The "armette test" for evaluating universal motor insulation systems

The reliability of electrical insulation in a high speed universal motor has for years been of concern to motor manufacturers and their insulation suppliers. Of main concern is lack of correlation between insulation suppliers´ test and the results obtained when evaluated in an actual universal motor. The motor manufacturers have over the years developed test fixtures and methods similar to the "Skil Torture Test" ∗ as outlined in the paper presented at 1962 Electrical Insulation Conference ¹ and in Insulation ² magazine of February 1964. Other motor producers refer to this type of test as "Stall Testing" or "Lock Rotor Testing." This type of testing involves a cycle which includes a locked or stalled rotor heating period, a no load, full speed period, and an off period. The number of cycles to failure indicates a value level for the insulation being tested. This type of testing, however, has two main drawbacks. First, there are a number of variables which can produce unrelated early failures and second, the cost and availability of test units, especially for the insulation manufacturer, makes it highly unattractive as a standard test procedure. Among the variables producing erratic results are field failures, shorts or opens, variations in no load speeds when the voltage is held constant, and variations in stall currents due to variations between fields.