- Output pulse voltage
- 300 V to 3000 V ±5% in 50 V steps on TH2882A-3; 500 V to 5000 V ±5% in 100 V steps on TH2882A-5; the model list gives the same 500 V to 5000 V for TH2882AS-5
- Pulse energy
- ≤90 millijoules on TH2882A-3; ≤250 millijoules on TH2882A-5 and, in the model list, on TH2882AS-5
- Inductance range
- ≥10 µH on TH2882A-3; ≥20 µH on TH2882A-5; ≥200 µH on TH2882AS-5
- Waveform sampling
- ≤40 MSPS at 8 bit
- Input impedance
- 10 MΩ resistive divider
- Comparison methods
- Area, area difference, corona discharge and phase difference
- Display
- 320 × 240 dot LCD with a 240 × 200 dot waveform area; a 480 × 272 TFT panel is also listed for TH2882A-3
- Measurement speed
- 5.5 per second with the waveform off and 3.3 with it on, for the two single-phase models; 2.0 and 1.5 per second for TH2882AS-5
- Waveform storage
- 60 reference waveforms inside the instrument and 500 on a USB drive
- Interface
- RS232C, USB HOST and HANDLER as standard, GPIB optional
A turn-to-turn short inside a coil can leave the inductance almost unchanged and the resistance untouched, which is why a winding that passes a resistance and an inductance check can still fail in service. An impulse test looks at the winding as a distributed network and judges the shape of its response.
Product overview
The range contains three instruments that share one method and differ in what they can drive. Two are single-phase machines that step up in pulse voltage and energy; the third is a three-phase instrument for stators and three-phase windings, and it needs a larger inductance before it can test at all.
All three store reference waveforms, compare a measured response against them, and return a pass or fail. The differences that matter in selection are the voltage the pulse reaches, the energy it delivers, and the number of phases the fixture has to contact.
How the impulse winding tester works
The instrument charges a capacitor and discharges it into the winding under test. The winding responds with a damped oscillation whose frequency and decay are set by its inductance, its distributed capacitance and the losses inside it. That response is digitised at high speed and stored as a waveform.
A known-good winding is measured first and its waveform is kept as the reference. Every part tested afterwards is compared against it. A short between two turns reduces the effective number of turns and changes the resonant behaviour of the network, so the two waveforms separate even where a single inductance reading would not.
The comparison can be made in several ways. Area comparison looks at the integral of the difference between the curves; area difference compares the areas under each; phase difference watches where the peaks fall in time; and corona comparison looks for the high-frequency burst that a partial discharge produces. Each method is sensitive to a different fault, which is why the instrument offers all four rather than one.
The voltage control mode matters here. In normal mode the instrument sets the open-circuit voltage, and a low-impedance winding pulls the actual pulse down. In constant mode the output is regulated so that the winding does not change the applied stress, which keeps the test comparable between parts of different impedance.
How to select the right impulse winding tester
Start with the part. A single-phase coil, a solenoid or a small transformer winding belongs on one of the single-phase models; a stator or a three-phase winding needs the three-phase instrument, because the test has to be applied across the phase pairs and compared in the same configuration.
Then set the voltage from the insulation the winding is built to withstand, not from the highest the instrument can produce. A pulse above what the insulation can take will damage good parts, and the damage may not appear until the part is in service.
Then check the energy and the inductance. A large winding needs more energy to reach the test voltage, and the instrument’s inductance floor has to be below the part being tested, or the response is too small to judge. The three models step up in both figures together.
Then decide how the instrument is operated. A bench with a foot switch and a fixture is one pattern; a line that sorts parts needs the handler interface and the pass and fail outputs; a line controlled by a computer needs the serial port, and the bus interface if that is what the controller speaks.
Send the winding drawing, the number of phases, the inductance and the insulation class, and a model, a lead and a fixture can be quoted against them.
Measurement limitations and selection considerations
The test compares a part against a reference, so the reference has to be right. A waveform stored from a marginal winding will pass marginal parts, and a reference taken at a different voltage or in a different fixture is not the same test.
The method is sensitive to the fixture as much as to the winding. Lead length, contact position and the capacitance of the jig are part of the network being measured, so a part that passes on one fixture can fail on another. The reference and the production parts have to be measured the same way.
The pulse is a stress, not a proof. A part that survives the test has survived that pulse; whether it survives the insulation test that follows, or the voltage in service, is a separate question.
Finally, the inductance floor is a real limit. Below the minimum the response is dominated by the fixture and the comparison loses meaning, which is why a small winding may need a different instrument rather than a different setting.
Every value below is taken from source documentation. Where a parameter is not stated in the available material it is marked as information required rather than estimated.
