By the end of this lesson, the technician will be able to:
The Drive That 'Failed' Three Times: A conveyor VFD has been replaced twice in four months. Each new drive "fails the same way": runs for weeks, then starts tripping OVERLOAD several times a shift. The third trip-fest lands on your bench, with purchasing asking which brand to try next.
You stop replacing and start reading. The drive's fault log stores the last eight faults — each with a snapshot: output current, frequency, bus voltage, and a run-time stamp. The pattern is eloquent: every overload trip catches the drive at full commanded speed, output current 112–118% of the motor-rated parameter, mid-shift, clustered in the afternoons. Then the parameter audit: the motor rated current parameter is set to 14.0 A. The motor's nameplate, out on the conveyor: FLA 18.2 A.
The story assembles itself. The original commissioning was done with default-guess motor data; every drive since inherited the same wrong number (cloned from a bad backup). The drive has been faithfully protecting a fictional 14-amp motor — tripping its electronic overload exactly as parameterized whenever the real motor pulled its normal afternoon load (warm bearings, heavier product). Three healthy drives were condemned for doing their job with bad facts.
The fix goes through change control: motor parameters corrected to nameplate, as-found/as-left documented, the cloned backup file quarantined and replaced. The trips never return. Drives rarely lie — but they believe whatever they're told. Parameter literacy is the difference between reading a drive's testimony and shooting the witness.
A modern VFD carries hundreds of parameters, but they organize into families a technician can navigate on any brand:
Rated voltage, full-load amps, frequency, base RPM, power, (and for vector modes, data from an auto-tune). Every protection and control calculation downstream assumes these are true. The Field Scenario is what wrong nameplate data does; mis-set rated current skews the electronic overload, mis-set base frequency skews the V/Hz curve (Lesson 7.7), and a vector-mode drive with stale tune data hunts and growls. Rule: motor data comes off the motor's nameplate, physically read, at commissioning and after any motor change.
Acceleration time, deceleration time, min/max frequency, skip frequencies (bands the drive refuses to dwell in, used to jump over mechanical resonances). Diagnostic relevance: accel too short → overcurrent trips on start; decel too short → bus overvoltage on stop (Lesson 7.7's regeneration); max frequency caps "the machine is slow" complaints; skip bands explain "speed jumps past 42 Hz."
Who may start the drive (keypad, terminal/hardwired, network/PLC) and what sets its speed (keypad pot, analog input 4–20 mA/0–10 V, preset speeds, network reference). The majority of "drive won't start" and "drive ignores the PLC" calls are source-configuration questions: a drive switched to keypad/local control during testing and never switched back is a plant classic. Knowing where these parameters live converts a mystery into a two-minute check — display says ready, won't respond to remote start → check control source parameter and the HAND-OFF-AUTO state.
Electronic motor overload (the thermal model built on Family 1's rated current), current limit, stall protection, ground-fault sensitivity, automatic restart behavior. Two field notes: current limit can mask mechanical problems (a drive holding at its current limit accelerates forever and never reaches speed — "conveyor crawls" with no fault logged); and auto-restart settings deserve respect — a drive configured to retry after faults will re-energize machinery without a human decision, which is a safety-relevant configuration owned by engineering, not a convenience knob.
Relay output assignments (run/fault/at-speed signals to the PLC), analog output scaling (speed/current telemetry), network setup (Lesson 7.9). When the PLC "sees" a stopped drive that's running, the assignment of the drive's status relay is the first suspect.
Drives store a rolling history (commonly the last 4–16 faults) and — the diagnostic treasure — at-fault data captures: output current, output frequency, DC bus voltage, sometimes temperature and run time, frozen at the trip instant. A fault name says what limit was crossed; the captured data plus the pattern across multiple events says why. Single fault = an event; the log = a behavior. Reading the log before touching anything is to drives what the panel survey (Lesson 7.2) is to PLCs.
The universal fault vocabulary, each entry read as what the drive measured + the usual physical stories behind it:
The pattern layer on top: repeating fault type at similar at-fault values = systematic cause (parameter, process, installation). Mixed faults clustered in time = look for a common upstream event (power quality, temperature). One-off years apart = event noise. The log turns anecdotes ("it trips sometimes") into data.
| Fault Displayed | Drive Measured | First Parameter Checks | First Physical Checks |
|---|---|---|---|
| Overcurrent | Instantaneous current spike | Accel time; current limit; (vector tune state) | Jam/mechanical bind; megger motor & leads (LOTO + bus discharge first) |
| Overload (thermal) | Sustained % of rated current parameter | Motor FLA vs. nameplate; overload class setting | Real load growth: bearings, product change, alignment (Module 4) |
| Bus overvoltage | DC bus above limit | Decel time; braking option enabled? | Overhauling load; line voltage if steady-state trips |
| Undervoltage | DC bus below limit | Restart/ride-through settings (engineering lane) | Line sags, loose lugs, input phase health, upstream events |
| Ground fault | Current imbalance to ground | GF sensitivity (engineering lane) | Motor insulation, cable damage, moisture |
| Drive overtemp | Heatsink temperature | Carrier frequency setting | Filters, fans, cabinet ventilation, ambient |
| "Won't start, no fault" | — | Control source / reference source; enable & interlock inputs | HOA position, hardwired enable jumper/safety string, PLC command path |
| Runs but wrong/capped speed | — | Max frequency; reference scaling; skip bands; current limit (masking) | Analog reference signal (Lesson 6.3 loop checks); mechanical drag |
Boundaries: keypad/software diagnostics are read-first work. Parameter changes are controlled changes — authorized per site policy, documented as-found/as-left; protection-related and auto-restart parameters are engineering's lane; safety-function parameters (safe-torque-off configurations and kin) are SME/engineering territory, full stop. All power-section verification per Lesson 7.7's discharge discipline.
The at-fault snapshot is the witness statement: 210% current at 0.4 Hz means the motor was commanded to barely creep and instantly drew a violent surge — the electrical picture of a shaft that cannot turn or a winding that is no longer a winding. Answers D and A both try to parameter their way past physics: raising the current limit feeds more energy into a short or a jam (the drive's protection exists precisely for this moment), and no ramp length helps a load that is mechanically seized — the current spike happens at the first volt-seconds, not because acceleration is hasty. Answer C invents spontaneous corruption to avoid the walk to the machine; parameters that ran 4,000 hours did not rot overnight, and the log's clean history argues against configuration causes, not for them. The clincher is the change-in-kind: clean history, then sudden, perfectly repeatable hard fault = something physical changed last night. The professional sequence is evidence (log harvested), isolation (LOTO + bus discharge), mechanics (bar the shaft), then electrical integrity (megger) — and only with the machine cleared would parameter work even be considered. Drives protect; technicians investigate; parameters are never a crowbar.