VA-TEK Engineering A VLF test set is not only a source of high voltage; it is a measuring instrument. The voltage it reports is the voltage the cable is actually tested at, and the test is valid only if that number, the frequency below it and the shape of the waveform are what the set claims. A set whose high-voltage measurement has drifted, or whose 0.1 Hz output is no longer a clean sine, produces a record that is wrong in a way nobody notices until a cable fails a test it should have passed — or passes one it should not.
This article covers the checks that keep a VLF test valid: what to verify before every test, what annual calibration covers, and the specific items that catch a drifting voltage or a distorted waveform.
Key takeaways
- Treat the VLF set as an instrument: the voltage it reads is the test result.
- Two failure modes matter — a wrong voltage (safety and validity) and a distorted waveform (invalidates comparison and PD).
- Field checks before every test; traceable calibration at intervals, typically annually.
- The high-voltage divider is the part most likely to drift, and the one most often left unchecked.
- Derived values — leakage current, tan delta — inherit any voltage error, so a drifting set corrupts more than the headline kV.
1. Why verification matters
A VLF set does two jobs at once. It generates the test voltage, and it reports that voltage as the test level. If the reporting is wrong, the record is wrong, and everything derived from it — whether the cable passed, what leakage current was normal, what tan delta was at 1.5 U0 — is wrong with it.
There are two distinct failure modes. A voltage error is a safety problem if the set under-reports (the cable is at a higher voltage than the operator believes) and a validity problem either way. A waveform or frequency error is quieter: a set that produces something other than a clean 0.1 Hz sine still returns a number, but the number is not comparable to a sinusoidal test, and the partial discharge behaviour changes with it.
2. Before every test: the field checks
These are quick, done on site, and catch the majority of problems:
- Record the instrument — set identifier and the date of its last calibration, on the test record. A test whose set cannot be traced to a calibration is weaker evidence.
- Inspect the high-voltage path — cables, connectors, the shield-ground bond, and the discharge rod. Look for damage, contamination and loose connections.
- Check the no-load baseline — energise the set with no cable connected and confirm the output and the meter read as expected at zero. A meter that reads something at no load is already suspect.
- Verify the ground path — a poor ground introduces noise, hides partial discharge and can distort readings.
- Confirm safety interlocks and the emergency stop function before an operator is near the far end.
3. Verifying the high-voltage measurement
The single most important item is the high-voltage measurement chain — usually a resistive or capacitive divider at the set output. This is the component that translates twelve thousand volts into a number on a screen, and it is the one most likely to drift with age, temperature and mechanical handling.
Periodically, the set’s reported voltage should be checked against an independent, calibrated reference — a separate high-voltage meter or a calibrated divider with its own traceable certificate. An internal self-check confirms the electronics are alive; it does not confirm the divider is accurate. Only a comparison against an external reference does.
4. Frequency and waveform
The output frequency should be confirmed at its nominal value (0.1 Hz, or whichever setting is used) rather than assumed. Beyond frequency, the waveform deserves attention:
- Confirm the set is producing a sinusoidal output when a sine number is being recorded. A set left in a cosine-rectangular mode still displays a voltage, and the number is not transferable to a sine result.
- Watch for distortion of the sine — flattening, clipping or added harmonic content — which changes the ratio between the rms value displayed and the peak voltage the insulation actually sees, and alters partial discharge behaviour.
As with voltage, the object is comparability: a record should state the waveform so that a later test on the same cable, or a standard’s voltage band, can be applied correctly.
5. Annual calibration
Traceable calibration is a separate exercise from the field checks and does not replace them. Typically performed annually, and more often for a heavily used or borrowed set, it establishes that the voltage and frequency measurements are accurate against a reference traceable to national standards, and it produces the certificate that makes the calibration defensible. High-voltage test equipment calibration generally follows the framework of IEC 60060; the cable test standard being applied may add its own requirements.
The two work together: calibration proves the instrument was right at a point in time; the field checks confirm it is still behaving on the day.
6. What a voltage drift does to the record
The reason to care about a small voltage error is that it propagates. Leakage current is a function of the applied voltage, and tan delta is derived from the ratio of the resistive to the capacitive component of that current. If the reported voltage is low by a few percent, the derived values are off by the same reasoning, and a cable can appear to be trending better or worse than it is. A test where the set was known to be within calibration is a defensible comparison point; one where it was not is not.
7. The habit
Treat the set as an instrument, record its identity and calibration status on every test, verify the obvious on site, and put the high-voltage chain in front of a traceable reference on schedule. The test voltage is the one number the whole test is built on; it is worth the few minutes to know the set is telling the truth about it. For how the verified voltage is chosen in the first place, see VLF test voltage selection, and for the overall procedure see the practical VLF guide.
Frequently asked questions
How often should a VLF test set be calibrated?
Carry out the field checks before every test, and a full calibration on the manufacturer’s interval or your own quality-assurance cycle. Confirm the interval against the instrument’s procedure and the requirements of the standard you test to.
What must be checked before every test?
Ground continuity before energisation, the integrity of the high-voltage measurement chain (divider, lead and instrument), and that the set regulates to the setpoint in closed loop. Where a known reference is available, check against it.
Why does voltage drift matter?
Because a result is only defensible if the applied voltage is known. A drifting measurement chain makes a pass/fail judgement and any trend meaningless, and the record cannot be compared from one year to the next.
