CitationDownload as .RIS
Emerald Group Publishing Limited
Copyright © 2006, Emerald Group Publishing Limited
How could I possibly join them on to the little bit (two inches wide)...?
Keywords: Computer software, Electronics industry
If you are like me, you will have an enormous quantity of information stored on your computer, in the form of files, e-mails, images, whatever. No matter how well you arrange them, retrieving a document takes time and effort. Of course, Windows has a search feature, but it sometimes seems to take a day and a half to find a file where you can remember only one or two keywords and searching e-mails is limited, to say the least. I have, over the last couple of months, been trying a free software called Copernic Desktop Search (www.copernic.com) and I am impressed. If I type the keywords into it, it will give all the files, e-mails and other documents containing them within a second, in separate categories and classed according to age. It is obvious, it cannot do this as a true search; it goes through a single keyword file. The latter is generated, in the background, after installation, without interfering with normal use. This process can take many hours but, once it has finished cataloguing them, it becomes fully functional. It also automatically updates itself as you introduce new documents. Its great strength is that I can find anything in my archives of the IPC TechNet, received as an e-mail list. This is interactive; if I type, for example, laminate, it finds 297 e-mails and 98 files. As soon as I add another letter or two, such as bro, it drops to 60 and 68, respectively; adding more letters to finish the word bromine, it has narrowed the search down to only 9 TechNet files and 36 of my own, classed by folder and date. This is the most useful tool I have found in a long time but it does have one weakness: although it claims to conduct internet searches, it is not a patch on Google or many other search engines, but it does remember where you have been in your browsing history, so you can quickly re-find a site that you visited some time previously. Highly recommended.
The “standard” tool for preparing photos for publication, whether on a web site, by e-mail or even as a printed document is Adobe Photoshop. No doubt, its latest version, baptised CS2 (Creative Studio) is excellent, but it is horrendously expensive (average internet prices $600, £450). I had thought to make a comparison between it and some of its rivals, but Adobe did not send me a review copy, although I did obtain the latest Corel Draw! Suite ($250, £300), with PhotoPaint and the Ulead PhotoImpact suite ($80, £50). I therefore decided to compare just these two and see how they measured up for internet applications. Before I do so, I would just like to mention the disparities between US and UK prices. The PhotoShop, at $600, is equivalent to £333 at the current exchange rate of 1.80 (time of writing), so the UK purchaser is paying a 35 per cent premium. Corel is worse: it translates to £139, not £300. How, by any stretch of the imagination, can a premium of 116 per cent be justified, just because one lives 5,000 km away? This is daylight robbery. Ulead is far less greedy, because $80 is only £44, just £6 less than the UK price, which is far more acceptable, even if the premium is still 14 per cent.
So, what do you obtain for your money? The Corel Draw! 12 suite does contain Draw! Itself as its key application and this is incomparable as a leading vector graphics system; if you need this, then do not hesitate to buy it. Then there is PhotoPaint, of course, which I will discuss presently. RAVE is a timeline-based animator for logos and other applications, also basically a vector graphics system, but bitmap data may also be imported and exported by conversion into any of the usual formats. This may be useful for animating titles or other items on a web site, but I recommend circumspection when adding bandwidth-intensive data. There are also some other utilities, one of which is excellent: Corel Capture is really useful for generating bitmap screen captures. All those you have seen in these columns since I started writing them have been captured using this utility (various versions).
Ulead PhotoImpact 10 suite contains two basic applications plus a number of utilities. Apart from PhotoImpact itself, there is also PhotoExplorer, which is a light and basic photo editor and filing system. An academically interesting utility is Cool360, which allows a series of photos to be seamlessly stitched together into a panorama of whatever angle you want, up to 360°, automatically correcting for position, angle and exposure. More usefully for web site designers, there is GIF Animator; this may be more practical than RAVE because it is a bitmap application and, as the name implies, it can use GIF natively, while also using objects and converting to other formats. I believe it is also more “user-friendly” for those using it for the first time.
Comparing the Corel and Ulead offerings for internet use, I would like to concentrate on two main jobs: preparing photographs for sending by e-mail and doing the same for inclusion in a web site. I must confess that I have used Corel Draw! Products for at least 15 years, so I am much more familiar with their software but I will try to be fair in my assessment.
