21 May 2010
I continued my quest in producing a comprehensive report on the feasibility of using upbent lead frames for SOT363F products. In the mean time, I was given a side chore to complete which was to update the Product Master Data (PMD).
24 May 2010
I had completed my project report on the feasibility of using upbent lead frames for SOT363F products. It is 58 pages long. I know quantity is not always commensurable with quality. Nonetheless, I would endeavor to debunk that as I had placed profound effort in conjuring up this report. Through this time-consuming process, I had assiduously undergone theoretical evaluation of the problem at hand; acquire actual values on actual products and finally simulating obtained data using software called Minitab. Besides that, I had enhanced my skills of writing an engineering report as well as significantly improved my Microsoft Word and Excel literacy.
In addition to project report, I spent the rest of the day updating Product Master Data (PMD) as well as detecting and correcting errors found in PMD updated by colleagues.
25 May 2010
I started off the day updating Product Master Data (PMD), before proceeding to take quite a number of pictures on decapped products, both top view and side view. Moreover, I also took pictures on die pushed products.
26 May 2010
I spent the whole day gaining cognizance about Macros, which is one of the most powerful features in Microsoft Excel. I was given a book to read. The objective of using Macros is to ease our works and shorten task completion time. I was asked to create Macros akin to the project that I had just completed so that whenever there is another pristine set of parameters to be evaluated, one does not need to come through all those tedious processes like calculations, actual value acquisition and Minitab simulations, to have a rough overview on the feasibility of implementing that set of parameters.
27 May 2010
I proceeded with my effort to engender Macros. I had decided to use divide and conquer strategy in coming up with usable Macros by dividing the whole project into noticeable parts and writing Macros for each one.
Besides that, I was requested by Miss Manggala to aid her in completing her engineering report. To accomplish that, I was required to use JMP software to simulate over five thousand values obtained from the back end testing, in order to produce multiple relevant graphs impregnating process stability. I was also required to tag along Puan Norham who was in charge of performing manual tests on failed products to authenticate that the products were indeed flawed, and were not due to machine malfunctioning.
Thursday, May 27, 2010
Thursday, May 20, 2010
17 - 20 May 2010
17 May 2010
I continued my exhilarating odyssey into the world of calculation. Today, I finished calculating the theoretical distance between capillary and the die under ideal die placement as well as under worst case scenario. By ideal die placement, the die is perfectly situated at the centre of the major flag. On the other hand, worst case scenario infers that the die is displaced 20um.
Next, I calculated the distance between die edge and wire, stretching from the line of ball bond on the die. Again, two situations are considered, that are ideal die placement as well as worst case scenario.
I finished off the day by evaluating the displacement area of wedge bond on minor lead. I took the center placement of the wedge, which is the ideal case as a reference point. Subsequently, I modeled each displacement as a circle centered about that reference point.
All calculations will be shown.
18 May 2010
I spent the day tabulating and simulating those data I had obtained from the assembly line into Minitab. This is a software that can depict many parameters akin to any process under inspection. With the helpful Miss June beside me, I was able to simulate, understand and encapsulate the graphs as well as information generated. To put it compendiously, the processes that I had been involved in are stable.
Besides Minitab, I also calculated the maximum distance that the die can deviate vertically before hitting the capillary. Again, all calculations will be shown.
19 May 2010
I had basically finish all the data acquisition as well as theoretical calculations relevant to my project. I was asked to detect errors in the method used by Miss June in calculating those similar parameters that I had calculated. After some meticulous inspection, I was able to detect some errors and remedied them. I spent the rest of the day gaining knowledge on how to conjure up a report regarding my project. I was given several reading materials to attain adequate information deemed relevant to producing an engineering technical report.
20 May 2010
My agenda today was indubitably to produce a quality report. I gained several skills dealing with graphics in Microsoft Word in my quest to engender a splendid report.
I continued my exhilarating odyssey into the world of calculation. Today, I finished calculating the theoretical distance between capillary and the die under ideal die placement as well as under worst case scenario. By ideal die placement, the die is perfectly situated at the centre of the major flag. On the other hand, worst case scenario infers that the die is displaced 20um.
Next, I calculated the distance between die edge and wire, stretching from the line of ball bond on the die. Again, two situations are considered, that are ideal die placement as well as worst case scenario.
I finished off the day by evaluating the displacement area of wedge bond on minor lead. I took the center placement of the wedge, which is the ideal case as a reference point. Subsequently, I modeled each displacement as a circle centered about that reference point.
All calculations will be shown.
18 May 2010
I spent the day tabulating and simulating those data I had obtained from the assembly line into Minitab. This is a software that can depict many parameters akin to any process under inspection. With the helpful Miss June beside me, I was able to simulate, understand and encapsulate the graphs as well as information generated. To put it compendiously, the processes that I had been involved in are stable.
