Comparison
How a PLC controls a VFD: five interfaces compared
Short answer
A PLC controls a VFD through one of five interfaces: discrete I/O, an analog speed reference such as 0-10 V or 4-20 mA, Modbus RTU or TCP, EtherNet/IP or BACnet/IP, or PROFIBUS or DeviceNet. Discrete and analog wiring carry commands only; the network interfaces also return status and fault data. Pick by how much the PLC must know about the drive, not by brand.
On this page
A PLC controls a VFD through one of five interfaces: discrete I/O for start, stop and direction; an analog speed reference, either 0-10 V or 4-20 mA; Modbus RTU or TCP; EtherNet/IP or BACnet/IP; or PROFIBUS or DeviceNet. The deciding factor is how much the PLC must know about the drive. Discrete and analog wiring carry commands only — a speed number and a few contacts — while the network interfaces carry command, setpoint, status and fault data over one link and report whether each command was received. Choose by information need, set the drive's command and reference sources to match the wiring, and configure the signal-loss behavior before the machine runs.
What each interface carries #
Discrete control is two- or three-wire start/stop and forward/reverse. The PLC sees nothing back unless you wire one of the drive's relay outputs to a PLC input. It is the whole answer when the machine only needs on/off and direction.
Analog control sends one number. The PLC scales an engineering value to the output, and the drive maps that range to its minimum and maximum frequency, so a scaling mismatch at either end shows up directly as a speed error. Neither analog signal carries a run command or any status.
Modbus RTU or TCP makes the drive a slave that answers every request, so the PLC knows whether a write was received, and one link carries the setpoint, start/stop, fault reset and monitoring registers. On a single serial port the drives are polled in turn, and that polling cycle sets how fast the whole list updates — read why Modbus RTU drive polling on one serial port is slow before committing several drives to one port.
EtherNet/IP, BACnet/IP, PROFIBUS and DeviceNet carry the same command, setpoint and monitoring set over a network connection instead of a serial bus. The drive confirms receipt and executes, and the PLC reads status and fault data over the same connection. EtherNet/IP fits industrial control, BACnet/IP building automation, and PROFIBUS or DeviceNet the installed industrial and process networks already in the cabinet.
| Interface | What the PLC sends | What comes back | Best fit |
|---|---|---|---|
| Discrete I/O (two- or three-wire) | Start/stop and forward/reverse commands | Nothing, unless drive relay outputs are wired to PLC inputs | On/off and direction control only |
| Analog, voltage or current loop | One scaled speed value | Nothing by itself | Speed control, ramping, modulation |
| Modbus RTU or TCP | Setpoint, start/stop and fault reset by request and response | Acknowledgment of each write, plus status and monitoring registers | Diagnostics on existing serial hardware |
| EtherNet/IP or BACnet/IP | Same command and setpoint set over Ethernet | Status and fault data on the connection | Industrial control; building automation on BACnet/IP |
| PROFIBUS or DeviceNet | Same command and setpoint set over the fieldbus | Status and fault data on the connection | Installed industrial and process networks |
Command source and reference source are separate settings #
Every drive picks where its run command comes from and, separately, where its speed reference comes from: terminals, keypad or network. A PLC can write a flawless setpoint over Modbus and achieve nothing while the drive still takes its reference from the analog terminal. Mixed schemes are common and legitimate — discrete run/stop with an analog reference, or network commands with a local reference — but both selections must match what is actually wired and written. The parameter names differ by manufacturer, so read them in the manual for the exact drive model. When a drive ignores a command, read its displayed command and reference source before touching the PLC code.
Worked example: sending a 30 Hz reference over both analog spans #
Inputs:
- Drive maximum frequency: 60 Hz — the configured maximum frequency for this example drive; substitute the value set in your drive.
- Requested speed: 30 Hz — half of maximum, the process demand in this example.
- Signal spans: 0-10 V and 4-20 mA — the two standard analog reference ranges.
Rule in words: the output signal is the requested speed divided by the maximum frequency, times the span width, plus the zero-speed offset (0 V for voltage, the live zero for current).
Arithmetic:
- Speed fraction: 30 Hz ÷ 60 Hz = 0.5.
- Voltage output: 0.5 × 10 V = 5.0 V; zero-speed offset is 0 V, so the PLC outputs 5.0 V.
- Current output: 0.5 × 16 mA = 8 mA; add the live zero, so the PLC outputs 12 mA.
Both signals command 30 Hz on a 60 Hz drive, and they differ only in what happens when the wire breaks. A cut voltage conductor reads zero at the drive — a valid zero-speed command — so the drive ramps to minimum speed with no fault anywhere in the system. A cut current-loop conductor reads below its live zero and outside the valid span, so the drive can detect reference loss and run its configured response. The decision follows directly: use the current loop whenever the reference leaves the enclosure or runs near power wiring, and keep the voltage reference only for short runs inside one panel where a silent drop to zero speed is acceptable.