Other than the photo-editing itself, to send a photograph by e-mail attachment requires, as a rule, cropping, resizing and compressing it, usually to a JPG format. To obtain the best results manually does require some skill. Cropping in PhotoPaint may be done by a rectangular selection, covering the unwanted part by a transparent red overlay, then crop to mask. There is a little problem with this technique: once you have made your selection, you cannot adjust the size without re-selecting. However, there is a better way that is adjustable, a crop tool, which is more flexible with a similar transparent grey overlay. Once you are sure of the selection dimensions and angle, you right-click to do the cropping. PhotoImpact has a very similar Crop Tool, except that there is no rotation possible; if it is necessary to correct a horizontal, then it would be better to rotate the image before using the tool, using the layer manager. On the other hand, dimensioning to fixed values or a locked aspect ratio or paper size is easier. For cropping, there is little to choose between the two.
Resizing to a suitable format for emailing in PhotoPaint is called Resampling and you may choose, say, 500 pixels wide at 72 dpi; this is straightforward. In PhotoImpact, this is unnecessary (see the next paragraph). However, you can do this manually, if you wish, by the resize feature, which also offers different methods of reducing the number of pixels. PhotoImpact wins on this score.
Compression is usually done to JPG. With PhotoPaint, JPEG Export is fairly simple, with little choice other than compression and smoothing. It works well. With PhotoImpact, you can do the same but with a much better set of controls including a wider choice of different compression algorithms and you can even choose to compress to any file size, which you can select. In both cases, you have “before and after” windows which allow you to check that the degree of “artefacting” is acceptable. However, there is one feature of PhotoImpact that is an absolute winner; there is an e-mail export/smart send feature which does a one button resizing and compression to generally acceptable values, then it opens your default e-mail client send attribute with your picture already set as an attachment. This, alone, makes the low-cost PhotoImpact an excellent choice and I can assure you that it will, in future, be my default application for e-mail preparation.
For preparing photographs (or other images) for web site editing, the procedure is very similar. Again, PhotoImpact is more versatile, although PhotoPaint does have a very useful web image optimizer. However, when publishing web photographs, it is often more desirable to edit them, rather than to put them on a web site “as is”. Without doubt, PhotoPaint has more artistic effects, but these are rarely used for the technical photographs that we are likely to publish. The possibilities for correcting brightness, contrast, hue, intensity, colour, cast and so on are similar in both applications, but they are easier to use in PhotoImpact, with better windowed and full-scale previewing, although the latter may be somewhat slower than one would sometimes like, in both applications. PhotoImpact also has a number of automatic enhancements that work well. Perhaps one of the most outstanding is termed high dynamic range. This is useful if you have an object which has a highlight: shadow ratio which exceeds the normal range but you wish to display detail in both the highlights and the shadows, while maintaining a normal contrast in each. This is done by a clever melding of two images, which may be either manipulated to optimise the highlights in one and the shadows in the other or, better, from two exposures.
Overall, I suggest that Ulead PhotoImpact 10 suite is the best overall buy for image editing for internet use (and other ones) and will do all, or at least most, of the work you are likely to need to do easily and well. I have found it to be a mature and a stable software and it comes with a 366- page manual. Perhaps the most telling argument is that it costs, in the UK, one-sixth of the price of the Corel Draw! 12 suite and one-ninth that of the Adobe PhotoShop CS2, neither of which are likely to offer significantly more for this kind of work. It should, perhaps, be mentioned that there is a cut-down version of PhotoShop at a much better price, but I am unable to judge whether the reduced features would be prohibitive or not.
Finally, for this prologue, Ulead will have started to market Media Studio Pro version 8, by the time you read this. This is a very significantly advanced software for video editing, including for internet streaming in several formats, much more than just a normal upgrade from v 7 but, to use a popular but almost meaningless expression, a paradigm shift.