Besides Minitab, I also calculated the maximum distance that the die can deviate vertically before hitting the capillary. Again, all calculations will be shown.
19 May 2010
I had basically finish all the data acquisition as well as theoretical calculations relevant to my project. I was asked to detect errors in the method used by Miss June in calculating those similar parameters that I had calculated. After some meticulous inspection, I was able to detect some errors and remedied them. I spent the rest of the day gaining knowledge on how to conjure up a report regarding my project. I was given several reading materials to attain adequate information deemed relevant to producing an engineering technical report.
20 May 2010
My agenda today was indubitably to produce a quality report. I gained several skills dealing with graphics in Microsoft Word in my quest to engender a splendid report.
17 - 20 May 2010
17 May 2010
I continued my exhilarating odyssey into the world of calculation. Today, I finished calculating the theoretical distance between capillary and the die under ideal die placement as well as under worst case scenario. By ideal die placement, the die is perfectly situated at the centre of the major flag. On the other hand, worst case scenario infers that the die is displaced 20um.
Next, I calculated the distance between die edge and wire, stretching from the line of ball bond on the die. Again, two situations are considered, that are ideal die placement as well as worst case scenario.
I finished off the day by evaluating the displacement area of wedge bond on minor lead. I took the center placement of the wedge, which is the ideal case as a reference point. Subsequently, I modeled each displacement as a circle centered about that reference point.
All calculations will be shown.
18 May 2010
I spent the day tabulating and simulating those data I had obtained from the assembly line into Minitab. This is a software that can depict many parameters akin to any process under inspection. With the helpful Miss June beside me, I was able to simulate, understand and encapsulate the graphs as well as information generated. To put it compendiously, the processes that I had been involved in are stable.
Besides Minitab, I also calculated the maximum distance that the die can deviate vertically before hitting the capillary. Again, all calculations will be shown.
19 May 2010
I had basically finish all the data acquisition as well as theoretical calculations relevant to my project. I was asked to detect errors in the method used by Miss June in calculating those similar parameters that I had calculated. After some meticulous inspection, I was able to detect some errors and remedied them. I spent the rest of the day gaining knowledge on how to conjure up a report regarding my project. I was given several reading materials to attain adequate information deemed relevant to producing an engineering technical report.
20 May 2010
My agenda today was indubitably to produce a quality report. I gained several skills dealing with graphics in Microsoft Word in my quest to engender a splendid report.
I continued my exhilarating odyssey into the world of calculation. Today, I finished calculating the theoretical distance between capillary and the die under ideal die placement as well as under worst case scenario. By ideal die placement, the die is perfectly situated at the centre of the major flag. On the other hand, worst case scenario infers that the die is displaced 20um.
Next, I calculated the distance between die edge and wire, stretching from the line of ball bond on the die. Again, two situations are considered, that are ideal die placement as well as worst case scenario.
I finished off the day by evaluating the displacement area of wedge bond on minor lead. I took the center placement of the wedge, which is the ideal case as a reference point. Subsequently, I modeled each displacement as a circle centered about that reference point.
All calculations will be shown.
18 May 2010
I spent the day tabulating and simulating those data I had obtained from the assembly line into Minitab. This is a software that can depict many parameters akin to any process under inspection. With the helpful Miss June beside me, I was able to simulate, understand and encapsulate the graphs as well as information generated. To put it compendiously, the processes that I had been involved in are stable.
Besides Minitab, I also calculated the maximum distance that the die can deviate vertically before hitting the capillary. Again, all calculations will be shown.
19 May 2010
I had basically finish all the data acquisition as well as theoretical calculations relevant to my project. I was asked to detect errors in the method used by Miss June in calculating those similar parameters that I had calculated. After some meticulous inspection, I was able to detect some errors and remedied them. I spent the rest of the day gaining knowledge on how to conjure up a report regarding my project. I was given several reading materials to attain adequate information deemed relevant to producing an engineering technical report.
20 May 2010
My agenda today was indubitably to produce a quality report. I gained several skills dealing with graphics in Microsoft Word in my quest to engender a splendid report.
Friday, May 14, 2010
11-14 May 2010
11 May 2010
Today, I gained a better understanding of the operation as well as errors akin to ADAT and PHICOM machines. First off, I observed how a technician attended to “Post Bond Total Skip Count Exceeded” error on an ADAT machine. Initial comprehension was that presence of tilted dies were detected, which rendered the PRS image darker than usual. The technician once thought it was just sporadic occurrence; however, subsequent series of same errors faulted his initial judgment. He then discovered that the push up needle was rickety and promptly replaced it with a new one. Push up needle serves the purpose of pushing up the designated die to be picked up by the collet.
