Power Quality Analyzer Not Recording Events? Check These Settings

The machine stopped twice overnight. Its control log has timestamps. The power quality analyzer, collected the next morning, has an empty event list. That result can mean the supply stayed within the configured event limits. It can also mean the instrument was watching the wrong quantity, at the wrong point, with the wrong settings.

Before extending the recording for another week, examine the measurement setup and the saved data. “No events recorded” becomes useful evidence only when the test could have captured the event being investigated.

Translate the complaint into a measurement target

“Power problem” is too broad a trigger specification. Start with the affected equipment, its alarm, the observed behavior and the time of occurrence. A contactor dropping out, a drive reporting undervoltage, a PLC restarting and a communications link failing may need different evidence.

Ask what electrical quantity would distinguish the competing explanations. For a suspected AC voltage dip, the relevant evidence may be short-interval RMS voltage, an event duration and a waveform. For a control-power failure, the important point may be a downstream power supply output. A mains analyzer at the building entrance cannot directly report every condition inside a machine.

Write one specific question: “Did the voltage at this feeder fall during the recorded drive alarm?” This gives the survey a boundary. It also allows a useful negative result: no qualifying dip was captured at that feeder under the stated settings.

Check the measurement point and connection mode

Confirm the actual location on the one-line diagram, including transformer, feeder and equipment boundary. A disturbance downstream of a loose connection may not appear in the same way at an upstream bus. If the complaint concerns the machine input, a recording at a different board needs an explanation.

Review the selected circuit topology and the reference used for each voltage channel. Phase-to-neutral and phase-to-phase voltages have different nominal values. An instrument configured for a different topology can produce plausible-looking numbers while applying inappropriate event references.

A current-only load study is also different from a voltage-quality study. Entering a nominal voltage to calculate estimated load quantities does not create a measured voltage waveform. Confirm the voltage leads were connected and those channels were part of the recording.

Where current is recorded, verify phase mapping and polarity too. The negative-kW and phase-mapping guide covers the checks that prevent misleading power calculations.

Convert trigger percentages into actual volts

Read the reference voltage and thresholds together. A percentage without its reference is incomplete. Fluke's logger setup guidance distinguishes the topology, nominal voltage, event limits and logging settings. The menu names on another instrument may differ, but these questions still need answers.

For an illustrative calculation, suppose a voltage-dip threshold is set to 90% of a 230 V nominal reference. That threshold corresponds to 207 V. A drop to 210 V would not cross it, even if particular equipment reacted. If the nominal value had mistakenly been entered as 120 V, the same 90% setting would correspond to 108 V. These are examples of settings, not recommended limits.

Check whether the analyzer uses a fixed reference or offers a sliding reference, and which one was active. Then inspect minimum duration, hysteresis, enabled channels and event type. A short excursion may fail a configured duration condition; a current alarm will not automatically create a voltage-dip record.

Do not lower every threshold until the report contains events. Set the monitoring criteria from the investigation and preserve the original settings. Excessively sensitive triggers can fill storage with unhelpful records and make the complaint harder to find.

A trend interval is not the same as event resolution

A five-minute trend can be useful for loading and energy. It does not show all the detail of a brief disturbance. Check whether the saved points are averages, minima, maxima or another statistic, and over what interval.

An illustrative 20% voltage reduction lasting 0.1 seconds in an otherwise steady 300-second interval occupies only one three-thousandth of that interval. A simple average would barely move. Actual instrument calculations vary, but the example shows why a smooth trend does not prove the absence of short events.

Dedicated event capture may operate separately from trend logging. It can have its own enabled state, thresholds, duration rules, waveform storage and capacity. Look in both the event table and the continuous recording. Confirm the instrument actually saves the waveform detail required for the investigation, rather than only an alarm timestamp.

If the suspected phenomenon is faster than the analyzer's documented capture capability, another measurement method may be needed. See the analyzer, logger and oscilloscope comparison before assuming that a different setting can supply missing bandwidth or isolation.

Prove that the recording was running

A display showing live voltage is not proof that a recording session is active. Inspect the session start and stop times, duration, file size and sample continuity. Check the battery or supply arrangement, available memory, automatic-stop settings and any warning history.

Then check the export. A session may exist in the instrument while the software is displaying another date range. Events may be excluded by a channel filter, severity filter or report option. Preserve the raw file before changing analysis settings.

Align the clocks before interpreting cause and effect

Record the analyzer time beside a trusted site time source when installing and retrieving it. Note the time zone and any offset. Do the same for PLC, drive, protection and building-management logs where available.

A one-minute difference can put the supposed cause after the equipment trip. A screenshot of two clocks is often more useful than a later assertion that both were “correct.” If the investigation depends on close timing, use instruments and synchronization methods appropriate to that requirement.

Preserve original timestamps when producing a corrected comparison. State the offset applied rather than silently rewriting the raw record. If the offset changed during the survey, a single correction may be inadequate.

Run a short acceptance check before leaving the analyzer

After reviewing the settings, make a short trial recording under normal operating conditions. Confirm that voltage and current channels, units, ratios, clock and exported data are correct. Check an ordinary scheduled load change against the event log if it falls within the configured criteria.

If an event-capture function needs positive verification, use the manufacturer's approved test method or a suitable controlled source in a safe test environment. Do not create a fault, interrupt a production supply or loosen a live connection to test the trigger.

Save the configuration alongside the trial file. An effective handover states which event types were enabled, the actual thresholds in units as well as percentages, minimum duration, selected channels, recording period and available waveform detail.

What to confirm on EK instruments

The EK-F500 public specification lists waveform display, disturbance capture, long-duration trend recording and configurable alarm limits. Treat these as separate functions when planning the survey. Confirm the current manual's trigger behavior, storage limits, recording mode and export workflow for the selected function.

The EK-F523A product page lists waveform and electrical-parameter recording. Its shorter public description does not establish every event threshold or capture detail needed for a particular fault. Ask for a demonstration of the required event and saved output before selecting it for that task.

Neither a broad function list nor the word “professional” establishes a measurement-class declaration. Where the contract names a class or standard edition, request supporting model-specific documentation. The Class A and Class S guide explains that procurement distinction.

Write a useful negative finding

A good report might say: “No voltage dip below the stated threshold, lasting longer than the configured minimum duration, was recorded at the monitored feeder during the 48-hour session. The equipment log contains two stops within that period. Clock alignment and data continuity were checked.” The actual numbers and evidence belong in the report.

That finding narrows the next investigation without claiming that every supply or equipment fault has been ruled out. It may justify moving the measurement point, investigating control power or reviewing equipment logs with the maintenance team.

When requesting technical help, include the model, firmware, connection diagram, monitoring location, nominal voltage, trigger configuration, a raw sample file and the equipment's timestamped symptom. Those details are far more useful than a photograph of an empty event list.