logo
blog
Home > blog > Company blog about How to Maintain High-Voltage Electrical Testing Equipment
Events
Contact Us

How to Maintain High-Voltage Electrical Testing Equipment

2026-11-19

Latest company news about How to Maintain High-Voltage Electrical Testing Equipment
Maintenance Guide

How to Maintain High-Voltage Electrical Testing Equipment

A working routine for hipot sets (high-potential withstand testers), VLF (very low frequency) generators, megohmmeters, relay testers and cable fault locators - written for the people who own the equipment, not only for the lab that calibrates it.

Reading time about 15 minutes | For procurement, plant and field test teams

A high-voltage test set rarely fails with a bang. It drifts. A unit that reads 2% low still switches on, still prints a curve, and still says PASS. The first sign of trouble is usually a failed joint that should have passed, or a good cable that was rejected on site. Maintenance exists to stop that from happening.

What maintenance is really protecting

First, a plain definition. A high-voltage test set applies a controlled voltage above the normal working level of a piece of equipment, then measures how much current leaks through its insulation. That single idea covers withstand tests, leakage tests and most diagnostic tests that follow.

For a multimeter, maintenance mostly means "keep it working". For high-voltage test equipment, three things are at stake at the same time, and they fail in different ways.

  • Operator safety. The set generates the hazard it measures. Its own insulation, grounding and interlocks are part of the safety chain, not accessories to it.
  • Trust in the reading. A test result is a decision: pass or fail, energise or repair. A 1% voltage error on a 100 kV test moves the stress on the specimen by 1 kV, which is enough to flip a marginal result.
  • Availability. Field work is scheduled around outages. A set that trips on the morning of a shutdown costs more than its repair invoice.

A useful way to think about it: a torque wrench that reads 5% high tightens every bolt too far, and nothing about the wrench looks wrong. High-voltage test sets behave the same way. The reading is the product, so the instrument has to be maintained as a measuring device, not just as a machine.

Start with the leads, because that is where most problems live

In service experience, the high-voltage generator is usually the healthiest part of the kit. The parts that fail are the ones that get coiled, dragged, stepped on and thrown into a van: output cables, ground leads, clamps, adapters, discharge rods and dividers.

What to look for

  • Intermittent conductors. A cable with broken strands inside an intact jacket looks perfect and fails only when it is bent. Flex the cable along its length during inspection, not just at the ends.
  • Surface contamination. Dust, oil and metal dust on a connector form a leakage path. On a 60 kV lead, a contaminated surface can draw more current than the specimen, and the reading becomes meaningless.
  • Oxidised clamps. Contact resistance at the clamp adds to the circuit and can heat under load. Clean contacts, and replace clamps that no longer spring.
  • Crushed or kinked cable. A tight bend damages the insulation from the inside. Keep the bend radius generous, and never let a vehicle door close on a lead.
  • Damaged dividers. A dropped divider may still work at low voltage and flash over at high voltage. Treat any impact as a reason to verify the ratio before the next job.
Practical habit: store high-voltage leads hanging or in a loose coil, separately from tools and sharp parts. A lead that is stored well lasts years longer than one that is stored fast.

Routine Checks Before and After Every Test

Most field failures are caught by a short, repeated check rather than by an annual service. The routine below is deliberately boring, which is the point.

Before the test

  1. Inspect the set, leads and connectors for damage, moisture or contamination.
  2. Confirm the ground connection first, before anything else is connected. Ground is not a formality; it is the return path that keeps the chassis at earth potential.
  3. Run the self-test or start-up check, and confirm the battery or supply is adequate for the planned duty.
  4. Check that the lead and divider ratings exceed the highest voltage you intend to apply, with margin.
  5. Confirm the correct divider ratio or correction factor is entered for the divider in use.

After the test

  1. Follow the manufacturer's shutdown procedure for the test you have just run, and reduce the output to zero before switching off where the instrument allows it.
  2. Discharge the specimen and the leads with a properly rated discharge device, then apply a direct earth connection where the procedure requires it.
  3. Verify residual voltage with an instrument suitable for that voltage, and read the result together with the discharge and grounding steps. Do not rely on the set's own display alone.
  4. Clean and dry the leads if the work was outdoors or in damp conditions.
  5. Record the result, the asset ID and anything unusual while it is still fresh.