Next, still within the vicinity of ADAT machines, I came across “Touchdown Push Up Too Soon” error. This means that the push up needle elevated the die too soon for the collet to pick up. The operator attended the problem by changing the process delay value.
Mr. Leong provided me with several additions to my knowledge of ADAT as well as die bonding process. He taught me that there the first die attach machine was invented by the Germans. Predictably, it was slow. Then, we have ADAT1, ADAT2 and ADAT3. For ADAT2, there are 3 different types: ADAT2 12k, ADAT2 14k, ADAT2 18k, with the last type being the fastest among the three. So, the notation of 12k, 14k and 18k simply impregnates the idea of different speed. Apart from that, Mr. Leong also talked about what SOT and SOD mean. SOT is the acronym for Small Outline Transistor, while SOD means Small Outline Diode. Furthermore, the helpful Mr. Leong told me that RF (Radio Frequency) products are relatively costlier as more convoluted process is implicated in which wafer back coat is required. Wafer back coating is similar to glue bonding.
Next on the agenda are errors associated with PHICOM wire bonding machines. When one encounters “Total Skipcount Exceeded” error, it probably indicates that the predetermined allowable number of units that are devoid of dies or “Total Skip Setting” value was reached or exceeded. There is always a “Skip Counter” indicator beside the “Total Skip Setting” to tell us about how many units have been skipped by the machine. Another setting worthy of mentioning is “Sequential Skip Setting” which is similar to “Total Skip Setting” except that it indicates the number of acceptable consecutive defective units to be skipped. Besides that, when the error “Spark Not Completed” pops out, it means that spark is not provided for wire bonding process, culminating in the failure to form Free Air Ball (FAB).
In addition, I also tried my hands on die shear. Previously, I had been performing ball shear. After several runs with ball shear, I can deem myself as an expert in ball shear. However, the same cannot be said to die shear. The first thought that came to mind was that it would just be a piece of cake. It turned out to be an exhilarating experience though it was tremendously difficult. In the process, I managed to obliterate many dies. Poor dies, rest in peace…
12 May 2010
Today was really a humdinger as I was assigned something totally different and challenging. Before I go into details of what I was asked to do, Miss June briefed me about the upcoming production of a new product which is based on upbent lead frames, which have the major as well as minor bond pads elevated slightly. Customarily, lead frames are completely flat. Miss June’s briefing was rather vague and convoluted which culminated in me enveloping in a mist of uncertainty and questions.
Uncertainty notwithstanding, after being provided with relevant information, I proceeded to calculate the length of that bent portion. Subsequently, with that aforementioned value in mind, I calculated the angle of bending. After obtaining those theoretical values, I went into the assembly line where I performed several measurements to acquire some practical values, in order to vindicate my theoretical findings. Fortunately, the values obtained theoretically and practically were coterminous at worst and exactly the same at best.
Next, I moved on to calculate the angle of R35 capillary tip. Before I could attain that, I did some concise research on the parts of the R35 wire bonding capillary, and understanding the parameters or dimensions involved. Based on the Outer Diameter (OD), Length (L), Cone Angle (CA), I was able to determine the outer angle of R35 capillary tip as well as the cone vertical height.
Moving on, I evaluated the remaining space available when a flat lead frame is to be used upbent. With that in mind, and information regarding die size, I also determined the final remaining space available after the die is attached to the bond pad. After that, I went into the assembly line again to measure the major and minor bond pad length.
13 May 2010
I continued where I left off yesterday, performing measurements of major as well as minor flags length. I was able to finish the measurements of three distinct lead frames, which are designated for SOT343F, SOT343R2 and SOT363F. Also, I finished measuring the bending angle and the bending base length of SOT343F and SOT363F lead frames. The same cannot be performed on SOT343R2 lead frames as this type of lead frame is not upbent. After lunch, I continued my unique relationship with the measurement device as I was assigned to perform 360 measurements in order to determine ADAT and PHICOM consistency. All of these are part of the challenging assignment given to me yesterday.
14 May 2010
I tabulated all the data I acquired in the past two days. For the 360 measurements involving horizontal and vertical distances from my designated reference point on the major flag to the die as well as from a designated reference point on the minor lead to the wedge bond, I also calculated the hypotenuse distance in order to determine the consistency of ADAT and PHICOM machines. For the machines to be regarded as consistent, the standard deviation is to be less than 10µm, which is rightly so for both ADAT as well as PHICOM machine.
