Guide to Switching-Mode Power Supply (SMPS)--Table of Contents

Home | Articles | Forum | Glossary | Books

[Note: more articles to be posted soon]

Introduction and Personal Note

Looking back, I realize there was actually never an exact moment when I said to myself "I have had it with Physics - let me do Electronics now"! My near mid-life career change was a rather gradual process. Faced with an exponentially declining interest in pi-mesons, Lagrangian multipliers, quantum electrodynamics, and so on, my grades had started scraping the bottom of the barrel. It didn't help that I perceived my last bunch of professors to be largely apathetic to students in general - it seemed that teaching just happened to be what they needed to keep doing, to be eligible for research grants, which is what they really enjoyed doing. And Physics itself, for all its initial undergraduate allure, had at the post-graduate level, turned suddenly very mathematical and abstruse, contradicting my inherent desire to be firmly embedded in reality (not virtual reality). Unfortunately, the disenchantment reached a culmination only during my second Masters degree program, in Los Angeles. Too late! So though I eventually did part company with Physics (as good friends, I may add), there was a slight problem - I didn't have a clue what to do next. I call that my Problem Number 1.

Back in hot, bustling and dusty India, it took me several years to figure things out. But finally, I did! The small bags of transistors, capacitors, resistors and inductors that I had lately started tinkering with, held the answer to all my problems. And hope for the future.

This was my long-awaited lifeboat. I could now feel, touch, build and test whatever I did.

No deceptive sense of comfort lolling around in lush minefields of equations and algebraic abstractions. This was the real world, the one that we live in every day.

Problem Number 2: I still didn't know the ABCs (or the NPNs) of electronics. So I had to teach myself very gradually, working days and often very late into the nights, barely stopping only to ask the elderly local components dealer, daring questions like - what is a transistor?! This act went on for a pretty long time - in fact I became the Rocky Horror Midnight show - you got to see me mostly at midnight for several years in succession.

But it would have still been impossible if I had not met a few very remarkable men along the way. Finally, with all the help at my disposal (most of it mine), I think I made it into the exciting world of electronics. And into power electronics.

Aha! I could start rolling down the shutters now. Or could I?

The above chain of hair-raising events is the one and only reason this guide ever got written in the first place! But wait! I have explained "how" this guide got written, but did I explain why? Actually I haven't yet. Because that has something to do with the last major problem I faced. I call it Problem Number 3 - encountering people who knowingly or unknowingly thwart the growth of the engineering discipline that gives us growth. Now, I had personally been through a rather life-changing process (of being rescued by Electronics). So perhaps it was more natural for me to always think I owed Electronics my best, in return for favors received.

But I realize not everyone thinks along those lines, at least not all the time. Maybe they had affluent fathers paying for their shiny EE degrees from MIT or CalTech. But I didn't have an affluent father nor an EE degree. However, at some stage, we all have to realize that we share the same forces of nature, and a common stake in its existence and further development. Our destiny is eventually common, and therefore we have a common responsibility to uphold it too. Anybody who has learnt enough about the immense mystical forces of nature realizes that he or she has really learnt nothing at all. It will therefore be very surprising if they don't end up imbibing the sense of humility that Newton once expressed in the following words:

"I know not what I may appear to the world, but to myself I seem to have been only like a boy playing on the sea-shore, and diverting myself in now and then finding a smoother pebble or a prettier shell, whilst the great ocean of truth lay all undiscovered before me." Power Electronics too, is just a small part of that infinitesimal part of the universe that we have just begun to understand. There is much, much more, just waiting to be discovered.