Why discharge is not a formality

DC tests charge the specimen like a capacitor, and a large part of that energy is still there after the set is switched off. The amount is easy to underestimate.

Stored energy in the specimen

E = 0.5 x C x U^2

For a 5 µF cable at 60 kV: E = 0.5 x 5e-6 x (60,000)^2 = 9,000 J (9 kJ). That is a dangerous amount of energy, and it explains why discharge is a procedure with steps, not a button press.

What happens after the test depends on the type of test and on the specimen. The three cases below cover most field work, but they are general guidance only. The instrument manual and your site safety procedure define the exact steps.

After DC withstand testing

Switching off the output, or watching the indicated voltage fall to zero, does not necessarily mean the specimen is fully discharged. Cable insulation, long test leads and the specimen's own capacitance can hold charge that the display no longer shows. A practical sequence:

  • Reduce the output voltage to zero according to the equipment manufacturer's procedure, rather than removing power under load.
  • Discharge the specimen with a discharge device rated for the voltage and stored energy involved. A probe that is adequate for a small specimen may not be adequate for a long cable.
  • After discharging, apply a direct earth connection to the specimen, and keep it in place for the period required by the applicable procedure.
  • For long cables and high-capacitance specimens, repeat the discharge check and watch for voltage recovery. Dielectric absorption can bring the terminal voltage back up after an apparently complete discharge.
  • Do not treat a fixed waiting time or a single low reading as universal proof of safety.

After insulation resistance testing

An insulation resistance test can leave dielectric absorption, or polarization-related charge, in the specimen even after the terminal voltage has fallen to a low value. This is most noticeable on large windings and long cables. Discharge to earth and keep the connection for the period your procedure specifies.

IEEE 43-2013 gives a minimum discharge-time recommendation of four times the voltage application time. That recommendation applies to the rotating-machine winding insulation tests within the scope of that standard. It is not a universal requirement for every megohmmeter test, for cables, or for other insulation systems. Where a manufacturer's instruction or a product-specific test procedure gives a different time, that instruction takes priority.

After VLF testing

A VLF test charges a cable in a different way from a DC test, but the ending deserves the same discipline. A common sequence is:

  1. Slowly reduce the test voltage to zero.
  2. Discharge the test object using the specified discharge arrangement.
  3. Apply a direct earth connection and keep it connected as required by the applicable procedure.

IEEE 400.2-2024 can be referenced for the general principle of slowly reducing the voltage, discharging the cable and maintaining grounding. It should not be read as a universal instruction to hold a specific voltage-reduction period for every cable and every set. Follow the edition of the standard, the cable manufacturer's guidance and your own procedure for the actual timings.

The safety point that applies to all three

Zero indicated voltage does not mean zero stored charge. The display shows what the instrument can measure at its terminals, not what remains inside a cable or a winding.

Discharge and grounding are two separate steps. Verification of residual voltage is necessary, but it has to be interpreted together with the discharge and grounding procedure, not instead of it. Never touch, short or disconnect a test circuit merely because the instrument reads zero.

Treat this section as general guidance, not as a universal operating procedure. Exact voltages, discharge device ratings, waiting times and grounding arrangements vary by instrument, specimen and standard. The equipment manual and the site safety procedure are the controlling documents.

Keep the high-voltage section dry

Moisture is the most common environmental cause of inaccurate readings and flashovers. The mechanism is simple: water films on insulation surfaces create leakage paths, and leakage current adds to the current you are trying to measure.

Where moisture comes from

  • Condensation after moving the set. A cold set carried into a warm, humid room will sweat internally. Let it reach room temperature before applying voltage.
  • Storage in unheated rooms or containers. Temperature swings drive moisture in and out of insulation and enclosures all year.
  • Outdoor work in rain, fog or high humidity. Even a brief exposure leaves films on connectors and bushings.

What works

  • Store the set in a dry, ventilated room, covered but not sealed in a way that traps moisture.
  • Use silica gel in the case and replace or regenerate it on a schedule, not when it is visibly wet.
  • Allow a warm-up period of roughly 10 to 15 minutes before high-voltage tests, especially after transport in cold weather.
  • Clean bushings and connectors with a dry, lint-free cloth. Use isopropyl alcohol on metal contacts only, never on epoxy, silicone or XLPE surfaces unless the manufacturer approves it.
Avoid solvents on insulation. Cleaning agents that are fine on a metal panel can attack polymer surfaces and leave a conductive residue. When in doubt, use a dry cloth and let the surface dry fully.

