A Self-Calibrating, Erroneous Measurement-Preventing Production Test System for Military Slip Rings

Abstract

Military Slip Rings have tight test requirements to meet; such as contact noise, dielectric discharge and end-to-end resistance. Especially, the contact noise and end-to-end resistance measurements are in the order of several milliohms. Therefore, the test system must be well-calibrated in order to take precise measurements. During mass production, more than one set of test equipment can be used on different manufacturing lines. Hence, calibration should be able to be done on-line by the test software. Additionally, the test process must take as minimum time as possible in order to meet mass production schedule. Erroneous measurements due to external electromagnetic noise is a serious agent that cause the UUT (unit under test) to fail the test, prolonging the overall testing time. The testing software should detect an erroneous measurement, dispel it and prevent it from failing the test of the UUT. In this paper, we explain the methods used for a self-calibrating, automated, erroneous measurement-preventing test system for the mass production of military slip rings. When testing a slip ring, some of the measurements such as contact noise should be measured while the slip ring is rotating. Therefore, a turntable with servo motor is used for rotation. While the slip ring is rotating, it is not feasible to take measurements from the connectors on the both sides (stator & rotor) since it can cause testing cables to be tangled and eventually jam the turntable, possibly harming the slip ring. Hence, in this setup all the test equipment are connected to the slip ring from one side (stator or rotor), while at the other side all the connectors are short-circuited. These connections add to the end-to-end resistance and contact noise while measuring them. Since the precision of the measurements should be in the order of milliohm, the effect of the connections change not only with the test setup, but also with the slip ring itself. Therefore, testing software initially measures the resistances of connections, updates its' calibration values and then starts the test procedure. The details of the algorithm as well as test results with and without using this method are presented. Another phenomena while taking measurements is the electromagnetic noise that manipulates the test results. Even if the test cables are twisted and shielded, this additional noise is hard to avoid due to the test equipment itself. It causes sudden peaks in the measurements and causes the test of that channel to fail. One way of preventing this is to measure the rms or the standard deviation of the noise instead of its peak-to-peak value. However, it gives excessively optimistic results which may cause some problematic channels to go unnoticed. The method used in this setup prevents the test process to fail due to erroneous measurement, while taking objective measurements. Statistical information about the efficiency of using this method is exhibited.