Should we be the ones to encourage that onward natural process, or thwart it (even momentarily) with our petty office-space personal agendas? Finally, when I had seen too much and heard too much, I wrote the following paragraphs somewhere on the web, in what is now a rather controversial opinion piece for some people who obviously don't understand the logic or the motivation for it (see last page of Sub-Section 1)

'Technology may never gain a foothold in a "king's court," where you are either rewarded with largesse for being vehemently agreeable, or unceremoniously sentenced to the dark dungeons for the rest of your life. Engineers like to speak out - but usually only when they were sure of their facts and have incontrovertible data to back themselves up. They therefore deserve and need a "peer environment," where they are judged (primarily) by the respect received from their peers - the king be damned (on occasion)! It must be kept in mind that this can really bother the king sometimes! So managers who supervise engineers, should be fairly competent at a technical level themselves and respect data and facts equally. They can't attempt to win a technical argument by throwing rank on their subordinates. Nor should they ever go around, God forbid, trying to subsequently shoot the "emotional and/or disrespectful" engineer down ("that'll teach him"). Surprisingly that does happen more than we dare admit. Not only does the good engineer pay the price, but so does technology in the long run.' The only reason that this piece (based largely on the old wisdom of my dear long-time mentor and former-former mentor) turned controversial was I suspect, because it had hit closer to home than even I had imagined. It’s always amusing that whenever someone has one-too-many skeletons in their closets, the very sound of a distant siren triggers off their worst fears. I was told to stop talking about things I didn't understand, and stick to my (humble) circuits. I was also refused the normal official Author Encouragement Program payment that I thought was due to me as per their guidelines - for this article and even for my other popular power electronics book, which they had already used freely to promote their products. Finally the best thing I did was to quit as soon as I could! Without notice. Then surprise, surprise! Just after I resigned, they went and restructured exactly as I had been preaching all along - by re-amalgamating their two erstwhile groups "Portable Power" and "Power Management" into one, saying privately that there would be "more sharing among the engineers finally". My words exactly (read the article)! Weighing all these events in mind, I found some peace knowing the net result of my article was that a few better-designed, more peer-reviewed products would ultimately emerge from the very same company in question (whether they cared to admit it or not). For sure, the winner wasn't me, certainly not some insecure small-minded manager in a hopelessly high position. It was electronics that had had its day. And that was enough for me.

Till a while ago, I had naively thought large corporations, especially those showcasing themselves in glitzy facilities headquartered in Silicon Valley, had woken up to the times and become more professionally managed. To me that meant things like not allowing race-related slurs to demoralize struggling engineers, not allowing chilling war-rhetoric indiscriminately sent via company E-mail (making employees wary of their own supervisor's basic sanity at times), and simply, simply, just rewarding all efforts fairly and without discrimination. Too much to ask! I wasn't too sure anymore that the field of electronics, the one that I was trying so hard to nurture, was getting even close to what it deserved. Sure, they had now started declaring "record gross margins" and so on. But behind this benumbing onslaught of pure PR, you have to remember that that their new-found exhilaration was a) borne mainly on the shoulders of a new breed of extremely talented, friendly and pro-active engineers and b) basically, they just stopped loss-making operations, in areas that were outside their "core competence" (in reality: those business units that had been so badly managed from start to finish, that even the engineers couldn't make a difference anymore). Further, almost without further thought, they kept laying-off several talented or promising engineers, some that I knew personally - often because their own managers had screwed up so bad they needed alibis to present before their equally bad supervisors, who needed alibis to present to theirs … and so on. Of course the last man standing was apparently just too busy counting the millions of dollars cash bonus he had just received for meeting the company's (short-term?) "targets". End of story. Not a tear for those engineers that were walked out one fine day: a) without the slightest warning; b) without even being given an opportunity to present their side of the story - quite unlike even a normal court of law anywhere in the world. I asked myself - what if Newton or Einstein had been similarly dogged by incompetent dishonest supervisors? Would the world have been a better place today? And come to think of it, how many potential Newtons and Einsteins had these companies already banned into the hell of dark obscurity, and possibly premature retirement, while chanting that their analog ICs were nature incarnate ("the sight and sound of information")? We never know the real casualty toll ever, do we? As notice, I can honestly say I have not found the solution to Problem Number 3 yet.

But I am still trying! And this guide is an effort to do just that.