Measure the set's own insulation, not only the specimen's

A megohmmeter check on the high-voltage circuit of the set, at a voltage the insulation can take, gives an early warning that moisture or contamination is building up. One reading proves little. The trend across months is what tells you something is changing.

The instrument's own high-voltage circuit is one thing; its internal electronics are another. An external megohmmeter should only be connected to test points and circuits that the equipment manufacturer permits users to test, and the test voltage must suit that specific circuit. Do not apply insulation resistance test voltage indiscriminately to internal electronic circuits, measurement modules or protective components. If the manufacturer's documentation does not describe an approved test method, consult the manufacturer or qualified service personnel rather than improvising one.

Polarization index, for insulation condition

PI = R(10 min) / R(1 min)

PI compares resistance after one minute with resistance after ten. A flat ratio suggests the insulation is stable; a ratio that falls over successive checks suggests moisture or contamination. Use the limits from the applicable standard and from the equipment manufacturer, and treat a single abnormal result as a reason to investigate, not as a final verdict.

Calibration and verification are not the same job

These two activities get mixed up, and the confusion causes most of the arguments about "calibration intervals".

Calibration is a comparison against a reference that is traceable to a national standard, performed by a competent laboratory, and reported with measurement uncertainty. It tells you how far the instrument is from the truth, and it does not necessarily adjust anything.

Verification is a quick in-house check between calibrations, using a divider, a reference meter or a known resistance. It tells you whether the instrument has moved since the last calibration. It does not replace calibration, but it catches drift early and it is cheap.

Correcting a reading with a divider ratio

Divider-based measurement is the normal way to read high voltage, and the ratio is where errors enter. The correction is simple once the purpose is clear: you are converting what the instrument displays into what is actually at the specimen.

Divider correction

k = U(reference) / U(indicated) and U(true) = U(indicated) x k

If a 1000:1 divider and its meter indicate 60.0 kV while the reference reads 60.6 kV, then k = 1.010 and every future reading should be multiplied by 1.010. Record k with the divider, not with the person who measured it. When k moves outside the instrument's specified accuracy - about 1% for many high-voltage sets, wider for resistance ranges - the divider or meter needs attention rather than a new correction factor.

A workable interval

ActivityTypical intervalNotes
Visual and functional checkEvery useLeads, ground, self-test, battery
In-house verificationMonthly, or before critical workDivider ratio, current reading, timer
Traceable calibration12 monthsShorten for heavy use, safety-critical work or harsh environments
Insulation check of the set6 to 12 monthsCompare with the previous record
Battery capacity check6 monthsReplace when runtime falls below about 70% of new

These are reasonable starting points, not universal rules. Your quality system, the manufacturer's instructions and how hard the set works all move the interval. A set used daily on a production line deserves a shorter cycle than one used twice a year.

Power, batteries and heat

High-voltage sets concentrate loss in a small package, and both batteries and insulation dislike heat. Two habits prevent most of the damage.

Battery care

  • Store lithium-ion packs at roughly 40% to 60% charge if the set will be idle for months. Full charge plus heat ages cells fastest.
  • Avoid running the pack to cutoff habitually. Recharge before storage, not after a long idle period.
  • Replace packs that swell, that charge unusually fast, or that deliver noticeably shorter runtime. A pack at end of life can also sag under load and cause spurious trips.
Follow the cell manufacturer first: where the battery manufacturer specifies a storage state-of-charge range, use that guidance. The 40% to 60% range above is a conservative default when no specific instruction is available.

Respect the duty cycle

High-voltage generators are rated for a working pattern, not for continuous operation. For example, the XHZG series DC hipot sets are rated for 30 minutes at rated load and 10 minutes at 1.1 times rated voltage. That is a product-specific working duty for that series, not a universal duty cycle for all high-voltage generators. The actual operating and rest periods for any instrument must follow its rated duty and the manufacturer's instructions. Exceeding the rating heats the transformer, the multiplier and the insulation, and heat shows up first as increased ripple (the small AC component riding on the DC output) and drift, then as failure.