While waiting for Miss June to assign me with further tasks, I spent my time taking images of die pushed units. Similar to die shear units, the die on each individual unit is removed by force. For die shear products, the die is removed by machine, however, for die push units; the die is removed by tweezers held by human fingers.
After lunch, after lucid explanations from Miss June, I began the intriguing calculation of the possibility of capillary hitting the die. To show this, I needed to calculate the distance between the capillary and the die during wire bonding process. I considered the ideal case, where the die is situated at the middle of the major flag, as well as the worst case scenario, in which the die is displaced for 20µm. This maximum displacement value is calculated by evaluating the process capability index of 4σ.
I will upload all the photos pertaining to the calculations that I had done when I gain internet access again. For your information, my home is without internet for one month. I know, it is agonizing. That is why I could only update my blog so sporadically
Today, I gained a better understanding of the operation as well as errors akin to ADAT and PHICOM machines. First off, I observed how a technician attended to “Post Bond Total Skip Count Exceeded” error on an ADAT machine. Initial comprehension was that presence of tilted dies were detected, which rendered the PRS image darker than usual. The technician once thought it was just sporadic occurrence; however, subsequent series of same errors faulted his initial judgment. He then discovered that the push up needle was rickety and promptly replaced it with a new one. Push up needle serves the purpose of pushing up the designated die to be picked up by the collet.
Next, still within the vicinity of ADAT machines, I came across “Touchdown Push Up Too Soon” error. This means that the push up needle elevated the die too soon for the collet to pick up. The operator attended the problem by changing the process delay value.
Mr. Leong provided me with several additions to my knowledge of ADAT as well as die bonding process. He taught me that there the first die attach machine was invented by the Germans. Predictably, it was slow. Then, we have ADAT1, ADAT2 and ADAT3. For ADAT2, there are 3 different types: ADAT2 12k, ADAT2 14k, ADAT2 18k, with the last type being the fastest among the three. So, the notation of 12k, 14k and 18k simply impregnates the idea of different speed. Apart from that, Mr. Leong also talked about what SOT and SOD mean. SOT is the acronym for Small Outline Transistor, while SOD means Small Outline Diode. Furthermore, the helpful Mr. Leong told me that RF (Radio Frequency) products are relatively costlier as more convoluted process is implicated in which wafer back coat is required. Wafer back coating is similar to glue bonding.
Next on the agenda are errors associated with PHICOM wire bonding machines. When one encounters “Total Skipcount Exceeded” error, it probably indicates that the predetermined allowable number of units that are devoid of dies or “Total Skip Setting” value was reached or exceeded. There is always a “Skip Counter” indicator beside the “Total Skip Setting” to tell us about how many units have been skipped by the machine. Another setting worthy of mentioning is “Sequential Skip Setting” which is similar to “Total Skip Setting” except that it indicates the number of acceptable consecutive defective units to be skipped. Besides that, when the error “Spark Not Completed” pops out, it means that spark is not provided for wire bonding process, culminating in the failure to form Free Air Ball (FAB).
In addition, I also tried my hands on die shear. Previously, I had been performing ball shear. After several runs with ball shear, I can deem myself as an expert in ball shear. However, the same cannot be said to die shear. The first thought that came to mind was that it would just be a piece of cake. It turned out to be an exhilarating experience though it was tremendously difficult. In the process, I managed to obliterate many dies. Poor dies, rest in peace…
12 May 2010
Today was really a humdinger as I was assigned something totally different and challenging. Before I go into details of what I was asked to do, Miss June briefed me about the upcoming production of a new product which is based on upbent lead frames, which have the major as well as minor bond pads elevated slightly. Customarily, lead frames are completely flat. Miss June’s briefing was rather vague and convoluted which culminated in me enveloping in a mist of uncertainty and questions.
Uncertainty notwithstanding, after being provided with relevant information, I proceeded to calculate the length of that bent portion. Subsequently, with that aforementioned value in mind, I calculated the angle of bending. After obtaining those theoretical values, I went into the assembly line where I performed several measurements to acquire some practical values, in order to vindicate my theoretical findings. Fortunately, the values obtained theoretically and practically were coterminous at worst and exactly the same at best.
Next, I moved on to calculate the angle of R35 capillary tip. Before I could attain that, I did some concise research on the parts of the R35 wire bonding capillary, and understanding the parameters or dimensions involved. Based on the Outer Diameter (OD), Length (L), Cone Angle (CA), I was able to determine the outer angle of R35 capillary tip as well as the cone vertical height.
Moving on, I evaluated the remaining space available when a flat lead frame is to be used upbent. With that in mind, and information regarding die size, I also determined the final remaining space available after the die is attached to the bond pad. After that, I went into the assembly line again to measure the major and minor bond pad length.
