Content area
Full text
Familiar to any electrical engineer 50 years ago, today it's seldom encountered
IT WAS A CLEVERLY PLOTTED SET OF LINES, arcs, and angles graphically depicting induction motor performance, the so-called "circle diagram," derived from simple no-load and locked-rotor tests.
Only rarely does a modern textbook even mention the circle diagram. But the basis of such a diagram, why it was used, and why it is no longer used, should be of interest to anyone concerned with motor behavior.
Winding temperature at full load is the most difficult characteristic to estimate without actually loading the machine (see "Evaluating the motor without a load," EA February, 1998). The implication is that other performance, including efficiency, may be predicted reasonably well without load testing. That's true. Verifying such an estimate, however, can require an exacting test procedure.
Years ago, testing was much less rigorous. Until the 1930s, no specific allowance was made for stray load loss. A motor's "conventional" efficiency was determined based on four losses only: stator and rotor I2R, core loss, and friction and windage. Each of them could be evaluated with acceptable accuracy from tests without shaft load, through use of a circle diagram.
One textbook, Elements of Electrical Engineering, now out of print, puts it this way: "Although the circle diagram. . . is not exactly correct, it gives results satisfactory for the usual requirements of engineering accuracy.... The results for motors larger than 10 hp are quite satisfactory, and it is in the testing of large motors that the circle diagram becomes most useful because of the difficulty of testing these motors at full load."
However, as the so-called "wire-to-water" efficiency of pumping systems became more important, and of course now when precise efficiency evaluation is essential for all sizes and types of machines, ignoring stray load loss has become impermissible. Unfortunately, that loss cannot be arrived at from simple unloaded tests. Nor can it readily be incorporated into a circle diagram. The various methods of synthetic load testing generally presume that total motor power input when rated current flows will equal rated output plus all five losses.
That isn't necessarily true, for two reasons: "Rated" or nameplate current does not always correspond exactly to full-load horsepower output, and, depending upon...