VA-TEK Engineering To prove or diagnose a medium-voltage cable in the field you need an alternating test voltage, but 50 Hz is impractical over any real cable length and DC damages extruded insulation. Two technologies answer that problem from opposite directions. VLF applies a continuous sine wave at 0.1 Hz; DAC (damped AC, sold in Europe as an OWTS or oscillating wave test system) charges the cable and then lets it ring down through an LC circuit at tens to hundreds of hertz. Both are AC, both avoid DC, and both are named in current standards — but they stress the cable differently and are good at finding different defects.
This article compares the two methods stage by stage: how each generates its AC voltage, what the evidence says about what each one detects, which standards govern them, and how to decide which belongs on a given cable.
Key takeaways
- VLF is a continuous 0.1 Hz sine wave; DAC/OWTS is a decaying AC burst at 20–500 Hz generated by DC pre-charging the cable, then discharging it into an inductor.
- Their frequencies sit at opposite ends of the AC spectrum, and that is the whole story: VLF’s low frequency makes it portable (charging current is ~1/500 of 50 Hz); DAC’s higher frequency makes partial discharge behave more like it does at service frequency.
- Defect selectivity differs: DAC tends to catch void-discharge defects, VLF tends to catch electrical-tree-initiated defects.
- A same-cable field study found only weak agreement between the two — most partial-discharge events were seen by one method and not the other. Results are not directly interchangeable.
- VLF is governed by IEEE 400.2-2024; DAC by IEEE 400.4-2015. GB/T 3048.8-2025 lists both, as informative annexes.
1. How each method makes its AC voltage
The two methods differ first in how the alternating voltage is produced, and everything else follows from that.
VLF is a power supply: it generates a continuous sinusoidal output at a selectable frequency — 0.1 Hz is the field standard, with 0.05, 0.02 and 0.01 Hz available on longer cables. Because the frequency is a hundredth to a thousandth of the mains, the capacitive current the cable draws collapses. Charging current is proportional to frequency, so at 0.1 Hz it is roughly 1/500 of the 50 Hz value (1/600 at 60 Hz). That is why a VLF set is a box of tens of kilograms rather than a substation transformer: the low frequency, not any magic in the electronics, is what makes field testing practical.
DAC / OWTS does not generate a continuous wave at all. It first charges the cable with a high-voltage DC source to the required peak voltage, then closes a fast switch. The cable capacitance and a chosen inductor form an LC circuit, and the stored energy rings out as a damped alternating oscillation that decays to zero in tens to hundreds of milliseconds. The oscillating frequency is set by the hardware and the cable:
f = 1 / (2π√(L · Ccable))
so a longer cable (more capacitance) rings at a lower frequency, and the operator switches inductor taps to keep the frequency in range. Typical working bands are quoted around 20–500 Hz — close to service frequency, which is the method’s central claim.
2. The two methods side by side
| Dimension | VLF (0.1 Hz sine) | DAC / OWTS (damped AC) |
|---|---|---|
| Voltage shape | Continuous sine wave | Decaying oscillating burst (LC ring-down) |
| Frequency | 0.01–1 Hz, usually 0.1 Hz | ~20–500 Hz, varies with cable length |
| How it is made | Direct AC power supply | DC pre-charge, then switched LC discharge |
| Charging current | Low (1/500 of 50 Hz at 0.1 Hz) | High during each burst |
| Strong at | Continuous withstand; tan delta trend; tree-initiated defects | Partial-discharge detection and localisation; void defects |
| Portability | High (tens of kg for MV sets) | Medium; inductor grows with voltage and cable length |
| Main weakness | Weak partial discharge can go unnoticed | Frequency drifts with length; results are system-specific |
| Governing standard | IEEE 400.2-2024 | IEEE 400.4-2015 |
One comparison caution: check the voltage basis before comparing numbers. A VLF set is rated by the peak of its sine wave; a cosine-rectangular VLF output has a different rms-to-peak relationship again, and DAC voltages are quoted on their own basis. Two “60 kV” machines from two methods are not necessarily stressed to the same level.
3. The frequency argument — both directions
Advocates of each method argue from the frequency, and both arguments are legitimate; they just point at different things.
The case for low frequency. Rejecting DC was about space charge, not about wanting a slow wave. A 0.1 Hz sine keeps the field alternating, avoiding the trapped-charge problem DC creates in extruded insulation, while keeping the charging current small enough to test a long cable from a portable unit. A continuous wave also lets the operator watch leakage current and tan delta evolve smoothly across the whole hold time, which is what makes a monitored withstand possible.
