How-to
How to map a 1747-SDN scanner for DeviceNet drive control
Short answer
Seeing the drive in an RSNetWorx browse is not an I/O connection. Put the drive in the 1747-SDN scanlist, read the M-file words the scanner mapping assigns, and reference those mapped words in the SLC 5/04 ladder. A node address is not an integer-file offset: the mapping view, not node 10, decides where each command and status word lives.
On this page
- Why does RSNetWorx see the drive but the PLC cannot run it?
- Is there an integer-file offset for a drive on node 10?
- What is the difference between the scanlist and the M-file mapping?
- How do I add the PowerFlex 70 to the scanlist in RSNetWorx?
- Which M-file words does the scanner map to node 10?
- How do I reference those mapped words in ladder logic?
- How do I prove the command path before starting the motor?
- Why will the drive still not run when the mapping is right?
- Questions asked
You can browse a PowerFlex 70 at node 10 in RSNetWorx for weeks while the motor never twitches, because browsing is device discovery, not an I/O connection. An SLC 5/04 talks to DeviceNet only through the 1747-SDN scanner's M-files: RSNetWorx decides which slaves sit in the scanner's scanlist and which M-file words carry each one's command and status data. So the work runs in a fixed order: add the drive to the scanner's scanlist and download it, read the M-file mapping RSNetWorx assigns, then write the ladder against those mapped words. There is no integer-file offset for node 10 — the mapping, not the node number, picks the addresses. This article gives you that order, the placement arithmetic, and the checks that prove the path before the motor turns.
Why does RSNetWorx see the drive but the PLC cannot run it? #
A browse result only proves the node answers DeviceNet identification requests; motor control needs three more things — the drive in the scanner's scanlist, an active polled I/O connection to it, and ladder logic reading and writing the exact M-file words the mapping assigns.
RSNetWorx for DeviceNet does two different jobs, and a browse touches only the first: it configures slave devices such as the drive's 20-COMM-D adapter, and it configures the 1747-SDN as the master scanner. Discovery proves the node answers on the wire. Control requires the scanner to hold an active polled I/O connection to that node, and that connection exists only when the drive sits in the scanner's scanlist with the configuration downloaded.
Keep the layers separate while you diagnose:
| What you observe | What it proves | What to check next |
|---|---|---|
| RSNetWorx lists the drive at node 10 | The node answers on the wire | The scanner's scanlist |
| The scanlist is downloaded with the drive in it | The scanner will exchange I/O with the drive | The mapping view's word assignments |
| Mapped M1 words change with the drive's state | The inbound path works | Command words in M0, then drive-side setup |
| Mapped M0 words are written but the drive idles | Outbound path or drive control setup | Adapter polled parameters and control source |
No single observation counts as commissioning. The chain — SLC ladder to scanner M-files, scanlist and mapping to DeviceNet connection, connection to drive, drive response — fails one link at a time, so test them in that order.
Is there an integer-file offset for a drive on node 10? #
No — the node address is a DeviceNet identity, not a data-table address, and the 1747-SDN never places slave data into your integer file on its own: RSNetWorx assigns each scanned device its M-file words, and those assignments are the only addresses your ladder can trust.
The scanner's user manual states it directly: the module exchanges slave data through its M-files rather than dropping anything into discrete or integer storage by itself, and the scanlist decides which devices participate. Two machines can both carry a PowerFlex 70 on node 10 with completely different mapped words, because placement follows scanlist order and connection size, not the address.
Worked placement example — the words below are illustrative; your mapping view shows the real ones. The scanlist holds a node 4 drive mapped first with a 2-word input connection, so it occupies M1 words 2 through 3. The PowerFlex 70 on node 10 is mapped second with a 4-word input connection: its first input word lands at 2 + 2 = 4 and its last at 4 + 4 - 1 = 7, so the inputs sit in M1 words 4 to 7 and the 4 command words sit at the same positions in M0. Reorder the scanlist so node 10 is mapped first and the same node address lands on words 2 to 5. Same node, different words — that arithmetic is exactly why no offset formula exists, and why the saved RSNetWorx configuration file is the only source of truth for addresses.
