How-to
Set up Micro850E implicit messaging to a Kinetix 5100 servo
Short answer
The implicit connection to a Kinetix 5100 is created in the Micro850E controller's EtherNet/IP module configuration under Ethernet > Modules, not by a UDFB. The drive is added as a natively supported motion slave, the UDFBs then run above the active connection producing commands and decoding status, and the same controllers accept up to eight generic EtherNet/IP slave devices.
On this page
- Does the Micro850E support implicit messaging to a Kinetix 5100?
- Why is there no assembly instruction UDFB?
- How do I add the Kinetix 5100 to the controller?
- What do I carry over from the Logix 5000 instructions?
- How do I know the implicit connection is live before running motion?
- How much traffic does the cyclic connection put on the network?
- When do I use the generic EtherNet/IP device path instead?
- Questions asked
The implicit connection to a Kinetix 5100 is built in the controller's EtherNet/IP module configuration, not by a UDFB, and that is why importing the UDFB library produced no assembly instruction to call: there is not one. On a Micro850E or Micro870E you add the drive under Ethernet > Modules as a natively supported motion slave, let the module configuration open the Class 1 cyclic connection when the project runs, and wire the UDFBs to that connection's data areas afterwards. The same controllers also take up to eight generic EtherNet/IP slave devices for hardware with no native definition. The steps below follow the order that gets the connection live before any motion logic depends on it.
Does the Micro850E support implicit messaging to a Kinetix 5100? #
Yes: the Micro850E and Micro870E run Class 1 cyclic I/O to a Kinetix 5100, and the same controllers also support generic EtherNet/IP implicit messaging with a documented limit of eight slave devices.
The E in the controller name is doing real work here. The earlier Micro850 reached drives over EtherNet/IP with explicit Class 3 messages, which are services you trigger and wait for; the E-series controllers added the Class 1 cyclic connections this setup uses. The updated Micro800 user manual, document 2080-UM002S-EN-E, is where the generic-device support is documented: up to eight generic EtherNet/IP slave devices per controller. The Kinetix 5100 and the PowerFlex 525 sit on the native side of that line, which is exactly why a motion UDFB set exists for them.
Why is there no assembly instruction UDFB? #
Because the cyclic connection belongs to the module configuration: the controller opens it from the configured module when the project runs, and a UDFB only reads or writes the data that connection carries.
A UDFB on these controllers is the same idea as an add-on instruction in Logix 5000: a packaged block of application logic with pins you wire. None of the blocks opens a connection, because implicit messaging is not a service you call once per scan. When the controller goes to run, it reads each configured module and establishes the cyclic exchange defined there: which assembly each side produces, how many bytes it holds, and how often it updates. From that point the input data area refreshes and the output data area is consumed on the configured interval without any ladder execution.
The connection you were looking for is therefore already half-built in the module list, and re-importing the UDFB library changes nothing about it. Save the UDFBs for the layer above: building commands, decoding status, sequencing motion.
How do I add the Kinetix 5100 to the controller? #
Open the controller in the project, add the drive under Ethernet > Modules, and let the native device definition supply the connection data; the PowerFlex 525 is added through the same list, and implicit messaging from a Micro850E to a PowerFlex 525 walks that variant of the same setup.
Work through the steps in this order:
- Confirm the controller is an E-series unit before anything else. The E must be part of the catalog number, because the workflow above it does not exist on the earlier Micro850.
- Add the drive under Ethernet > Modules and give it the network address it answers on. A drive the software lists natively fills in the connection data it knows; a generic device leaves the assembly entries for you.
- Verify the connection data against the Kinetix 5100 documentation even when the definition filled it: input assembly, output assembly, byte count in each direction, configuration data if the drive requires one, and the update interval.
- Map the data areas: drive status into the input area, commands into the output area, matching the configured byte counts exactly. Every field's width and position counts, and one added or missing byte shifts everything after it.
- Wire the UDFB pins to those variables, command pins to output area members and status pins to input area members, following the block interface rather than assuming it matches a Logix add-on instruction pin for pin.
- Download, put the controller in run, and read the module diagnostics before issuing any motion command.
What do I carry over from the Logix 5000 instructions? #
Carry the connection contract, not the screens: the input and output assembly instances, the byte count in each direction, any configuration data, and the update interval are the values that must match.
The Kinetix 5100 manual walks a Logix 5000 workflow: a module profile from the catalog, an add-on instruction on a rung. The Micro800 software has no screen that looks like that, so copying the manual page by page gets you nowhere. What must be identical is the contract behind the screens: which assembly instance the controller consumes from the drive, which one it produces to the drive, the byte count each way, any configuration assembly, and the update interval. Those values correct, the connection opens; one of them wrong, the connection either fails to open or opens carrying the wrong bytes.