So now, it’s time to tell you what exactly I have tried to achieve with this book. One unique aspect about designing power supplies is that the "devil is in the details". In other words, I, as a technical writer, can either put in everything, including the supporting Math, and come up with a guide that (only) professionals would like. Or I could try making it very simple and straightforward for the beginner. But then the chances are very high I would miss out on the very essence of what power supply design is all about - the optimization, and design trade-offs. To strike a meaningful compromise between simplicity and depth requires a very carefully considered structure of presentation, one that I have really tried hard to achieve in this book. For example, several books out there, try to give a step-by-step detailed design procedure for DC-DC converters. However, they seem to routinely miss out on the important fact that the input is rarely, if ever, a "single-point" input voltage level. It’s usually a "wide-range input" and we need to be very clear which converter stresses are at their worst at the highest end of the input, and which ones at the lowest input end. We also need to know which stresses we need to give priority to during a particular design step within that procedure. Clearly, designing a good power supply is not a trivial task! In Section 2 I have presented a universal design procedure for DC-DC converters that hopefully fulfills the simultaneous demands of rigorous detail as well as simplicity.

So what did we do in Section 1? That to most readers is just an introduction that they can readily skim over. Wrong! Let's take a step (and page) back. This particular introduction actually starts at the component level, not at the topology level as most other books do. The hope is that now, even a beginner, can understand the mysteries of a capacitor and inductor, then tie them up synergistically, to derive a switching converter topology. In fact, it will become clear that all topologies evolve out of a basic understanding of how, in particular, the inductor works. Here, advanced readers should beware. Because, while interviewing even senior engineers for applications engineer positions, I found that many of them are still quite uncomfortable with the very concept of an inductor. Therefore, I think it’s a good idea for every reader, novice or advanced, to read the guide in the order of sections presented, starting from the very first section. Just don't be caught reading it (by your perception-driven supervisor!). The temptation of jumping straight into an advanced section to "save time" may just end up slowing things down even more in the long run (besides causing avoidable bruising of self-confidence for some, along the way). Basic concepts always need to be brought in at the right time, exemplified, and then firmed up to last a lifetime.

In Section 3 I have tried to start at a fairly basic level again, but then ramped up steeply to provide one of the most detailed step-by-step procedures available for designing off-line converters and their associated magnetics. This includes the dreaded Proximity Effect analysis. I have broken up the basic procedure into two separate iterative strategies - one for foil windings and another for round windings, because their respective optimization procedures are really very different. There are also generous amounts of curves and plots thrown in to quickly help the engineer visualize and design the magnetics optimally.

I have included a section devoted largely to switching losses in MOSFETs, since this topic has become increasingly vital as switching frequencies are increased. But it has been presented with some of the most carefully prepared and detailed graphics probably seen in related literature - highlighting each phase of the turn-on and turn-off individually.

Common simplifying assumptions have also been made whenever appropriate, and the user should thereafter have no trouble anymore practicing this rather poorly understood area of power conversion. There is also some interesting parameterized graphical information available that can come in handy either for an applications engineer selecting external MOSFETs, or an IC designer trying to optimize the driver stages of the chip.

The section on loop stability is likewise presented from scratch to finish, with very detailed accompanying graphics. My hope is that for the first time the reader will have easy access to almost all the equations required for loop compensation. Now, even a novice, can very quickly get very deep into this area (as I once did).

There are also seven sections on EMI, starting from the very basics and moving up to a full mathematical treatment. This is again a topic that has been almost studiously avoided in most related literature, and yet is needed so badly today. It needs much more elaboration I thought.

To cap it all, there is an "interview-friendly" FAQ, several Mathcad files, and various design spreadsheets thrown in.

As notice, the guide has been designed to try to live up to its name "A to Z". Of course that is never really going to be possible, least of all in an all-encompassing area such as Power Conversion. But hey, I did give it a shot! The stage is now set. I hope you like this book, even if it’s A to Z with some of the alphabets missing along the way, and go on to make a small but noticeable difference, using it. Though I do strongly suggest you choose where you attempt to do it, because that that makes a big difference in the long run - to technology and to its committed practitioners: you the engineers. And of course, it’s to you that this guide is solely dedicated.

Also see: Electromagnetic Interference (EMI) in Switching-Mode Power Supplies

Top of Page

PREV: (none) NEXT: The Principles of Switching Power Conversion Guide Index HOME