How to Verify VFD Analog Reference Scaling
When a drive takes its speed command as a 4-20 mA or 0-10 V analog signal from a PLC or controller, the scaling that maps that signal to frequency has to be right, or the motor runs at the wrong speed for every command and nobody knows until the process misbehaves. This page is for the technician commissioning or checking an analog speed reference. It walks the verification in order: confirm the two endpoints map correctly, check the direction and any offset, and prove the drive does the safe thing when the signal is lost.
VFD Analog Reference Scaling in one line: To verify a VFD analog reference scaling, inject the two endpoints of the signal and confirm the commanded frequency at each. Apply the minimum, such as 4 mA or 0 V, and confirm the drive commands the intended minimum frequency, then apply the maximum, such as 20 mA or 10 V, and confirm it commands the maximum frequency. Check a mid-point to confirm the mapping is linear and in the right direction, verify any minimum-speed clamp, and test that the drive responds safely, holding or stopping as configured, when the analog signal is removed or drops out of range.
Confirm the Two Endpoints First
The endpoints define the whole scale, so establish them before anything else. With the drive able to display its commanded frequency, inject the minimum signal the controller will send, typically 4 mA on a current loop or 0 V on a voltage input, and read what frequency the drive commands. Then inject the maximum, 20 mA or 10 V, and read that. These two points fix the line the drive draws between signal and speed, and if either is wrong the entire range is wrong. Use a signal source, a loop calibrator or a precise voltage source, so you know exactly what you injected.
Confirm the signal type matches the wiring and the parameter. A drive analog input can usually be configured for current or voltage, and for the low endpoint to sit at 0 or at a live-zero like 4 mA, so verify the drive is set for the signal the controller actually sends. A 4-20 mA loop wired into an input configured for 0-20 mA reads 4 mA as a small nonzero speed instead of zero, which is a common and confusing scaling error. Match the input mode to the field signal before trusting any endpoint.
Read the endpoints against the process intent, not just the drive's default. The minimum frequency at 4 mA might be zero, or it might be a deliberate minimum-speed floor the process needs, and the maximum at 20 mA should be the intended top speed, which is not always the motor's base frequency. Confirm the two ends map to what the process actually wants, because a drive that scales perfectly from 4-20 mA to 0-60 Hz is still wrong if the application needed 20-50 Hz.
Check Linearity, Direction, and the Minimum-Speed Clamp
With the ends set, prove the middle. Inject a mid-scale signal, such as 12 mA or 5 V, and confirm the drive commands roughly the midpoint frequency between your two endpoints. A clean linear input lands close to the middle; a reading well off suggests a nonlinear scaling parameter, a gain or bias set wrong, or an input that is not the type you assumed. Checking a third point turns two endpoints into a verified line rather than a hopeful assumption that the drive interpolates the way you expect.
Verify the direction, because an inverted scale is easy to configure and dangerous to miss. Some applications intentionally invert the reference so that a low signal means high speed, but far more often that inversion is a mistake. Confirm that increasing the signal increases the speed unless the process specifically wants the reverse. An inverted reference that goes unnoticed sends the motor to full speed on a signal the operator thinks means slow, which is exactly the kind of surprise that damages a process.
Check any minimum-speed clamp and skip-frequency interaction. Many drives clamp the commanded speed to a minimum floor regardless of a lower reference, so a signal below the floor still runs the motor at the minimum, which is correct behavior but must be understood. If the drive also has skip frequencies configured to avoid a resonance, described in the page on VFD skip-frequency avoidance, the commanded speed will jump across those bands, so a reference landing in a skip window will not produce exactly the frequency the linear scale predicts.
Test the Lost-Signal Behavior and Tie It to the Loop
A live-zero signal exists so the drive can tell a real zero command from a broken wire, so prove the drive uses it. On a 4-20 mA reference, a current below the live zero, near 0 mA, means the loop is broken rather than a legitimate low command. Configure and then verify the drive's response to a lost signal: depending on the application it should stop, hold the last speed, or go to a safe preset, and you confirm it by actually removing or shorting the signal and watching what the drive does. A drive that ramps to full speed on a broken wire is a hazard waiting to happen.
Tie the scaling back to the controller that produces the signal. The drive's input scaling has to match the analog output scaling in the PLC or controller, so a full understanding checks both ends: the controller's engineering-unit-to-milliamp mapping and the drive's milliamp-to-frequency mapping must be consistent. If the process runs a closed loop, such as a pressure or flow controller driving the speed, confirm the loop as a whole, since a scaling error anywhere in that chain shows up as a control loop that will not settle, related to the topic in the page on the irrigation pump VFD pressure-control loop.
Document the verified scaling and connect it to how the system is monitored. Record the two endpoints, the direction, the minimum-speed clamp, and the lost-signal action, so the next person does not have to rediscover them. Because the commanded speed and the actual running frequency are both values a monitoring system can trend, a scaling that drifts or a reference that stops matching the running speed becomes visible in the recorded data, giving an early warning that something in the reference chain has changed since commissioning.
Frequently Asked Questions
Why does my VFD run at a small speed when the command is zero?
The usual cause is a mismatch between a 4-20 mA field signal and a drive input configured for 0-20 mA. On a 0-20 mA scale the drive reads the 4 mA live-zero as a small nonzero speed instead of a true stop, so the motor creeps when the controller means zero. Reconfigure the input for 4-20 mA so the live zero maps to minimum speed. It can also be an intentional minimum-speed clamp, so confirm which behavior the parameters describe.
How do I check that increasing the signal increases the speed?
Inject the minimum signal and read the commanded frequency, then inject the maximum and read it again. If the frequency rose with the signal, the direction is correct; if it fell, the reference is inverted. Confirm a mid-scale point lands near the middle to prove the mapping is linear. Some applications deliberately invert the reference, but far more often an inverted scale is a mistake that would send the motor to full speed on what the operator thinks is a slow command.
What should a VFD do if the 4-20 mA reference is lost?
It should do whatever the application defines as safe, which is why the live-zero exists: a current below 4 mA, near zero, means a broken loop rather than a low command. Depending on the process the drive can stop, hold its last speed, or go to a safe preset, and you must both configure that response and verify it by actually removing the signal. A drive that ramps to full speed on a broken wire is a hazard, so never leave the lost-signal behavior untested.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.