The impulse voltage withstand test fires a short, steep surge across the insulation to see whether it survives the kind of spike that arrives from a lightning strike or a switching event. The waveform is 1.2/50 µs — roughly 1.2 µs to reach peak, 50 µs to fall to half.
Here’s the part most write-ups skip. This is not one test. It’s the same waveform doing two different jobs, on two different things, with different rules for each. Run it as one test and you will get the pulse count wrong, apply an altitude correction where it doesn’t belong, and misread what counts as a failure.
Table of Contents
Same waveform, two jobs
On a clearance, you’re testing air. Air is self-restoring. If it flashes over, the gap is undamaged once the arc clears. You’re asking one question: is the air gap big enough for the rated impulse withstand voltage?
On solid insulation, you’re testing material that does not recover. Every puncture is permanent. You’re asking whether the material has a weak spot that a fast, steep front will find.
Where the two diverge:
| Clearance | Solid insulation | |
|---|---|---|
| Impulses per polarity | 3 minimum | 5 |
| Interval between impulses | At least 1 s | At least 1 s |
| Altitude correction | Applies | Does not apply |
| Waveshape recording | Not required | Required |
| Partial discharge during test | Disregarded unless specified | Allowed |
| Failure | Disruptive discharge | Puncture or partial breakdown |
| Can be replaced by AC/DC test | Yes | No, not for AC stress |
The altitude line is the one that catches people. Clearance depends on air density, so a gap tested at 3000 m needs correction. Solid insulation doesn’t care about air pressure — the correction factors simply don’t apply to it. Applying them anyway means you’re testing at the wrong voltage.
Why multiple impulses, both polarities
Impulse breakdown scatters. Fire the same voltage at the same gap ten times and you won’t get ten identical results. A single shot proves very little. The repeat count is there to push through that scatter.
Both polarities matter because breakdown is not symmetric. A non-uniform field — a sharp edge facing a flat plate — has a different withstand level depending on which electrode is positive. Test one polarity only and you may test the easy direction.
The 1 s minimum between impulses lets the insulation and the circuit settle so each shot starts from the same condition.
Generator and setup
Generator output impedance stays at or below 500 Ω. Where the equipment has components sitting across the test circuit, a much lower effective impedance may be called for, and then resonance becomes a live concern — the peak at the insulation can end up higher than what the generator thinks it’s delivering.
That leads to the setup rule that matters most on assembled equipment: make sure the specified voltage actually appears at the point you’re testing. A surge propagating through a real circuit gets attenuated or amplified on the way. The number at the generator terminals is not automatically the number at the clearance.
Practical points for testing complete equipment:
- Disconnect surge protective devices before any dielectric test. An SPD does its job and clamps your test.
- Tie the external terminals of each circuit together.
- Put switchgear in the closed position, or bypass it.
- Short across voltage-blocking components like rectifier diodes.
- Keep RFI filters in circuit for the impulse test. They may need disconnecting for AC tests.
- Cover non-conductive accessible surfaces with metal foil. On a large enclosure, foil the parts that provide shock protection.
- Pre-tested plug-in boards and modules can be pulled or replaced with dummies, as long as the test voltage still reaches everywhere it needs to.
Reading the result
For clearances, a pass means no disruptive discharge — no sparkover, no flashover, no puncture. Partial discharges in the clearance that don’t lead to breakdown are disregarded. An oscilloscope on the impulse is the practical way to catch a disruptive discharge.
For solid insulation, the criterion is different in a way that’s easy to misapply. No puncture and no partial breakdown. But partial discharges are allowed during the test — they’re not a failure here. That’s what the separate partial discharge test exists for.
Partial breakdown shows up as a step in the recorded waveshape, and the step appears earlier in each successive impulse. That progression is the tell. It means damage is accumulating shot to shot. This is exactly why the waveshape has to be recorded for solid insulation and why one impulse would never reveal it.
Discharges in voids can also produce very short notches on the trace, repeating within a single impulse. Those are not the same as the stepping pattern of partial breakdown.
When AC or DC stands in
An AC or DC test can substitute for the impulse test on clearances, at the same peak value. Both prove the gap holds.
The catch: they hold that peak far longer than a 50 µs tail does. Any solid insulation sharing the path takes much more stress than the impulse would have applied, and some materials will be damaged or destroyed by that. Before substituting, look at what solid insulation is in the path.
For AC, three cycles at peak equal to the impulse test voltage. For DC, three applications, with the average value equal to the impulse test voltage and ripple held tight.
The substitution does not work in reverse. An AC test on solid insulation cannot be replaced by an impulse test. Two reasons: impulse and power frequency voltages distribute differently through a complex circuit, and solid insulation withstand depends heavily on how long the stress lasts. A material that shrugs off a 50 µs surge can fail under 60 seconds at the same peak.
One shortcut does exist. If the peak of the AC test voltage is at or above the rated impulse withstand voltage, the AC test already covers the impulse test, and a separate impulse test isn’t needed.
Conditioning before the test
Testing a brand-new specimen can hide faults that only appear after service exposure. Conditioning is applied before type testing to surface them.
Four methods, each targeting a different defect:
- Dry heat — settles the material into a stable state that may not exist straight off the line.
- Temperature cycling at a controlled rate — creates the voids that develop through storage, shipping, and use.
- Thermal shock — induces delamination inside the insulation system.
- Damp heat, steady state — evaluates what absorbed moisture does to electrical properties.
For impulse and AC voltage tests, dry heat and damp heat are the relevant ones. Temperature cycling and thermal shock matter more ahead of PD testing, since they’re the ones that make voids.
If a component has already been type tested with conditioning, you don’t repeat it.
Test parameters at a glance
| Parameter | Value |
|---|---|
| Waveform | 1.2/50 µs |
| Impulses, clearances | 3 minimum per polarity |
| Impulses, solid insulation | 5 per polarity |
| Minimum interval | 1 s |
| Generator output impedance | 500 Ω or lower |
| Impulse voltage accuracy | ±5 percent |
| Impulse timing accuracy | ±20 percent |
| Ambient temperature | 15 °C to 35 °C |
| Air pressure | 86 kPa to 106 kPa at sea level |
| Relative humidity | 25 percent to 75 percent |
| AC substitute | 3 cycles, peak equal to impulse test voltage |
| DC substitute | 3 applications, average equal to impulse test voltage |
FAQ
Why five impulses for solid insulation but three for a clearance? Air is self-restoring, so the extra shots add little. Solid insulation accumulates damage, and partial breakdown only reveals itself across successive impulses as the step in the waveshape moves earlier.
Does altitude correction apply to the impulse test? For clearances, yes — air breaks down more easily at altitude. For solid insulation, no. The correction factors don’t apply to it at all.
Are partial discharges a failure during an impulse test? No. On solid insulation they’re explicitly allowed. On clearances they’re disregarded unless they cause breakdown. Sustained discharge behavior is a separate test.
Can I use an impulse test instead of the AC test on solid insulation? No. Impulse and power frequency voltages propagate differently through a circuit, and solid insulation weakens as stress duration grows. The impulse test doesn’t cover what the AC test looks for.
When can I skip the impulse test entirely? When the peak of the AC test voltage is at or above the rated impulse withstand voltage. The AC test then covers it.
Why record the waveshape? Because partial breakdown in solid insulation doesn’t show as an obvious failure. It shows as a step that arrives progressively earlier across the impulse sequence. Without a record you’d pass the part.