7.1.29 SURG - Surge Stress Test
This test may be used to highlight a short-circuit between adjacent turns in a winding. It is applicable to any transformer, but is particularly suitable for transformers with a large number of turns using very fine wire.
For such wire, the enamel coating is very thin, and there is a danger that it will be scratched, giving rise to exposed copper. In some cases, the scratch does not immediately cause a shorted turn, but will leave a weak spot which may eventually fail.
By applying a higher than normal voltage across the winding, any weakness in wire insulation will be encouraged to fail.
Each SURG test can be programmed to consist of a number of impulses. For each impulse, the AT3600 will charge an internal capacitor to the high voltage specified. This stored charge will then be suddenly discharged into the windingunder test, and the resulting transient voltage will be analysed.
The product from the discharge will be a sinusoidal wave with decaying amplitude.
Where ts = The time of releasing the impulse
VP= The peak voltage just after switch-on
At the start of the surge test, the AT3600 performs an initial run to compensate for the effect of the capacitance of the transformer winding. Without this compensation, the peak voltage would be reduced by charge sharing between the winding capacitance and the reservoir capacitance within the AT3600, and would not be the value you require.
Therefore, the full test sequence is a follows:
||The value of VP is measured, and the starting conditions
changed to compensate for the charge sharing effects.
||The value of VP is again checked. If it is as specified for the
test, this becomes the first impulse of the sequence, and the
transient is analysed. (If not it is treated as a second
preliminary impulse, and impulse #1 is repeated.)
...: Repeated impulses, up to the number programmed for the
||The value of VP is re-checked, and the transient analysed on
each impulse of the sequence.
During the decay phase after the impulse has been fired, the AT3600 measures both the voltage amplitude along the transient, and the time of decay.
A good transformer will have a clean and sustained transient, with a long decay period. A transformer with a shorted turn will have a heavily damped response, with a shorter decay period.
The calculation performed is to calculate the ‘area’ underneath the graphical plot of the decaying transient. (For the calculation used, both negative peaks and positive peaks add to the total area.) The area, measured in Volts-seconds, is much smaller for the faulty winding with a shorted turn.
SPECIFYING THE TEST LIMITS
It is very difficult to predict the Volts-seconds ‘area’ under the curve from theoretical calculations.
The recommended method is to use the Measure Mode (see Chapters 3 & 6) to obtain some values. The procedure is as follows:
Measure the area on a known good transformer; let this result be area AG.
Wrap an additional single turn round the core, short the two ends together, and re-measure the area; let this result be area AF.
Set the limits as follows:
Max Area = 3AG/2
Min Area = (AG + AF)/2
Remember that these limits are taken from only one transformer, and may need to be revised after more have been tested.