Thermal accelerometer based predictive drop sensor

Abstract

The thermal accelerometer measures internal changes in heat transfer caused by acceleration. It has been the most reliable sensor employed for hard disk drive (HDD) drop protection system because this kind of device is capable to survive high shocks. The frequency response is limited to 17Hz, however, so that protection functions only when the drop distance exceeds 17cm, considering the HDD park time being 70ms. At this drop height, the impact force shock of up to 246.6g could happen without any sensing signal which will trigger the protection function into operation. The HDD could thus be damaged since the allowable shock g rating in operation mode is only 200g/1ms. Based on the transient response modeling of the thermal accelerometer, a predictive filter was developed to predict the drop by proprietary algorithms. The predictive drop sensor, composed of a commercial thermal accelerometer and micro controller chip, can achieve fast response as it alarms impending fall and enables the protection function before real drop occurs. By the optimization of the filter weighting coefficients, accurate prediction of the impending drop event has been shown in this study. Predictions of up to 35ms have been demonstrated and that can provide reliable protection against drop shock.