Rule of thumb: if the set is hot to the touch or the ripple reading has risen during a long session, stop and let it cool. Cooling is part of the test procedure, not a delay.

Cooling and environment

Keep vents and fans clear, clean dust filters, and do not operate the set inside a closed vehicle or a sealed cabinet. As a reference example, the XHZG series is specified for an ambient range of about -10 °C to +40 °C, relative humidity up to 85% and altitudes below 1500 m. Those figures describe typical reference conditions for that product family, not universal ratings for every instrument. At high altitude, thinner air reduces both cooling and withstand margin, so derating may be necessary; check the nameplate and manual for the equipment you actually have.

Records: the part most teams skip

A maintenance programme without records becomes a set of opinions. Three records carry most of the value:

  • An asset register with one ID per instrument, including leads and dividers, because accessories travel separately and get mixed up.
  • A calibration and verification log tied to that ID, with the correction factor in use at the time.
  • A test-result baseline for the set itself - for example, leakage current at a fixed voltage, recorded the same way each time.

Trending turns a single number into a warning. A leakage current that rises steadily suggests the set's insulation is absorbing moisture or contamination, even while it still passes a go/no-go check.

Leakage trend

Delta I(%) = (I(current) - I(baseline)) / I(baseline) x 100

Compare like with like, or the trend means nothing. A baseline is only useful if it was taken under conditions you can reproduce: the same test voltage, the same warm-up time, the same ambient temperature and humidity, the same lead arrangement and connections, and the same test configuration. Establish the baseline when the equipment is known to be in good condition - preferably after calibration or a satisfactory verification. Take at least three consecutive measurements under those consistent conditions and use a stable representative value, rather than a single reading.

As an in-house guide, a rise beyond roughly 20% is worth investigating and beyond 50% is a reason to stop using the set for critical work until it is checked. Those two numbers are internal maintenance policy examples. They are not IEC or IEEE requirements, and an abnormal trend is not automatic proof that the insulation has failed. Repeat the measurement, confirm the conditions, and investigate before drawing a conclusion.

Instrument memory is not a backup. Many sets hold a limited number of records - ninety is typical on some VLF units - and then overwrite the oldest. Export results monthly and keep the calibration certificate with the asset record.

What to check on each type of instrument

EquipmentMaintenance focus
DC hipot / withstand setOutput cable and multiplier condition, ripple, voltage and current accuracy, duty cycle, discharge function
VLF generatorWaveform quality, frequency stability, cable and connector condition, cooling, stored records
High-voltage dividerRatio stability after impact or transport, surface cleanliness, stable mounting and ground
Insulation oil BDV (breakdown voltage) testerElectrode condition and gap, oil cup cleanliness and dryness, calibration of the voltage ramp
MegohmmeterLead insulation, guard terminal, battery, resistance accuracy at the top of each range
Relay protection testerOutput accuracy on current and voltage channels, channel balance, firmware, connector wear
Circuit breaker analyserContact timing channels, travel transducer calibration, clamp condition, timebase accuracy
CT / PT analyserRatio and excitation accuracy, lead sets, grounding practice, software version
Cable fault locator and surge generatorCapacitor bank health, discharge gap, earthing, surge cable condition, safe discharge after use

When something looks wrong: first actions

SymptomLikely causeFirst action
No high voltage outputInterlock open, ground not connected, supply or battery faultCheck grounding and interlocks before suspecting the generator
Ripple higher than normalOverload, overheating, failing filter capacitorStop, cool, retest at reduced load; compare with baseline
Reading lower than expectedWrong divider ratio, contaminated leads, poor groundVerify ratio and clean connectors; check with reference divider
Trips immediately on a cableGenuine specimen fault, or a fault in the lead setDisconnect the specimen and test the set alone to separate the two
Erratic current displayLoose connector, moisture, EMI from nearby equipmentDry and reseat connections; move away from noise sources
Over-temperature shutdownBlocked vent, high ambient, exceeded duty cycleClean vents, allow cooling, review the test plan
Short battery runtimeAged cells, cold conditions, charger faultCapacity test; replace pack if below about 70% of rated runtime
Software will not connectCable, driver, firmware or port conflictSwap cable and port first; then check firmware and driver versions

