Musings on Publication — and Zero Sequence Modulation

• Miscellaneous
• Power Electronics

Perhaps you don’t think about it, but in order for you to read these articles, someone has to do something.

And I don’t just mean writing them. Stephane Boucher has set up this website so that it’s automatic, for the most part — at least from my end of things, as an author. When I get an idea for an article, I open up a new IPython Notebook, write my article, save it in a Mercurial repository, run a Python script to convert from IPython Notebook format to HTML, open up embeddedrelated.com, click on the Write New Article link, click on Tools → Source Code, paste in the HTML code, review + make final corrections, and click Publish. If it’s a short article (and that never happens anymore), it might take an hour or two from start to finish. Stephane gives us free rein, for the most part — I suppose if I posted death threats to someone, or disdainful comments on the hairstyles of Justin Bieber and Britney Spears, or classified plans for the U.S. Army’s secret autowwe\( %pitjwoi16^jroi we&+33rator program, he’d quietly take my post offline and ask me WTF is going on, but that hasn’t happened, at least not yet. So I have total power (BWA HA HA HA!) and can get something up and published in a jiffy.

Publications from “real” sources are more formal, because there are consequences to saying something stupid. So corporations have attorneys and public relations teams. It’s a well-oiled machine behind the scenes. Not necessarily an impersonal one, but a team of people working together to present a unified message.

As an applications engineer at Microchip Technology, part of my job is contributing to the development of technical publications. (Obligatory disclaimer: the messages and opinions expressed herein are my own, and are not that of my employer.) Datasheets. Presentations. Application notes. We call those “collateral”. And they take a lot of work to publish. A 500-page datasheet represents man-months of effort to compile and review for correctness. Application notes are man-weeks, maybe man-months if they’re based upon a larger hardware or software development project. We have technical writers that work full-time to take draft copy from engineers, whether it’s a carefully-crafted document or chicken-scratch drawings, and turn it into a professional-looking document that is consistent with corporate style guidelines. E. and J. are the two tech writers in our department, and they make it look easy. After they get a draft of a document ready, it goes through review from our technical staff, marketing, and legal departments, to ensure that the document is correct and consistent with corporate standards. The document gets an official title and document number; when everyone approves for publication then the document moves forward and is posted on the corporate website. Even if it took me only an hour to put together a high-quality draft of an application note (it takes a lot longer), and even if E. or J. took only an hour to convert that into the proper formatting (it takes a lot longer), it would still take a few weeks for review and final publication.

One interesting tidbit of information is the publishing software used by various semiconductor companies. Next time you download a PDF datasheet, look at the metadata, specifically the Creator field. In Adobe Acrobat Reader, it’s under File → Properties. Just from looking at a few datasheets, it seems like most companies (e.g. Linear Technology, Maxim, International Rectifier, Renesas) use various types of Adobe software, many of them (including Microchip, Intersil, Analog Devices, Cypress, and AustriaMicrosystems) using Adobe FrameMaker. Some are using custom publishing software, or you can’t tell which software they use to publish these documents, because the metadata just says the document came from a PDF library like iText or Adobe Distiller. Others are using content management systems with integrated PDF publishing tools. TI was using a program called TopLeaf; ON Semiconductor appears to be using BroadVision; Freescale is using Antenna House Formatter. Maxim was using QuarkXPress but is now on the Adobe bandwagon.

In any case, the semiconductor industry relies on datasheets and appnotes so that its customers (YOU!) can understand how to find and use the right products. In this environment, the importance is on the customer. There are a lot of anonymous workers that bring the documentation to a PDF file. We try to weed out the mistakes before publication, but sometimes we miss a few. So if you find something that’s confusing or inconsistent in a datasheet or application note, bring it to the attention of your local sales representative or applications engineer. I know at Microchip we always appreciate this feedback; it helps us fix mistakes and make our documentation better.

Sign in