13 May 2010
I continued where I left off yesterday, performing measurements of major as well as minor flags length. I was able to finish the measurements of three distinct lead frames, which are designated for SOT343F, SOT343R2 and SOT363F. Also, I finished measuring the bending angle and the bending base length of SOT343F and SOT363F lead frames. The same cannot be performed on SOT343R2 lead frames as this type of lead frame is not upbent. After lunch, I continued my unique relationship with the measurement device as I was assigned to perform 360 measurements in order to determine ADAT and PHICOM consistency. All of these are part of the challenging assignment given to me yesterday.
14 May 2010
I tabulated all the data I acquired in the past two days. For the 360 measurements involving horizontal and vertical distances from my designated reference point on the major flag to the die as well as from a designated reference point on the minor lead to the wedge bond, I also calculated the hypotenuse distance in order to determine the consistency of ADAT and PHICOM machines. For the machines to be regarded as consistent, the standard deviation is to be less than 10µm, which is rightly so for both ADAT as well as PHICOM machine.
While waiting for Miss June to assign me with further tasks, I spent my time taking images of die pushed units. Similar to die shear units, the die on each individual unit is removed by force. For die shear products, the die is removed by machine, however, for die push units; the die is removed by tweezers held by human fingers.
After lunch, after lucid explanations from Miss June, I began the intriguing calculation of the possibility of capillary hitting the die. To show this, I needed to calculate the distance between the capillary and the die during wire bonding process. I considered the ideal case, where the die is situated at the middle of the major flag, as well as the worst case scenario, in which the die is displaced for 20µm. This maximum displacement value is calculated by evaluating the process capability index of 4σ.
I will upload all the photos pertaining to the calculations that I had done when I gain internet access again. For your information, my home is without internet for one month. I know, it is agonizing. That is why I could only update my blog so sporadically
Monday, May 10, 2010
10 May 2010
I started off the day by gaining cognizance about the term “partial wafer”. Partial wafer is exactly the same as ordinary wafer except that only a certain amount of dies from a partial wafer will be used or picked up during production. That amount will be dependent upon customers’ demands. After a designated amount of dies had been picked up, that wafer will be kept for further use.
Besides that, I went to the front end where Mr. Leong is stationed. He was monitoring the production of a new product. An ineluctable predicament surfaced in the form of the absence of numbering of die pick up starting position. To rub salt on wound, some of the dies in propinquity of the edges were not entirely defective. This will pose some impediment to the effectiveness of the Pattern Recognition System (PRS) in identifying good (usable) dies. After numerous attempts, the indefatigable spirit of Mr. Leong prevailed as he successfully estimated a proper starting position by trial and error.
Mr. Leong also taught me that the Pattern Recognition System (PRS) of ADAT2 die bonding machine works on the principle of binary scale. In other words, it discerns patterns based on black and white images of the die. On the other hand, PRS of ADAT3 machine not only works on binary scale, but also grayscale. This renders the PRS of ADAT3 machine significantly efficacious than ADAT2 machine. Nonetheless, higher cost of ADAT3 machine prevents its widespread use.
My learning experience continued after lunch break. After the front end process, the die bonded as well as wire bonded lead frames will be channeled to the mid-end process. Before that transfer takes place, unusual lead frames will be identified using stickers with different colours to inform personnel in mid end. Firstly, red coloured sticker indicates that certain part of the lead frames contain new device or new lot. Within front end itself, lead frames having new device will be noted with pen marking on lead frames to avert or assuage mixed device possibility. Next, green coloured sticker is used when a portion of the lead frames is unwanted scrap of production. Lastly, blue coloured sticker is used to represent lead frames that had been welded. Welded parts may be uneven and therefore it is not advisable to mold them.
During a casual stroll around front end, I came across several customer complaints. A customer from Korea complained about partial lifted ball bond. The cause of this is believed to be floating of lead frames due to copper build-up at heater block. During wire bonding process, as the capillary exerts lateral force on the ball at the bondpad, the force will not be spread evenly. Next, a Dutch company complained about offset ball bond, which means that the balls bonded on dies are out of designated positions. Thorough investigation reveals that the probable cause is the failure during Pattern Recognition System (PRS) after lamp burnt (dark field) incident.
Circular pulley-like parts are ubiquitous between each machine. According to Mr. Leong, this is meant to provide operators with more time to troubleshoot and amend errors in case of machine breakdown. During that time, other functional, parallel machines can continue production using inputs stored or winded on the pulleys.