A yearly plan that fits real schedules

FrequencyTasks
Every useVisual check, ground first, self-test, duty planning, discharge and residual voltage check
MonthlyIn-house verification, lead and clamp inspection, data export, battery top-up
QuarterlyDeep clean, fan and filter check, insulation resistance of the set, record review
AnnuallyTraceable calibration, full functional test, accessory replacement review, firmware update
After any impact or faultVerification before return to service; investigate rather than reset and continue

Safety rules maintenance must never break

  • Never open the high-voltage enclosure or adjust internal parts while the set is connected to a specimen, or while any charge may remain.
  • Ground first, disconnect last. Treat every conductor as live until it has been discharged and verified.
  • Do not bypass interlocks, even temporarily, to "finish the job". They exist because the failure mode is fatal.
  • Use a properly rated discharge device and then a direct earth connection where the procedure requires it; a zero display is not proof that the circuit is dead.
  • Work with a second person present for high-voltage tests and maintenance, and follow your site's lock-out/tag-out procedure.
  • Keep the working distance, use the specified PPE, and never reach into a test circuit to reposition a clamp while voltage is applied.

Repair or replace?

Repair usually wins when the fault is in a lead, a connector, a battery or a fan, and when spare parts and service are available. Replacement becomes the better decision when calibration cannot be brought back within specification, when high-voltage insulation is damaged, when the unit is unsafe, or when the repair cost approaches half the price of a new instrument. Age alone is not a reason; a well-maintained 15-year-old hipot set can be more trustworthy than a new one with no records.

Key takeaways

  1. Maintain the instrument as a measuring device: safety, accuracy and uptime are three separate goals.
  2. Leads, clamps and dividers fail far more often than the high-voltage generator.
  3. Ground first, discharge and verify residual voltage after every test.
  4. Dry storage and a short warm-up prevent most moisture-related errors.
  5. Verify in-house between traceable calibrations, and keep the correction factor with the divider.
  6. Respect the duty cycle; heat shows up as ripple and drift before it shows up as failure.
  7. Trend the set's own leakage current and keep records tied to an asset ID.

Frequently asked questions

How often should high-voltage test equipment be calibrated?

Twelve months is the common default, but it is a starting point rather than a rule. Shorten the interval if the set is used daily, if results are safety-critical, or if it works in heat, humidity or vibration. Lengthen it only with evidence from verification records, and only if your quality system allows it.

Can I use the set if the calibration certificate has expired?

For non-critical, internal checks you may choose to continue with a documented risk decision, supported by an in-house verification. For acceptance testing, customer witness tests or safety-related work, an expired certificate is usually not acceptable. The decision belongs to your quality procedure, not to the person holding the test lead.

What causes a high-voltage set to read low?

The three usual causes are a wrong divider ratio, contamination on the high-voltage leads creating a leakage path, and a poor ground connection. Check those before touching any internal adjustment. A set that reads low because of a ratio error will read consistently low, while a contamination problem often varies with humidity.

Is it safe to store the equipment in an unheated container?

It is common but not ideal. Temperature swings cause condensation, and condensation on high-voltage insulation is exactly what you do not want. If outdoor storage is unavoidable, use a sealed case with fresh desiccant, allow the set to reach room temperature before use, and check insulation resistance more often.

How do I know when a battery pack needs replacing?

Measure runtime under a realistic load, not idle time. When usable runtime falls below roughly 70% of the original figure, or the pack swells or becomes unreliable under load, replace it. A weak pack can cause intermittent shutdowns that look like instrument faults.

Do I need to recalibrate after replacing a lead or divider?

Replacing a lead normally needs only a verification, not a full recalibration. Replacing or repairing a divider changes the measurement chain, so verify the ratio against a reference before returning the set to service, and record the new correction factor.

Equipment referenced in this guide

These are examples of the instrument types covered above, with the specifications that matter for maintenance planning.

Need a maintenance schedule for your test set?

Tell us the instrument type, the highest test voltage and how often it is used, and we will help you set a practical verification and calibration plan - including the lead set, divider and accessories that usually get forgotten.

Scope note: this guide describes general good practice for high-voltage test equipment. Always follow the manufacturer's instructions and your site's safety procedures. Intervals and thresholds quoted here are typical starting points, not mandatory requirements.

Send your inquiry directly to us