You will remember, perhaps, in my last Commentary, my review section was largely devoted to the problems of EU legislation and how it affected us. The implications are far-reaching, of course, because it affects everyone: the designer, the PCB fabricator, the purchasing manager and even the vendors, as well as the assemblers, like ourselves. But how many assembly plant managers have tried to persuade “Upper Management” that this problem must be viewed holistically? This is partly symptomatic of the lack of concurrent engineering, so little practised in our industry. A few days ago, at the time of writing, there was a very heated argument going on in the IPC TechNet list server forum about whether this list should not be broken into two, one for fabricators and one for assemblers. There was one school stating that it is essential that both parties speak the same language, and bring in the designers and engineers, while we are about it. Sadly, the other school consisted mainly of some of the fabricators who seemed to be saying that they had nothing to learn from and nothing to give to assemblers. This has inspired me to look into how the various parties in our industry should interact with each other, beyond a simple vendor/client relationship.
Without much doubt, Earl Moon is one of the most controversial figures in our industry and he has what must be one of the worst web sites, into the bargain (Plate 1). Notwithstanding, even if you have to wait what seems hours for something to download, the wait must be worth it, because there is a veritable wealth of information on concurrent engineering – and its sister, design for manufacturing – to be found here. I should like to make one short quotation from Earl's Introduction, just to give you a taste of things to come – and to show I forgive him for his horrible multi-megabyte pages, especially the most hideous home page I have seen for many a long day:
Plate 1 A small part of Earl Moon's Home Page. This is probably the most informative industry-specific site on concurrent engineering available, but counts amongst the world's worst web designs. If you think it looks bad in greyscale, just wait until you see it in colour!
“Developing a manufacturable product, meeting all customer specified requirements, is critical to achieving low costs, high quality, and long term reliability. DFM must be addressed as early as possible in the product development cycle. This means starting at the concept phase, to be truly successful. This must continue to be addressed until product is delivered meeting all customer contract requirements, thereby being validated as proof of design.”
“Initiation begins with management's complete DFM program understanding, and its total commitment to DFM/CE principles and requirements. This is the most important process required to assure the primary objective (satisfy customer contract requirements) is reached. The primary objective, and the process required to reach it, must be communicated to the entire organization as this is a major management responsibility. POD's DFM training provides requirements for understanding the concepts and essential elements of a DFM approach. This book provides much of what is needed to develop, implement, and execute an effective DFM program vital to successful new product and process development.”
“Management must create and lead a team-based organization to establish and support DFM program development. POD, and this book, helps define roles and responsibilities to support DFM, as required for success. POD works with management and the team to facilitate their use of DFM principles, practices, procedures, and processes.”
“DFM guidelines must be established based on product types, components and materials but, no matter, DFM principles cannot vary. Also, process types, process capabilities, process times (takt times – not cycle times, within a lean manufacturing system). All costs must be considered but not compromised for apparent cost's sake alone. Within a DFM/CE and Lean system, real costs always are reduced when waste is eliminated and value is added (see my lean manufacturing presentation in this book's part 3). It costs so much less to do it right the first time than to do it over and over.”
All this is epitomised in this page from the site of a manufacturer of custom actuators. A short extract states:
The six groups/disciplines are:
product cost management;
program management/program coordinator; and
quality assurance/quality control.
“Each group works concurrently towards achieving a common goal – working with you to ensure full conformance with your requirements and an on-time prototype delivery.”
Granted that, in this case, we have neither PCB fabricator nor assembler, but the whole of this page is worth the visit as a clear grasp of the subject and how it is of benefit to all parties concerned.
This is the index of a detailed case study of the application of concurrent engineering within the context of the electronics industry (there are many other similar case studies in this best manufacturing practices website, but I chose this one at random from them). Although this study is more than 10 years old, it is still as apposite today as it was then and it covers more areas than my basic proposition, above. One concept that is new to me is the design for producibility, as opposed to design for manufacturability – yes, there is a subtle difference! Well worth spending half-an-hour or so reading the 14 pages!
This is a reprint of a magazine article (usually devoted to car manufacturing) describing how an anonymous electronics manufacturer (called reliant communications for the article) implemented concurrent engineering. In this case, the important point is not to go the whole hog just once, but to develop teams that will ensure that it is applied to all products and processes over time, by the naming of “process coaches”. These will ensure that the limitations of each process are made known to all. The result is that the company has reduced a 200þpage “how-to” book into nine simple modules on the firm's intranet, so that anyone can find the information needed to ensure the competitive manufacture of a product.
