Research On The Defect Detect System Of ABS On HDD Slider

Along with the fast development of information technology, market demandson data storage is becoming more and more rigorous. HDD(hard disk driver), as themain data storage medium, has a big increase in storage space and read/write speedin recent years. All those are required to be more rigorous in HDD production.MR(Magnetoresistive) head is one of the most important parts in mechanicalhard disk driver. It also contains the most advanced technologies in hard disk driver.MR heads'ABS(air bearing surface) is the contact surface of MR heads and thesurface on magnetic disk. So the quality of MR heads'air bearing surface is veryimportant for its direct influence on the read and write functions of HDD.In this thesis a defect detect system is established to evalue the condition ofMR heads ABS and improve the quality and production rate of MR heads. First ofall, machine vision defect detect system of ABS on MR heads is analyzed. A systemsolution is given. Secondly, the defect detect process is classified into three phases:acquiring of the images of MR head ABS, the adjusting of the MR heads ABSimage and recognizing of the defined defects.For the MR head is very small and delicate part, the clear image of ABS can beaquired by automatic focus technology. Then a method based on BP neural networkis to be established to classify the characteristic datas.At last, experiments for MR heads ABS defect automatic detection are carriedout. With the defect detect experimental results, the defect can be recognizedaccurately by using the automatic detection system.