Besides that, I went to the front end where Mr. Leong is stationed. He was monitoring the production of a new product. An ineluctable predicament surfaced in the form of the absence of numbering of die pick up starting position. To rub salt on wound, some of the dies in propinquity of the edges were not entirely defective. This will pose some impediment to the effectiveness of the Pattern Recognition System (PRS) in identifying good (usable) dies. After numerous attempts, the indefatigable spirit of Mr. Leong prevailed as he successfully estimated a proper starting position by trial and error.
Mr. Leong also taught me that the Pattern Recognition System (PRS) of ADAT2 die bonding machine works on the principle of binary scale. In other words, it discerns patterns based on black and white images of the die. On the other hand, PRS of ADAT3 machine not only works on binary scale, but also grayscale. This renders the PRS of ADAT3 machine significantly efficacious than ADAT2 machine. Nonetheless, higher cost of ADAT3 machine prevents its widespread use.
My learning experience continued after lunch break. After the front end process, the die bonded as well as wire bonded lead frames will be channeled to the mid-end process. Before that transfer takes place, unusual lead frames will be identified using stickers with different colours to inform personnel in mid end. Firstly, red coloured sticker indicates that certain part of the lead frames contain new device or new lot. Within front end itself, lead frames having new device will be noted with pen marking on lead frames to avert or assuage mixed device possibility. Next, green coloured sticker is used when a portion of the lead frames is unwanted scrap of production. Lastly, blue coloured sticker is used to represent lead frames that had been welded. Welded parts may be uneven and therefore it is not advisable to mold them.
During a casual stroll around front end, I came across several customer complaints. A customer from Korea complained about partial lifted ball bond. The cause of this is believed to be floating of lead frames due to copper build-up at heater block. During wire bonding process, as the capillary exerts lateral force on the ball at the bondpad, the force will not be spread evenly. Next, a Dutch company complained about offset ball bond, which means that the balls bonded on dies are out of designated positions. Thorough investigation reveals that the probable cause is the failure during Pattern Recognition System (PRS) after lamp burnt (dark field) incident.
Circular pulley-like parts are ubiquitous between each machine. According to Mr. Leong, this is meant to provide operators with more time to troubleshoot and amend errors in case of machine breakdown. During that time, other functional, parallel machines can continue production using inputs stored or winded on the pulleys.
Sunday, May 9, 2010
7 May 2010
I gained a deeper understanding on the wafer sawing process and the corresponding machines involved. First off, I noticed that the amount of wafer sawing machine is considerably more than that of wafer washer machine. After some lucid explanations from Miss Manggala, this is because one wafer washer can be shared by several wafer sawing machines as the washing process is a brief process. In a wafer washer machine, water is shot at high velocity at sawn wafer which is rotated swiftly. Although such conditions may seem to be destructive to the dies, the dies are actually unaffected! However, if we were to touch the surface of the wafer, though this condition seems to be less extreme than the previous instance, the dies will be damaged by our electrostatic charge as well as oxidation.
Next, in each wafer sawing machine, chemicals are used to clean impurities on the wafer surface. When the sawing blade is blunt or different blade is needed, the suitable blade will be changed manually by operators. After the new blade is installed, a dummy wafer is used for blade dressing. In other words, it means blade seasoning or to smoothen the blade edge. If this step is neglected, wafer sawn would have cracked dies.
After the sawing process, as I had mentioned earlier, wafer washing ensues. Then, that piece of wafer will be stored in cupboards. These cupboards are annexed with gas hose that channels nitrogen gas into the cupboards which can prevent oxidation of dies.
Oh! Ya! How can I not talk about the irksome bonding diagrams? I managed to complete the re-updating of SOT457P and SOT753P bonding diagrams.
Next, in each wafer sawing machine, chemicals are used to clean impurities on the wafer surface. When the sawing blade is blunt or different blade is needed, the suitable blade will be changed manually by operators. After the new blade is installed, a dummy wafer is used for blade dressing. In other words, it means blade seasoning or to smoothen the blade edge. If this step is neglected, wafer sawn would have cracked dies.
After the sawing process, as I had mentioned earlier, wafer washing ensues. Then, that piece of wafer will be stored in cupboards. These cupboards are annexed with gas hose that channels nitrogen gas into the cupboards which can prevent oxidation of dies.
Oh! Ya! How can I not talk about the irksome bonding diagrams? I managed to complete the re-updating of SOT457P and SOT753P bonding diagrams.