What the drive does when the link drops #
Every network or reference loss leaves the drive doing exactly what its communication-loss action says: stop, hold the last reference, or fault. PowerFlex 753 and 755 drives, for example, document configurable EtherNet/IP communication-loss action parameters, and equivalent loss settings exist across drive families for both network and analog signal loss. Decide it during commissioning rather than accepting the default. For machine control, set stop or fault; hold-last-reference belongs only where a stopped process is more hazardous than a drive nobody is commanding.
Commissioning the chosen interface #
- Verify the pair: the drive has the built-in port or option card, the PLC has the matching port or module, and the manufacturer documents the data map for that combination.
- Set the drive's command source and reference source to match the chosen interface before any wiring test.
- Wire and configure both ends: analog voltage or current type and range matched, discrete commons checked for sink or source compatibility, and fieldbus address, baud rate or IP address plus data format set identically. Route analog cable away from motor leads and use shielded cable for the reference.
- Program the PLC: permissives, run request, a scaled setpoint clamped to the drive's frequency limits, fault reset, and a comm-fault response.
- Check the analog chain with a meter at the PLC output terminals and again at the drive input terminals; the two readings must agree, or the fault is in the wiring, not the scaling. Then compare the drive's displayed frequency against the PLC value at zero, mid-scale and maximum.
- Test the failure paths: pull the run signal, the analog conductor and the network cable, and confirm the drive follows its configured loss action each time.
Keep the motor decoupled or the load otherwise safe for steps 5 and 6; the loss-action test is a stop path, and it only counts if you have watched it work. For a full single-drive sequence over serial, the steps in controlling an ATV12 drive from a TM200 PLC over Modbus RTU walk the same order on real hardware.
Which interface for which machine #
- The PLC only starts, stops and reverses, and a relay contact back is enough feedback: discrete two- or three-wire control.
- The PLC sets one speed and nothing else, and the wiring stays inside the panel: the voltage reference. If the reference leaves the enclosure: the current loop, for the broken-wire detection in the worked example.
- The PLC needs fault codes, run status or more than two drives, and it speaks EtherNet/IP natively: put the drives on EtherNet/IP. The procedure to start and stop a PowerFlex 700 over EtherNet/IP from a CompactLogix shows the mapping work that involves.
- Serial ports are what the hardware gives you: Modbus RTU multi-drop, after checking the polling cycle against your update requirement.
- PROFIBUS or DeviceNet scanners already sit in the cabinet: keep them. Both carry command, status and diagnostics, and replacing working fieldbus with Ethernet pays only when node counts or spare availability force it.
- The controller front end is a building automation system: BACnet/IP.
Questions asked
Can one drive use discrete start/stop with a network speed reference?
Yes. A drive takes its run command and its speed reference from separately selected sources, so discrete run plus a Modbus setpoint is a valid mixed scheme. The rule is that both source selections must match what is actually wired and written. If the reference source is still on the analog input, network writes to the setpoint register are accepted but never used.
Which interface uses the least PLC I/O?
Modbus RTU on the drive serial ports: one PLC port or gateway addresses every drive on the multi-drop cable, against one analog output plus two discrete outputs per drive for hardwired control. The trade is update rate, because every drive is polled in turn, so each additional drive slows the whole polling cycle. Run that numbers check before choosing serial for a long drive list.
Is the current loop always the right analog choice over the voltage reference?
Use it whenever the reference leaves the enclosure or runs near power wiring. The live zero is the reason: a severed current-loop wire reads about zero, below the valid span, so the drive can flag reference loss, while a severed voltage-reference wire reads zero volts, a perfectly valid zero-speed command. Keep the voltage reference for short runs inside one panel where that failure mode is acceptable.
What happens to a running drive when the PLC drops off the network?
Exactly what its communication-loss action is set to do: stop, hold the last reference, or fault. PowerFlex 753 and 755 drives, for example, expose the EtherNet/IP communication-loss action as configurable parameters. Set it deliberately during commissioning and prove it by pulling the cable; an unset action leaves the drive running at its last reference with no controller watching.
Do the PLC and the drive have to be the same brand?
No. Any PLC and drive that share a documented protocol and data map work together, which is the point of Modbus, EtherNet/IP, PROFIBUS and DeviceNet being open specifications. Same-brand pairs usually mean a ready-made data map and a single toolchain, which shortens commissioning, but the actual requirement is documentation for the exact model pair, not a matching logo.