transceiver chip relative to DC Common, and the difference between those signals is the "1" or "0" on the data bus.

Not connecting the +24V DC (the red wire) makes each MCC's DeviceNet draw power only from its own power supply. If you connected them, they might work in parallel or they might not... it would depend on the power supply design and features.

I have always powered each MCC network from its own supply, and put a big tag on the DeviceNet bus cable coming from the other segment that indicates not to connect the red wire.

Edit: This summer at a family picnic I was carrying my little niece around on my shoulders when she looked down at her brother, stuck her tongue out, and said "I'm taller than you !". I put her down and explained about elevation datum lines.
 
Good advice from Ken...

Many people think they are putting 24V on all the devices on a network from a single power supply, but you have to consider the volt drop along the network, caused by resistances in the cable and the connectors - all are widely documented.

You have to make sure that all devices receive the minimum spec. voltage on the "red wire", so it's no good putting a second power supply on the network at the same location as the first, the volt drop will still ensure the distant devices may not receive the minimum voltage (19V, or thereabouts).

Most modern power supplies will happily "share" on a network, but the spec. for DeviceNet calls for segregation of the positive supply to the network when using multiple supplies. Do it by the book, and don't cut any corners, and you will have no problems.

There are network connectors that provide LED indicators of the supply volts at that point in the network, which is a great diagnostic tool, but is it worth the extra cost? Properly designed and installed networks have run perfectly, and unattended, for many many years.

Proper design and install are the key issues.
 
thanks on the info daba,

See based on the device net manuals that i went through it it mentions like we already discussed to not wire the positive, also that one power supply the one in the middle of the network be grounded. Also the shielding cable that is in the multiconductor cable used for device net network needs to also be grounded wherever the power supply is grounded.

I may have issues because the current wiring that the customer has is only a 4 wire multiconductor cable in that their is no shielding wire which is technically the 'fifth' wire.

If i do experience issues with this i think just running everything from one power supply should work, and hopefully the voltage drop won't be an issue and all the devices will recieve their required voltage minimum. Ill see what happens next week.
 
Do it by the book, and don't cut any corners, and you will have no problems.

I spent ten years hammering that point home to DeviceNet users. The ones who listened, I didn't know well because they only had to call once.
 
So if i connected this cable thats coming from MCC1 and plug it into MCC2 without plugging in the positive wire should it work?

I would like to add that you should be careful about where you connect this cable into each of the two MCCs. You want to pick up one end of each MCC cable run and connect those together. You want to connect the two main trunk lines together so you have one long trunk. Connecting somewhere in the middle of the trunk will create problems. Terminating resistors at these ends will need to be removed also.
 

Similar Topics

Hello Team, Is there a way to explore DH+ Networks like there is for EIP and Dnet? Thanks! -PreLC
Replies
3
Views
1,718
Hello all, I have a problem to set up correct configuration with Schneider Altivar 340 in RSNetWorks for DeviceNet, CPU is CompactLogix L33ER...
Replies
2
Views
1,900
We have 11 (AB)Armor Start 284D PF4 drives on DeviceNet. 1. If I open up RSNetWorks and go line---do the network upload & save? 2. I will have...
Replies
11
Views
3,415
Hi all, I am new in Devicenet and AB's VFD,Now i am iam in a trouble with a project,One of our existing VFD in devicenet network has a problem,it...
Replies
5
Views
4,553
Hi Guys, I want to connect my 1769-L36ERM to a different network domain, is it possible to do it using only the dual ports in the Controller? Or...
Replies
3
Views
226
Back
Top Bottom