That contract is the same object a controller deals in whenever it owns a device it has no profile for, and sizing a Generic Ethernet Module connection for WAGO 750 I/O covers the identical instance-and-byte-count exercise on the Logix side. Read the instance numbers and sizes from the drive's EtherNet/IP connection documentation. Never guess them and never infer them from a working Logix project's tag list: a connection is rejected outright when either endpoint requests a size the other did not offer.
How do I know the implicit connection is live before running motion? #
The module's connection diagnostics must show the connection active with no configuration fault, and the raw input buffer must change when the drive state changes; only after that should motion commands run.
Check from the bottom up:
- Reach the drive over the network first: correct address, correct subnet, link up. A correct UDFB cannot repair an unreachable drive.
- Open the module diagnostics in the programming software and confirm the connection state is active with no configuration error.
- Monitor the raw input area live. Values should move when the drive state changes; a buffer frozen at zero means the connection never opened, the wrong assembly is selected, or the field you expect sits at a different offset.
- Compare the configured input and output lengths against the drive's documented assemblies in both directions.
- Map one known status field and decode it, width and signedness included, before mapping the whole structure.
- Test one output only with the axis uncoupled from the load or the machine guarded: an implicit connection can enable a drive with no operator action between your write and its motion.
- Call the motion UDFB last. If a command still does nothing once cyclic status is credible, inspect its enable chain, its inputs, and the servo's operating state rather than rebuilding the network.
How much traffic does the cyclic connection put on the network? #
Almost nothing: eight devices at 32 bytes per update and a 5 ms cycle produce about 51,200 bytes per second of cyclic traffic, roughly 0.4 percent of a 100 Mbit/s link.
Illustrative worked example, with the byte counts standing in for the numbers you read from your drive's connection documentation:
- Combined cyclic data per update: 24-byte input assembly + 8-byte output assembly = 32 bytes.
- Update rate: 1000 ms/s ÷ 5 ms/update = 200 updates/s.
- Per device: 32 bytes/update × 200 updates/s = 6400 bytes/s.
- Eight generic devices: 6400 bytes/s × 8 = 51,200 bytes/s.
- In bits: 51,200 bytes/s × 8 bits/byte = 409,600 bits/s.
- Share of a 100 Mbit/s link: 409,600 bits/s ÷ 100,000,000 bits/s = 0.0041, about 0.4 percent.
The decision follows from the arithmetic: set the update interval for the motion response you need, because bandwidth will not be what stops you. The eight-device count is the ceiling that actually bites.
When do I use the generic EtherNet/IP device path instead? #
Use it for devices the software has no native definition for: third-party drives, remote I/O adapters, gateways and instruments, with up to eight generic slaves per controller.
| What you are adding | Configuration path | What fills the connection data |
|---|---|---|
| Kinetix 5100 or PowerFlex 525 | Native motion slave under Ethernet > Modules | Device definition; UDFBs run above the connection |
| Third-party drive, remote I/O, gateway | Generic EtherNet/IP device, counted against the eight | Assembly instances and byte counts from the device documentation |
| A ninth generic device | Not supported on one controller | Split across a second controller or aggregate behind a gateway |
Pick one model per device and stay in it. A Kinetix 5100 configured as a generic device while its motion UDFBs are also called is the worst of both: the UDFBs expect the native slave's data layout, and a hand-built assembly will not match it. Configure the drive natively when a definition exists, use the generic path only when it does not, and let the count decide when the project outgrows the controller.
Questions asked
How do I tell an E-series Micro850 controller from the older Micro850?
The E is part of the catalog number on the housing and in the controller properties, not a mode you switch on: an E-series unit reads, for example, 2080-L50E-48QBB. A controller without the E in its number does not gain Class 1 cyclic I/O through firmware; it reaches drives with explicit Class 3 messages instead, and the module-based workflow does not apply to it.
Can one Micro850E control a Kinetix 5100 and a PowerFlex 525 at the same time?
Yes. Each drive is added as its own entry under Ethernet > Modules with its own address and connection data, and the two connections run independently. Both drives belong to the natively supported group that uses the motion UDFBs; the eight-device limit is documented for generic EtherNet/IP slave devices, not for these native drive definitions.
Where do explicit Class 3 messages still fit in this setup?
Use them for one-time and low-rate work: reading drive parameters, clearing faults, reading fault history, or commissioning edits, while the cyclic connection carries the run-time control. An explicit message is a service you trigger and wait for, so it is the wrong tool for position and status that must refresh every few milliseconds.
What are my options when a project needs more than eight generic EtherNet/IP devices?
Split the list. Keep the devices that need tight coordination on the Micro850E and move slower devices to a second controller, or place them behind a gateway that aggregates several devices into one connection. The limit is a connection count in the controller, not a bandwidth ceiling, so slowing the update rate does not create a ninth slot.