The case for near-service frequency. Partial discharge is frequency-dependent. A DAC burst sits at 20–500 Hz, closer to the 50/60 Hz the cable lives at, so detectors argue the discharge behaviour it excites is more representative of service conditions. The short, high-dV/dt burst also spreads PD pulses in time, which helps separate a defect in the cable body from one in a joint and supports reflection-based localisation along the length.
Neither argument wins outright. VLF wins on continuous conditioning and portability; DAC wins on PD realism and location. The next section is where the difference actually shows up in the field.
4. What the evidence shows
The most useful comparison is a same-cable test: one cable, cured, measured by both techniques. A CIGRE study on a 22 kV distribution network in Australia did exactly that across 78 circuits and roughly 700 joints. The result was sobering for anyone hoping to swap one method for the other — of the partial-discharge events recorded, only about 40% were seen by both methods; the majority were detected by one and missed by the other. VLF sine readings were on average about 50% higher in magnitude, and the DAC frequency drifted across the network with cable length. Localisation, where both could localise, agreed.
Physical selectivity follows the same split. A foundational IEEE study on defect selectivity found that DAC is sensitive to defects driven by void discharge, while VLF is sensitive to defects that start with an electrical tree. That is a mechanism difference, not a quality difference: the two methods are looking at overlapping but not identical halves of the defect population.
Treat all of this with the usual caution about commercial positions. Much of the material comparing the two methods is published by manufacturers of one or the other, and some widely repeated claims about detection rates are vendor positions rather than findings endorsed by IEEE or CIGRE. The published guidance — CIGRE technical brochures on after-laying tests and on-site PD assessment, and the IEEE 400 series — treats both as valid AC approaches with different roles, not as a ranking.
5. Standards
| Standard | Covers | Status |
|---|---|---|
| IEEE 400.2-2024 | VLF (<1 Hz) field testing of shielded cable | Active |
| IEEE 400.4-2015 | DAC field testing, 5 kV and above | Inactive – Reserved; revision in progress |
| IEC 60060-3 | Field test voltages, incl. VLF and damped AC | Current |
| GB/T 3048.8-2025 | AC voltage test methods — lists VLF (Annex A.1) and oscillating wave (Annex A.2) side by side | Published 2025 |
Two standard notes matter for procurement. First, IEEE 400.4 is currently Inactive – Reserved with a revision project under way, so a buyer should confirm the current status before citing it as the governing document for a DAC test. Second, GB/T 3048.8-2025 carries both methods as informative annexes — useful for showing that the two sit alongside each other in national practice, but not a mandate that either be applied to a given cable.
6. Choosing between them
The honest answer is that the choice follows the question you are asking.
- Routine MV acceptance and maintenance on extruded cable — a VLF sine withstand is the default: portable, standard-backed, and it conditions the cable continuously. See how the VLF test voltage is selected.
- Insulation ageing and water-tree assessment — VLF combined with tan delta tracks a trend over time far better than a single pass/fail.
- Finding and locating a specific defect, especially on long circuits — a DAC/OWTS test, with its near-service PD behaviour and reflection-based distance measurement, has the sharper tools.
- When both are available — do not assume one result substitutes for the other. The same-cable evidence says the two see different defects.
One rule holds regardless of method: record the waveform, the frequency and the voltage basis with every result. A “2.5 U0” figure means different things at 0.1 Hz sine, at cosine-rectangular, and at a 200 Hz damped oscillation, and a number without its basis cannot be compared, trended or defended later.
7. Where this leaves you
VLF and DAC are not rivals so much as two AC answers to the same field problem, chosen for different strengths: VLF for continuous, portable, standard-anchored withstand and condition testing; DAC where partial-discharge realism and defect location drive the decision. Whichever you choose, the cable is stressed with alternating voltage, so both sit on the right side of the DC-versus-AC question that reshaped cable testing, which is set out in VLF versus DC testing.
If the next step is a VLF set — sine output, 0.1 Hz standard with 0.05/0.02/0.01 Hz for long cables, and models covering 30 to 90 kV peak — the VLF hipot test set range lists the voltage classes and loads each covers. For the wider VLF method, see the practical VLF guide to IEEE 400.2-2024, and for reading discharge under VLF, VLF partial discharge testing.
Frequently asked questions
Are VLF and DAC results interchangeable?
No. A same-cable field study found that only about 40% of partial-discharge events were seen by both methods, and the two tend to detect different defect populations. Do not treat one result as a substitute for the other.
Which standard governs each method?
VLF is governed by IEEE 400.2-2024. Damped AC / oscillating wave is governed by IEEE 400.4-2015, which is currently Inactive – Reserved with a revision in progress, so confirm its status before citing it.
So which one should I choose?
Follow the question you are asking. VLF for routine acceptance and maintenance and for a continuous monitored withstand; DAC where near-service partial-discharge behaviour and defect localisation drive the decision.
