BGA_OPEN_ISSUE

1、FAILURE ANALYSIS REPORT P4B-MX BGA OPEN (一一) 不良現(xiàn)象不良現(xiàn)象3月25日,P4B-MX R1.05 ,F/B測(cè)試不良31pcs,顯示不良現(xiàn)象:(1)FF無(wú)法開機(jī)(2)80卡顯示異常(3)上機(jī)有異音(4)D5無(wú)法開機(jī)(5)D1無(wú)法開機(jī)(6)WINDOWS當(dāng)機(jī)結(jié)論:此上述不良現(xiàn)象與CPU socket478P及北橋相關(guān),故對(duì) CPU socket及北橋作分析. (二)工程分析11.此上述不良現(xiàn)象訊號(hào)線與北橋及CPU socket 478P相關(guān),因北橋訊號(hào)點(diǎn)位難以測(cè)量,故先測(cè)量CPU socket與GND之相連點(diǎn)位(主要為HA,HD信號(hào)點(diǎn))的保護(hù)二極體是否

2、與正常PCB之相同點(diǎn)位顯示數(shù)值有不同.工程針對(duì)不良PCBA做驗(yàn)證,發(fā)現(xiàn)有的訊號(hào)點(diǎn)位測(cè)試值比正常PCBA顯示數(shù)值偏大.結(jié)論:故判定CPU socket或北橋有本體不良或焊點(diǎn)OPEN狀況. (二)工程分析22.選取不良PCBA 北橋周圍線路位置點(diǎn),量測(cè)線路與CPU socket 478P相連的點(diǎn)位保護(hù)二極體是否有異常.量測(cè)結(jié)果無(wú)異常,表明CPU socket 478P本體無(wú)不良且焊接點(diǎn)無(wú)OPEN狀況,故排除CPU socket 478P不良. 結(jié)論:不良的產(chǎn)生是因?yàn)楸睒虮倔w不良或北橋焊點(diǎn)有OPEN狀況. (二)工程分析 3 3.對(duì)不良PCBA做如下分析:(1)首先使用X-RAY觀察焊點(diǎn),未觀察到異

3、常;(2)再用手下壓BGA本體進(jìn)行測(cè)試,F/B測(cè)試PASS;(3)對(duì)不良PCBA之BGA進(jìn)行重熔,測(cè)試PASS;(4)將不良PCBA之BGA用BGA rework設(shè)備拆除,對(duì)拆除的BGA本體進(jìn)行量測(cè),其量測(cè)值正常.結(jié)論: 此不良的產(chǎn)生是因BGA 焊點(diǎn)OPEN引起的. BGA區(qū)域信號(hào)點(diǎn)分布1代表空焊點(diǎn)位 BGA區(qū)域信號(hào)點(diǎn)分布2 INTEL 82845BGA點(diǎn)陣圖代表空焊點(diǎn) (三) 切片過(guò)程1一.切割: 1 操作流程：準(zhǔn)備工作調(diào)整各定位機(jī)構(gòu)開啟電源切割關(guān)機(jī)清理1.1 準(zhǔn)備工作 (1) 查看臺(tái)面是否整潔，切刀附近有無(wú)雜物，如有，即時(shí)清理。 (2) 查看機(jī)內(nèi)是否切削過(guò)多，妨礙機(jī)器正常運(yùn)轉(zhuǎn)，如有，即時(shí)清

4、理。(3) 將擋削蓋壓下，以防止切削飛出。 (三)切片過(guò)程21.2 調(diào)整各定位機(jī)構(gòu)(1) 根據(jù)要求板材切割位置，對(duì)照機(jī)臺(tái)刻度標(biāo)尺，確定橫向定位桿位置。擰松橫向定位桿之定位螺釘，將其移動(dòng)至要求位置，擰緊螺釘。(2) 如要求斜向切割，需先參照角度定位桿之圓角刻度，將角度定位桿調(diào)整至要求角度，擰緊定位螺釘。 (三)切片過(guò)程 31.3 開啟電源：接上電源，將啟動(dòng)開關(guān)按至“ON”檔，開啟切割轉(zhuǎn)刀。1.4 切割：將板子放于操作平臺(tái)，邊緣緊貼定位桿，雙手壓住板子，並向前送至切刀進(jìn)行切割。1.5 關(guān)機(jī)：切割完成後，取回板子，將啟動(dòng)開關(guān)按至“OFF”檔，關(guān)閉機(jī)器。 (三) 切片過(guò)程 41.6 清理 (1) 清理

5、機(jī)臺(tái)上剩餘切削，使其保持清潔 (2) 每星期五下班前擰開機(jī)臺(tái)蓋上螺絲， 打開機(jī)臺(tái)蓋，清理機(jī)器內(nèi)削渣，以保 証機(jī)器正常運(yùn)轉(zhuǎn)。 2 注意事項(xiàng)：安全操作，避免操作時(shí)，身體與切刀接觸。 (三) 切片過(guò)程5二:研磨 1 開啟電源開關(guān) 2 旋轉(zhuǎn)轉(zhuǎn)速開關(guān)，使其達(dá)到研磨時(shí)的速度，并調(diào)節(jié)水流速度。 3 粗磨：將壓克力模在200目的砂紙上研磨，若是磨線路，則只要磨平就可，若是磨孔則要磨到孔邊。4 中磨：將壓克力模在600目的砂紙上研磨，若是磨線路，則只要磨平就可，若是磨孔則要磨到接近孔半徑位置。 (三)切片過(guò)程 6 5 細(xì)磨：將壓克力模在1200目的砂紙上研磨，若是磨線則要磨光，若要磨孔，則磨至半圓，同時(shí)也使其磨