Thursday, May 6, 2010
5-6 May 2010
5 May 2010
It looks like my time here in NXP Semiconductor is preordained to be filled with a deluge of bonding diagrams. After completing 1119 bonding diagrams, with each consuming up to 5 minutes to complete, I was again assigned the unenviable task of updating bonding diagrams. I was ordered to add details, make circumlocutory resize actions on pictures and other petty stuffs. I was able to finish re-updating SOT346 (SC59), SOT323 (SC70), SOT457 (SC74) and SOT457P bonding diagrams which involved adding the notation “1 die size” as well as adding information table. My day just couldn’t get any worse. I was again asked to refine SOT353P, SOT363P, SOT457P and SOT753P bonding diagrams by adding yellow colour to highlight package, family and crystal information, as well as expanding ball size, subsequently centralizing them. It was an unimpeachable waste of my time! The only thing I learned is to suppress my anguish and tried not to be too antagonized.
6 May 2010
Nothing learned today. I finished re-updating SOT353P and SOT363P bonding diagrams amid utmost exasperation and frustration.
It looks like my time here in NXP Semiconductor is preordained to be filled with a deluge of bonding diagrams. After completing 1119 bonding diagrams, with each consuming up to 5 minutes to complete, I was again assigned the unenviable task of updating bonding diagrams. I was ordered to add details, make circumlocutory resize actions on pictures and other petty stuffs. I was able to finish re-updating SOT346 (SC59), SOT323 (SC70), SOT457 (SC74) and SOT457P bonding diagrams which involved adding the notation “1 die size” as well as adding information table. My day just couldn’t get any worse. I was again asked to refine SOT353P, SOT363P, SOT457P and SOT753P bonding diagrams by adding yellow colour to highlight package, family and crystal information, as well as expanding ball size, subsequently centralizing them. It was an unimpeachable waste of my time! The only thing I learned is to suppress my anguish and tried not to be too antagonized.
6 May 2010
Nothing learned today. I finished re-updating SOT353P and SOT363P bonding diagrams amid utmost exasperation and frustration.
Tuesday, May 4, 2010
3-4 May 2010
3 May 2010
I continued my nexus with bonding diagrams. I was given 4 new packages to update today, SOT23, SOT343, SOT343F and SOT346 (SC59). I was able to finish all of them before lunch break. The following is the comprehensive lowdown of the packages and the corresponding amount of bonding diagrams that I had done.
After lunch, I went to the line with Miss Norham. To be more specific, we went to the back end portion of the line. While waiting for our turn to perform tests on products, she mentioned that there are currently four companies supplying lead frames to NXP Semiconductor. They are Sumitomo, Enomoto, Possehl and Jade. The test that she intended to perform was to gain cognizance of the feasibility of using Possehl lead frames for production purpose for a particular device. To put it compendiously, she will put the trimmed and formed units using current production lead frame as well as those using Possehl lead frames into NT16 machine, to garner results or summary regarding the number of functional and substandard devices.
In NXP Semiconductor, as I had mentioned earlier during previous blog entries, there are two types of machines involved at the back end process. One being the IBE, and the other being NT16 or NX16 or TDD. When NT16/NX16/TDD comes into mind, ATF machine is indispensable. ATF stands for Auto Trim and Form. For more details on the differences in terms of operations between the two types, kindly refer to my previous blog entries. For this entry, I want to elaborate on the advantage of NT16/NX16/TDD over IBE. Firstly, NT15/NX16/TDD is faster and can be used for matrix as well as bi lead frames. IBE on the other hand, can only be used for bi lead frames. In addition, another difference between the two types of machine is that one PARSET is needed for an IBE, while a number of NT16/NX16/TDD can share a same PARSET. PARSET can be deemed as the repository of the information and operations for carrying out acceptance test (AT) and final test (FT). For further information on AT and FT, please refer to the following paragraphs.
From my observation as well as from the explanations provided by Miss Norham, for NT16/NX16/TDD, trimmed and formed units will be dumped into a rotating, vibrating bowl, where the units will be subsequently fed to a linear track. Here, there are sensors to sense the presence of units. Next, the units will go through a position sensor that indicates whether the units are oriented correctly or otherwise. Those units with wrong orientation will be corrected in the rotary portion. Then, the units will undergo final test (FT) to be followed by acceptance test (AT).
Besides that, there are two tests generally performed. The first being the final test (FT), and the other being the acceptance test (AT). Both are coterminous. Nonetheless, final test entails stricter or more rigorous test limits. Final test specifications are set by us, while acceptance test vicariously mirror what our customers want. Moving on with the similarities, both tests involve copious test specifications. For instance, akelvin and fkelvin. Notice that the first letter is the only discrepancy. The letter “a” refers to acceptance test, while the letter “f” indicates final test. Those units that failed either akelvin or fkelvin are regarded as units that may have missing legs or contacts, bent legs such that the clamper part of the test machine is not able to clamp the unit legs properly. If there were to be excessive amount of failed units under akelvin or fkelvin, the clamper contacts may be filthy. Next, I encountered units failing open test. These units may be devoid of dies (dummies), wires or have edge dies. Moving on, short test reveals units that have wires short circuited. Moreover, dvth and avth tests present us with a notion that units failing these tests may have flawed die attachments. Failed units of the respective tests will be guided into designated bins where they can be collected for further tests.