What is the difference between the scanlist and the M-file mapping? #
The scanlist is the scanner's subscription list — only devices added to it get I/O exchanges — while the mapping is the word-level address book that says which M-file words carry each device's inputs and outputs; both live in the scanner, and RSNetWorx edits both.
The scanlist answers which devices participate; the mapping answers where their data lives. On the SLC side the directions are fixed by the module design: the processor writes the scanner's M0 words, which the scanner transmits to the drives, and reads the scanner's M1 words, which hold what the scanner received from the drives. Getting the direction backwards produces healthy-looking logic that talks to itself, so record direction next to every word number.
Word counts are not arbitrary either. Each device's count follows the I/O connection its adapter is configured for, and on the drive side that polled connection size is set in the 20-COMM-D adapter's configuration — change it there and the mapping moves, along with every rung that references it.
How do I add the PowerFlex 70 to the scanlist in RSNetWorx? #
RSNetWorx edits the scanlist inside the 1747-SDN's properties, so the whole job is: open a configuration, go online to the network, select the scanner, add the node 10 drive to the scanlist, and download it.
- Start RSNetWorx for DeviceNet and create a configuration file for the network; that file becomes the permanent record of your scanlist and mapping.
- Go online and browse. Confirm the PowerFlex 70 answers at node 10 and the 1747-SDN appears as a scanner. This step is discovery only — it establishes nothing about I/O.
- Select the scanner in the configuration view and choose Device > Properties to open its configuration.
- Edit the scanlist and add the drive's node 10 entry. The click-level procedure for adding a DeviceNet drive node to an SLC 500 covers this same edit in detail.
- Note the polled connection size RSNetWorx shows for the drive; that word count comes from the adapter's polled configuration and is what the mapping will consume.
- Download the scanlist to the scanner and save the configuration file. The saved file holds the word assignments you read next, so treat it as the source of truth for every ladder reference.
Which M-file words does the scanner map to node 10? #
The scanner's mapping view is the authority — it lists every scanned node with its input and output word blocks in the M1 and M0 files, so record the start word, the word count, and the direction for the drive before any ladder work begins.
Open the same scanner properties you used for the scanlist and read the mapping. For the illustrative layout above that is M0 words 4 to 7 outbound and M1 words 4 to 7 inbound; on your machine the numbers are whatever the view says. Each mapped word is one 16-bit SLC data-table word, and the count follows the adapter's polled connection, so a polled change on the 20-COMM-D moves the mapping with it.
Write down four things before closing RSNetWorx: the scanner's slot, the mapped M0 start word and count for commands, the mapped M1 start word and count for feedback, and the direction of each. With those four lines on paper, the next person re-runs this discovery in thirty seconds instead of starting cold.
How do I reference those mapped words in ladder logic? #
Stage the drive's command words in integer storage, copy them into the mapped M0 words every scan, read drive status from the mapped M1 words, and gate the start command on the drive-ready and no-fault bits that status carries — always against the mapping you recorded, never from assumptions.
This answers the referencing question directly: the mapped words are the source and destination of ordinary move and copy rungs. Copy the mapped M1 block into your N-file status words for the HMI and alarms, build the command words in N-file staging words, and copy them out to the mapped M0 block each scan. Keep the direction in the rung comments (for example, node 10 inputs) so nobody later swaps command and feedback.
Decode the command and status layout from the adapter manual for the polled connection you selected: which bit is run, which is fault, how the speed reference is scaled, and what range it accepts. Gate the start on drive-ready true and fault clear, hold the reference at zero until checkout, and keep machine permissives in their own rungs instead of burying them in the transport copy. The same command-out, status-in discipline applies over EtherNet/IP, where starting and stopping a PowerFlex 700 over EtherNet/IP from a CompactLogix does the job with controller tags instead of M-files.
How do I prove the command path before starting the motor? #
Prove the data path with the machine safe to move and the reference held at zero: watch the mapped M1 words change when you stop the drive from its keypad, then watch the drive respond to writes on the mapped M0 words, and only then try a slow reference.
