Welcome to the Reads Electric Technical Notes Page.

 Home   Contact Us   About Reads   Services   Facilities   Quality   Representation   Associations   Ayrodev

High Voltage Surge Comparison Testing is the most effective test available to check the quality of electrical windings. Two high voltage pulses are discharged into two identical windings and the reflections of both signals are displayed on a cathode ray tube (CRT).

Good Winding

If the windings are identical and without faults, the reflected pulses will be identical and overlap each other.

If one or both of the windings are faulty, then two dissimilar traces will appear on the CRT.

Surge Testing is an important and effective test because it not only tests ground insulation, but more importantly it also tests for the following:

Other faults indicated include Coil to Coil and Phase to Phase Shorts. As a result, if as few as one turn is shorted the surge test will detect this fault. Most winding failures start with a turn-to-turn short. Circulating currents in that shorted circuit generate excess heat which eventually deteriorates the ground insulation. High Voltage Surge  Comparison Testing can detect faults before they shut down the motor. This test can apply to AC and DC motors, wound rotors, armatures and fields.

Surge Testing is most effective on shorted turns and coils. DC High Potential Testing is most effective locating ground faults. In the DC High Potential mode, the test voltage is displayed on the left side of the CRT. Leakage current is displayed on the right side of the CRT. The combination of both High Voltage Surge Comparison Testing and DC High Potential

This shows a good pattern with high voltage and relatively low leakage current.

This indicates a faulted winding where the leakage current is high and the test voltage is low.

Tan Delta

In view of the increasing interest in these aspects of insulation assessment, both within the Ayrodev organization and amongst their clients, the following notes will serve to illustrate the basis of these tests, the basic theory, and their application to our control methods. In the Ayrodev control system, using the Marconi Bridge, we regard the winding as a capacitor, the electrodes being on one side the winding, and the other frame, with the insulation structure, with or without the varnish, as the dielectric. The chemical changes taking place during the process are reflected in changes in the electrical parameters, and it is these which are monitored. Diagrammatically, a winding could be shown. thus:

If a d.c. voltage is applied across A and B, there is an initial surge as the condenser charges up and then the current stabilizes, at a value inversely proportional to the value of’ R. This is the basis of the Megger method of testing. If an a.c. voltage is applied, however, another factor comes into operation, due to losses in the ddielectric structure.

If C were a perfect capacitor, with no losses, the current would lead the voltage by 90ø, i.e., there would be no phase displacement, and the current flowing through the circuit would be due to the resistive component R. only, and would be equal to the d.c. current. However, we are dealing now with a.c. and the capacitor first charges up as before, on the positive half cycle, discharges, and then becomes charged again on the negative half-cycle. As no capacitor is perfect, however, some of the energy is absorbed into the dielectric, resulting in a phase displacement. The absorbed energy is dissipated in the form of heat within the dielectric, and it will be evident that the higher the energy absorption, the greater will be the phase displacement angle, as an indication of the dielectric absorption. If this is low, we have a good dielectric. If it is high, we have a poor dielectric, independent of R. In other words, it is possible to have a poor insulation for a.c. conditions, which may show a good d.c. insulation on a Megger. The Vector diagram will make this clear.

The ideal capacitor shows 90ø displacement between Ic and V, but as the absorption increases, the angle between Ic and V decreases from the phase displacement and angle increases. The significance of tan delta, however, is mainly applicable to high voltage machines and too much importance should not be attached to it in the case of normal medium

voltage impregnated machines, except as a general indication of the quality of the insulation structure and the stability or otherwise of a winding in process

Polarization Index (PI)

The test is again intended to assess the condition of the insulation structure of large high voltage machines, and is not basically applicable to the average run of medium voltage windings. However, it has been found in many instances that an Ayrodev Process III rewind has developed such a high value of insulation resistance that a P.I. has been obtainable and becomes relevant again as an indication of quality. The method is to measure the insular on resistance at 500 volts, keeping the winding at this pressure for 10 minutes. Polarization Index is defined as:

R. at 10 Min Divided By R. at 1 Min

For insulation values of the order up to 100 megohms or so, and machines up to about 300 hp, there is little change, resulting in a P.I. of 1. With a large machine, particularly with a very high insulation resistance, however, there will be a gradual creep upwards of insulation resistance and this is what is measured. The reason is as follows:

If we revert to the original capacitor diagram under d.c. conditions, the initial charge surge was mentioned, and for reasons of simplicity a single capacitor / resistor combination was shown. In practice however, a machine represents a complex network of capacitors and resistances, which can be shown diagrammatically thus:

The main slot insulation may be regarded as C1 and the basic insulation resistance R1. The other capacitors and resistances are representative of stray capacitances at the end turns, inter-phase insulation, etc., and the resistive components due to various minute surface leakage paths. If the value of C is high, as in a large machine, and R in the  thousand-plus  megohm range,  then  a

time constant factor appears, as C2 will not charge until C1 is charged, and similarly the others along the chain will not charge until their predecessors are charged. The higher the capacity or insulation resistance, the longer this will take. The 1and 10 minute intervals have been internationally agreed as giving a reasonable and practical interval although there have been machines in the 80,000-megohm region which have continued to improve for five hours. Again it cannot be stressed too strongly that this test is primarily intended for large mica-insulated machines, where a P.I. of three is considered to be good, but it is a useful check on quality, as the fact that the I.R. is high enough to give a P.I. reading is an indication of a high-quality insulation structure, whereas on a normal machine, the fact that the P. I. is 1 should not condemn it provided the I.R. is adequate.

With three processes that can dramatically revitalize machines at varying stages of breakdown, Ayrodev might have called it a day. But there still remained on site testing, using the conventional Megga test as a rough guide to which motor is in trouble. In fact the Ayrodev engineer may not take out the motor with the lowest IR.  We have found that it might be the machine with the highest I.R that is ready to fail, because the tan delta was high but was not being measured. With preventative maintenance of this order of accuracy Ayrodev lets you plan ahead.

Two years ago we started testing machines at Tate and Lyle with Process Two and managed to get 100% recovery on 25-year-old machines. The treated machines were shown up later as those with high I.R and low tan delta. No other test equipment can predict potential insulation failures as fast and as cheaply as Ayrodev. So preventative maintenance can eliminate the causes of electrical failure and cut routine servicing. The tests do not involve major dismantling and the average test time is under an hour.

For more information on our Ayrodev Process go to:

The Main Ayrodev Page