Chip Scale Packaging for Modern Electronics
Chip Scale Packaging for Modern Electronics
Keyword: Chip scale packaging
Joseph Fjelstad, Reza Ghaffarian and Young-Gon KimElectrochemical Publications Limited, IoMISBN 0 901150 43 6438 + xxiv pp.GBP 98.00, USD 198.00267 figures, 101 tables and 265 references
Do you know how to define a Chip Scale Packaging (CSP)? Do you know how many types of CSPs are there? Well, this great book will give you the answers if you are as ignorant as I was until I picked it up[1].
Joe Fjelstad needs no introduction; he is well known in our industry as the guru on flexible circuits, including their use for TAB technology (hence the connection to CSP). He has written several books on these and other related subjects. Dr Reza Ghaffarian, also a prolific author, works for the Jet Propulsion Lab of the California Institute of Technology, particularly in semiconductor packaging technology, advisory to NASA. Having previously worked for a major Korean semiconductor manufacturer, Young-Gon Kim is a director of Tessera Inc., responsible for packaging design. These three principal pens are complemented by a few well-qualified chapter writers and co-authors.
The book very clearly conforms to the excellent Electrochemical Publications tradition, appearance and quality. I have nothing to say on the physical side except, perhaps, some colour photographs and diagrams may have enhanced the appearance, but would they have enhanced the value of the book? Technically,probably not. I found one point that I would have liked to have seen improved:the index. A ratio of 60:1 pages for a book of this nature is a little meagre. While writing this, I am aware that indexing is a very vexatious subject. Technical indexers of quality, if they can be found, do not grow on trees and command very high fees. The publishers cannot therefore be held to account on this matter, even though one sometimes has to search to find mention of a related subject. Oh! I found only one typographical error! There is just one general point I would like to say to the publishers; this book is well-referenced, using the numbered chapter end-note technique. For future books, I suggests the more modern Harvard style of chapter referencing, e.g.(Smith, 2001b), offers the advantage that the reader can see who is being referenced and how old the reference is, without leaving the page, he is reading.
The authors have cleverly divided the book into three distinct sections as follows.
CSP market and infrastructure (six chapters, 82 pages).
Chip scale packages (ten chapters, 276 pages).
CSP reliability (two chapters, 68 pages).
Each chapter covers a different aspect of the subject of the section. Reading through the book, any thought about arbitrariness of the division of sections and chapters is soon dispelled, as they follow a clear logic. The subdividing within the chapters, up to five levels, is exemplary, as well. In fact, the table of contents, about twice as long as the index, is very useful. Furthermore, the editing has been done well, as is clear from the lack of a feeling of disparity between the different authors' chapters.
The strengths of the book are too numerous to detail. The weaknesses are few. I did feel, I would have liked a bit more on alternative assembly techniques and on the chemistries of fluxes, pastes, underfills and encapsulants and their potential interactions – this left me a little thirsty for more knowledge. The very thorny subject of cleaning after attachment appears to have been mentioned only for flip-chips where it is said to be almost impossible, but I could find nothing for the packages with a larger stand-off, such as BGAs, nor was this subject mentioned in the chapters on reliability. These are very minor niggles, compared with the masses of useful and valid information, but this reviewer would not have the reputation if he could not find something to complain about: he would have to invent something! He would like to make one specific comment, though. This book is not just for consumer-goods applications or aerospace ones, it covers the whole gamut between these extremes. Some of the examples in the illustrations are telling in this respect, including a photograph of what is said to be the densest flip-chip circuit in the world,taken from a Sony miniature camcorder, indeed impressive. I have always said that it needs better engineering to produce consumer goods than hi-tech hardware produced in small quantities, if only because the holistic cost becomes such a major criterion.
Of course, CSP often goes hand-in-hand with assembly onto high density interconnect structure (HDIS) printed circuits. It is obviously well beyond the scope of this book to detail how these should be designed. Notwithstanding,exactly the same techniques are often used within the CSP packages themselves. Chapter 4 therefore gives a summary of the basics, which can be extrapolated to the substrates on which CSPs can be mounted.
At the moment, the main use of CSP techniques, alongside embedded components in the interconnection technology, is for "miniaturised" (horrible word, but so expressive!) electronics. The obvious applications are for portable equipment,such as PDAs, camcorders, cellular telephones and the like. However, it is hardly less obvious that any electronics, where either space or weight is at a premium, is an evident application. Aerospace immediately springs to mind, but reliability is synonymous with this. The last section of this book is therefore very important to determine how to achieve this.
I should like to express a personal opinion here. We have seen how CSP techniques have been used in small consumer equipment. It is my belief that we are at the threshold of a new era in electronics packaging. Twenty years back,surface mount technology using discrete IC components, such as small outline packages, was at the cutting edge. I forecast at that time, incorrectly, that SMAs would be limited to those applications which benefited from the reduced size and weight, while through-hole techniques would remain for "ordinary" work. I was wrong, mainly because I did not foresee the advent of what are now current techniques in placement and joining technologies and equipment, and their consequent cost savings. I am not going to make the same mistake again! My present forecast is that mainstream electronics will be replaced by CSPs on HDISs within a few years and that the "conventional" SMA on ordinary two-sided or multilayer boards will shortly seem as antiquated as all PTH assemblies today. I think this will become cost-driven as the technologies evolve. This book is therefore timely and will doubtless be the "bible" for at least a few years. Therefore, reading is required for all those in our industry with a desire not to be left behind, even though some may not see the immediate need.
Brian Ellis
Note
- 1.
In case you would like to know, there is more than one definition of CSP,but this book evokes a maximum linear dimension of 1.2 times the die size. As for the number of types, Table 1.2 evokes 63 (wow!) but it is mentioned elsewhere that there may be another 70 in development. Let us be clear, these are types of CSP, each of which may have many sizes and pin-out configurations!