Before trusting any mapped data, check the scanner itself: its own node address and health sit in M-file word 210, and decoding the 1747-SDN node address and status from M-file word 210 breaks that word down bit by bit.
Then run the sequence with the load disconnected. Stop the drive from its keypad and watch the mapped status word change — inbound path proven. Write a command word of zeros and confirm the drive sits fault-free — the outbound words are not landing somewhere destructive. Write a small nonzero reference with the start bits, watch the drive take the network reference, and keep everyone clear of the driven machine: once the path is proven, the shaft turns the moment that reference lands.
Decision: mapped M1 words that never move at the keypad mean the input mapping or the scanlist download is wrong, so return to the mapping view. M1 words that move while the drive ignores M0 writes put the fault at the drive side, which the next section covers.
Why will the drive still not run when the mapping is right? #
When mapped words move and the drive still idles, work the remaining four links in order: the adapter's polled I/O parameters, the drive's control-source selection, the command bits the adapter passes through, and the health of the node itself on the wire.
A drive that runs from its keypad but not from the network is usually still owned by its local control source, and the drive manual covers switching control so DeviceNet commands are accepted. If the mapped command word arrives but particular functions stay dead, the adapter's polled parameters define which bits and words pass through to the drive — check them against the connection you selected in the scanlist. And when one misbehaving device drags the whole network down, temporarily disabling DeviceNet scanlist nodes on 1747-SDN and 1769-SDN scanners shows how to cut a single node out without rebuilding the scanlist while you replace it.
Final decisions. Mapping unread: stop, because no rung you write can be right until it is. Mapping read, words moving both ways, drive accepting network commands: commissioned — guard the rungs with permissives and store the mapping file with the program. Words moving one way only: the silent direction points at the scanlist download first and the drive-side control source second, and checking them in that order settles the fault.
Questions asked
Can I download a changed scanlist to the 1747-SDN while the SLC is running?
Yes. The scanner installation guide, publication 1747-IN058C-EN-P, documents downloading scanlist changes to the scanner while both the SLC processor and the scanner are in RUN mode. Save the RSNetWorx configuration file before the download, expect the edited node's I/O connection to re-establish as the new scanlist takes effect, and re-read the mapping view afterwards, because adding or resizing a device can move other devices' word assignments.
Do I need an EDS file for the PowerFlex 70 before it can join the scanlist?
RSNetWorx uses each device's EDS file to show its parameters and I/O connection choices. The EDS for a 20-COMM-D adapter comes from the drive manufacturer's own support or literature site, not from third-party file collections; register it in RSNetWorx and the drive appears with its connection options when you edit the scanlist. If the drive already shows in the browse with its proper identity, its EDS is already registered on that workstation.
Which manuals cover this scanner, drive, and adapter chain?
Three publications: 1747-IN058C-EN-P is the scanner installation guide, 20A-UM001J-EN-P is the PowerFlex 70 user manual, and 20COMM-UM002C-EN-P covers the 20-COMM-D DeviceNet adapter, including its node address setting and polled I/O layout. All three are on the manufacturer's literature site. Read the scanner's mapping chapter and the adapter's polled I/O chapter before you write a single rung against M-file words.
Does the same method apply to a 1769-SDN scanner on CompactLogix?
The scanlist work in RSNetWorx is the same, but the addressing is not: a 1769-SDN on a CompactLogix or MicroLogix exchanges its slave data through the controller's own I/O image instead of M-files, so rungs written against 1747-SDN M-file words do not carry over. Re-derive every address from the new module's configuration; the placement principle — scanlist order and connection size decide the words, never the node number — stays identical.
What baud rate and cable length limits does DeviceNet give me?
All nodes must run the same rate: 125, 250, or 500 kbit/s. Maximum trunk length drops as the rate rises: about 500 m at 125 kbit/s, 250 m at 250 kbit/s, and 100 m at 500 kbit/s. Set the scanner and the 20-COMM-D adapter to the same rate; 500 kbit/s suits a panel-local network, while a machine spanning a building bay leaves more length margin at 125 or 250 kbit/s.