6、光滑. 6 拋光。 7 在拋光紙?jiān)诜派弦恍伖夥郏淖饔檬鞘贡砻孢_(dá)到光滑，便于后繼顯微鏡。 (三)切片分析1不良焊點(diǎn)正常焊點(diǎn)結(jié)論: 不良焊點(diǎn)狀況:PCB之PAD與BGA之錫球分離 正常焊點(diǎn)狀況:PCB之PAD與BGA之錫球與錫膏熔為一體. 不良焊點(diǎn)正常焊點(diǎn)不良焊點(diǎn)與正常焊點(diǎn)放大後比較 (三)切片分析2 (四)BGA不加熱直接拆除分析空焊之PCB PAD空焊之BGA PAD (五) PCB廠商分析1.針對(duì)BGA不加熱直接拆除之PCBA,請(qǐng)統(tǒng)盟用EDS元素分析儀進(jìn)行分析:(1) 拒焊本身由EDS元素分析結(jié)果，錫墊表面除錫&鉛成份物質(zhì)外並無(wú)其它非金屬性拒焊雜質(zhì)成份在，造成因素排除與污染物質(zhì)有關(guān)。(

7、2) 空焊異常PCB生產(chǎn)製作週期橫跨3個(gè)週期、再加上空焊問題不良點(diǎn)僅集中發(fā)生於北橋固定區(qū)域位置，因PCB板本身是採(cǎi)整片浸泡噴錫的作法，空焊不良並不會(huì)集中固定發(fā)生於相同區(qū)域。結(jié)論:拒焊發(fā)生原因點(diǎn)並非於PCB本身 . (六) BGA廠商分析1Provide Unit without SMT and X-section unit for vendor to analysis: (1)Unit without SMTnNo any other element can be detected except for flux (C,O) and NaCl; nFlux residue due to non

8、-clean process during ball attach and this would impact solder ability;nNaCl may came from sweet hand; (六)BGA 廠商分析2 (2)X-section unitnPoor wetability between one solder sphere and motherboard pad, which resulted in void around pad edge area due to no eutectic solder bulk filled in during SMT process

9、;nStress area was formed due to void and which initiated the crack line along the interface, this is the reason of open;nElements comparison between non-wet interface and good interface shows they are equivalent;nThe Al element may be from fine polishing by chemicals. So it is not clear for the form

10、ation of the non-wet;nNo other element was detected on IMC layer, suspect that there is no other material on top of Cu layer. Based on Intel experience, the IMC formed by Cu/Sn/Pb may not have good joint strength. Dark area due to flux residue.NaCl may de due to touch with hand.Sn, Pb2Sn, C, O, Na,

11、Cl, Pb1 elementSpot12Detected Ball location EDX On Raw Part Solder Ball Surface-1 Detected Ball location Sn, C, O, Na, Cl, Pb, K2Sn, C, O, Na, Cl, Pb, K1 elementSpotDark area due to flux residue.NaCl may de due to touch with hand.EDX On Raw Part Solder Ball Surface-2 FESEM On Non-wet BallCrackvoidvo

12、idcrackcrackPb/SnCumotherboardCrack can be seen between solder ball/motherboard.There is void around pad edge due to no eutectic solder.EDX result refer to page 10.The interface layer (IMC) thickness is around 1.53um averagely. Left edge non-wet solder Ball right Edge ; 1 23 4Sn, C, N, O4Sn, Pb, C,

13、O, N3Cu, Sn, Al, C, O2Cu, Sn, C, O, N1 elementSpotAl may due to chemical when fine polishSuppose the C, O, N, from Flux?EDX On Non-wet Solder Ball Interface FESEM On Good Ball No Crack can be seen between solder ball/motherboard.There is void on solder ball bulk. EDX results refer to page 12.The int

14、erface layer (IMC) thickness is around 2.84um averagely.Cu 1 23 4Sn, C, N, Cu4Sn, Pb, C, O, N3Cu, Sn, Al, C, O, Pb, N2Cu, Sn, C, O, N1 elementSpotAl may due to chemical when fine polishSuppose the C, O, N, from Flux?EDX On Good Solder Ball Interface (七) 改善前迴焊爐曲線 (七)改善後迴焊爐曲線 (七)改善前與改善後的差異 改善前 改善後1.FI

15、RST RAMP-UP :1.3/SEC 1.FIRST RAMP-UP :1.2/SEC 2.SOAK TIME :75.2SEC 2.SOAK TIME :64SEC3.PEAK TEMP : 3.PEAK TEMP : (1)BGA (BIG) TOP TEMP :213 (1)BGA (BIG) TOP TEMP :217.5 (2)BGA (BIG) BOT TEMP :210.5 (2)BGA (BIG) BOT TEMP :216.5 (3)BGA (LITTLE) TOP TEMP :214.5 (3)BGA (LITTLE) TOP TEMP :216 (4)BGA (LITTLE) BOT TEMP :213.5 (4)BGA (LITTLE) BOT TEMP :215.5 (5)QFP 128 PIN TEMP :214 (5)QFP 128 PIN TEMP :221 (6)CPU SOCKET TEMP :208 (6)CPU SOCKET TEMP :211.54.OVER 183 TIME :73.6SEC 4.OVER 183 TIME :73.6SEC注:迴焊爐曲線調(diào)節(jié)的主要參數(shù)為SOAK TIME由原5090sec,更改為 5065sec,以防止錫膏