Lastly, I learnt about the meanings behind the notation used in marking each unit.
4 May 2010
What I did for today were just repetition of the things I had done before. Someone please succour me away from this purgatory. It is slowing emasculating my brain power as the things I am doing are just repetitive in nature. An overview of what I had done: Wafer inspection, updating of company’s Product Master Data (PMD) and acquisition of images using high power scope. The subjects of image acquisition are die sheared lead frames. Die sheared lead frames mean that the dies that were attached to the bondpads on lead frames were pushed away in a process known as die shear to evaluate the strength of the die bonding.
I continued my nexus with bonding diagrams. I was given 4 new packages to update today, SOT23, SOT343, SOT343F and SOT346 (SC59). I was able to finish all of them before lunch break. The following is the comprehensive lowdown of the packages and the corresponding amount of bonding diagrams that I had done.
After lunch, I went to the line with Miss Norham. To be more specific, we went to the back end portion of the line. While waiting for our turn to perform tests on products, she mentioned that there are currently four companies supplying lead frames to NXP Semiconductor. They are Sumitomo, Enomoto, Possehl and Jade. The test that she intended to perform was to gain cognizance of the feasibility of using Possehl lead frames for production purpose for a particular device. To put it compendiously, she will put the trimmed and formed units using current production lead frame as well as those using Possehl lead frames into NT16 machine, to garner results or summary regarding the number of functional and substandard devices.
In NXP Semiconductor, as I had mentioned earlier during previous blog entries, there are two types of machines involved at the back end process. One being the IBE, and the other being NT16 or NX16 or TDD. When NT16/NX16/TDD comes into mind, ATF machine is indispensable. ATF stands for Auto Trim and Form. For more details on the differences in terms of operations between the two types, kindly refer to my previous blog entries. For this entry, I want to elaborate on the advantage of NT16/NX16/TDD over IBE. Firstly, NT15/NX16/TDD is faster and can be used for matrix as well as bi lead frames. IBE on the other hand, can only be used for bi lead frames. In addition, another difference between the two types of machine is that one PARSET is needed for an IBE, while a number of NT16/NX16/TDD can share a same PARSET. PARSET can be deemed as the repository of the information and operations for carrying out acceptance test (AT) and final test (FT). For further information on AT and FT, please refer to the following paragraphs.
From my observation as well as from the explanations provided by Miss Norham, for NT16/NX16/TDD, trimmed and formed units will be dumped into a rotating, vibrating bowl, where the units will be subsequently fed to a linear track. Here, there are sensors to sense the presence of units. Next, the units will go through a position sensor that indicates whether the units are oriented correctly or otherwise. Those units with wrong orientation will be corrected in the rotary portion. Then, the units will undergo final test (FT) to be followed by acceptance test (AT).
Besides that, there are two tests generally performed. The first being the final test (FT), and the other being the acceptance test (AT). Both are coterminous. Nonetheless, final test entails stricter or more rigorous test limits. Final test specifications are set by us, while acceptance test vicariously mirror what our customers want. Moving on with the similarities, both tests involve copious test specifications. For instance, akelvin and fkelvin. Notice that the first letter is the only discrepancy. The letter “a” refers to acceptance test, while the letter “f” indicates final test. Those units that failed either akelvin or fkelvin are regarded as units that may have missing legs or contacts, bent legs such that the clamper part of the test machine is not able to clamp the unit legs properly. If there were to be excessive amount of failed units under akelvin or fkelvin, the clamper contacts may be filthy. Next, I encountered units failing open test. These units may be devoid of dies (dummies), wires or have edge dies. Moving on, short test reveals units that have wires short circuited. Moreover, dvth and avth tests present us with a notion that units failing these tests may have flawed die attachments. Failed units of the respective tests will be guided into designated bins where they can be collected for further tests.
Lastly, I learnt about the meanings behind the notation used in marking each unit.
4 May 2010
What I did for today were just repetition of the things I had done before. Someone please succour me away from this purgatory. It is slowing emasculating my brain power as the things I am doing are just repetitive in nature. An overview of what I had done: Wafer inspection, updating of company’s Product Master Data (PMD) and acquisition of images using high power scope. The subjects of image acquisition are die sheared lead frames. Die sheared lead frames mean that the dies that were attached to the bondpads on lead frames were pushed away in a process known as die shear to evaluate the strength of the die bonding.
Subscribe to:
Comments (Atom)