This is a paper, from the Hewlett- Packard stable, describing a software architecture that also places the emphasis of a company's intranet as a means of implementation and communication. This is not specifically for the electronics industry; the example used is an installation at the Ford Motor Company in Germany.
Although this is a commercial page from the site of PCB design service, it is perhaps an example of where concurrent engineering may fall down, if due care is not taken. The point is that the design is made according to a set of design rules and tested according to another set (or possibly the same set). The rest is up to the designer, provided that he does not violate these rules. Nowhere is it mentioned that a simultaneous dialogue is set up between the company's designer and the engineer, the fabricator, the assembler, etc. to make sure that everyone's requirements are met. It is true that many problems are avoided by a clever set of design rules, also written conjointly by all the parties concerned, but the design rules may not necessarily make for the best or lowest-cost solution. I am not saying that this will happen; the company may be very competent in the essential dialogue, but they do not say so.
This is a catalogue of projects undertaken within Loughborough University, about electronics assembly issues. The one that caught my eye was entitled, “Modelling and experimental investigation of reflow soldering” and the objectives read, “To construct from a proof of concept modelling system, a concurrent engineering tool to simulate the reflow soldering process, incorporating process knowledge and understanding derived from an extensive experimental programme. This will enable pre-production determination of IR profiles, facilitate design for manufacture of complex electronics assemblies and also increase the understanding of the physical and chemical phenomena associated with reflow soldering.” This had my mind whirring in top gear; the notion of being able to write an interactive modelling algorithm that allowed, for example, the designer to incorporate it into his ECAD system and, at the same time, tell the reflow operator the optimal settings for his machine for that assembly is simply mind boggling. It would go quite a way towards taking the capital letters out of concurrent engineering. I tried to find more details about this project on the internet but found nothing, unfortunately; however, those interested could contact the University.
Denso is a leading supplier to the Japanese automobile industry of electrical and electronic equipment, including engine computers. I visited one plant and research division of this company a number of years ago (when it was NipponDenso) and was very impressed then at their efforts to obtain maximum quality and reliability. Maybe this is one reason why you rarely see Japanese cars at the side of the road, waiting for help, because of an electrical fault! (Some readers here may remember that, in the past, I used the term WHS – Walk-Home Syndrome – to describe this problem, so common in some makes of cars, even expensive ones, from other countries). This document describes in detail how the company applies concurrent engineering from the conception of a new product. However, this is not an electronics assembly, but an alternator, but there is little difference in the general application of the concept.
Although only an abstract of a paper with the title Concurrent Engineering In High 1/0 BGA Implementation [sic! Of course, it should read I/O and not 1/0!], it is an interesting read. The author claims that there were lessons learned applicable to board design, assembly, rework, heatsink attachment and reliability testing, reducing costs and implementation time. The full paper may be purchased from the SMTA for the modest sum of $10, if you are not already a member – free of charge if you are.
Concurrent engineering, design for manufacturability and their associated themes are not buzzwords; they are an investment into the future that requires careful study to implement. Each company has its own, individual, infrastructure and the start-up CE team must be committed to the notion, to saving cash and thus improving profitability, to reducing the concept-to-finished product time, to reducing faults and rejects, to increasing reliability of the finished product, while ensuring the health and safety of the operators and minimal impact to the environment. This requires that the management be involved from the highest level and the members of the team selected from employees with the greatest spirit of polyvalence and broadmindedness. There is no room for the traditional “empire-builder” in such a team. In the electronics assembly industry, there is a natural tendency for departmental separation, for example, between the “mounting” and the “soldering” operations. The CE team member from the assembler must be able to make his contribution to help the designer, the PCB fabricator, the purchasing manager and all the rest of the team make the right decisions first time, without thought that his factory has departments. I am not saying it is easy to implement CE within AND without an assembly plant; the investment is considerable, but the results become apparent within a matter of months, in the balance sheet.
Note1 “What should I do with your strong, manly, spirited sketches, full of variety and glow? How could I possibly join them on to the little bit (two inches wide) of ivory on which I work with so fine a brush, as produces little effect after much labour?” Jane Austen, Letter to J. Edward Austen, 16 December 1816, in R. W. Chapman (ed.) Jane Austen's Letters (1952).