Advances in DSP over these past couple decades have been responsible for some of our  most important technological leaps — MRI scanners, mobile phones, digital  image sensors, etc. — but despite that, it doesn’t always get the attention or interest it deserves.   Unfortunately, the reason for this bad reputation is woefully clear: most of the literature out there on DSP is extremely opaque, heavily focused on the theoretical not the pratical, and  it all tends to be written by domain experts for domain experts, steeped in their own distinct domain language.  There’s room and a need for all of that … but it does make it pretty rough for non DSP people to take advantage of some of the enormous benefits even the simplest DSP algorithms can offer.

Thankfully, people interested in improving the reliability and usefulness of their data have a friend in Dr. Stephen W. Smith, author of “The Scientist and Engineer’s Guide to Digital Signal Processing“.  Dr. Smith not only provides an excellent overview of many of the key concepts and filters you’ll use solving problems in the real world day to day, but he does it in an extremely easy to understand and accessible way, with an emphasis on implementation not just abstract theory (though the theory is often explained with clarity and reasonable depth as well).  A lot of the most common DSP ‘filters’ and algorithms are covered — FFT, FIR, IIR, moving averages, etc. — and they are all accompanied with intentionally ‘basic’ implementations in code in … well … Basic.

In an particularly enlightened stance, Dr. Smith also takes the unusual step of letting you decide entirely for yourself if this is the right book for you, since the entire book is available online free or charge, chapter by chapter in PDF format!  You can get printed copies of Amazon, and if you find the book useful it’s worth support the author (I ordered my copy 20 minutes after digging into a single chapter), but if you’re on a tight budget, you’ll be bound to appreciate his generous and open gesture!

Any other good DSP resources for beginners out there that your found noteworth?  Toss them up in the comments below so other people can benefit as well!

While this is somewhat of a specialised book, if you have to do any work with image processing (OCR, shape recognition, etc.), there are some good frameworks out there, but there’s hardly an abundance of easily accessible information on the subject if you need to implement something yourself.  Algorithms for Image Processing and Computer Vision (2nd Edition) is one of the better books available, and while — like most specialised technical books — it isn’t cheap, it is a good value given the amount of information it covers in one place.

Microchip has a great app note where they’ve compiled all kinds of useful little SW and HW hacks as a ‘tips and tricks’ guide.  wondering how you can connect a bunch of switches or buttons to a single pin?  They’ve got you covered.  Have a look at Compiled Tips ‘N Tricks yourself to see what problems you can solved differently on your next project.

Digging around for some information on MEMs gyroscope manufacturing processes, I came across this excellent white paper that explains a lot of the options that manufacturers have making low cost gyroscopes.  MEMs is a really fascinating business to be in, with very real manufacturing challenges, but some very unique opportunities as well.  Most of the driving force behind MEMs in recent years has been smart phones — by far the single biggest market for MEMs gyros! — but the DIY and HW world definitely benefits from this since it’s never been cheaper or easier to get high quality motion sensors.  Curious how MEMs gyros are made?  Have a look at: A Critical Review of the Market Status and Industry Challenges of Producing Consumer Grade MEMS Gyroscopes from Invensense.

Our TSL2561 digital light sensor  has always been a popular breakout.  Part of what makes this board so useful is actually the Adafruit driver, and that extra bit of effort that went into providing useful values like ‘lux’ instead of 0..1023 etc.  It’s easy to get raw data out of any sensor, but that’s not always very useful if you want to do something useful with these sensors.  Raw data needs to be converted to real world values, but this often takes more work than everything else in the driver put together, at least if you care about accuracy.

Using the Lux Equation from Taos gives a good overview of what’s involved in taking that raw sensor data and making it useful, taking into account easy to overlook values like the glass attenuation factor, etc.  If you’re interested in things like color correction, Calculating Color Temperature and Illuminance also has some good details on accurately converting RGB sensor data to color temperature.

Part of the motivation behind the new Adafruit Unified Sensor system is to make it easier to give people real world values without having to do all kinds of matrix multiplication yourself … but sometimes it’s useful to see what goes on behinds the scenes, and know how to calculate some of this stuff yourself as well.  All the real magic is in the fine print and the details!

Switching power supplies can offer much better efficiency than linear supplies, but they’re also a bit trickier to get working and require more components and more precise component selection.  How do you evaluate switching more power supply circuits?  In reality there are a lot of parameters to test often requiring specialised equipment — things like a DC load to test efficiency, special probes for your scope, etc. — but there’s a good article at Scope Junction on the topic: Stability Testing A Switching Power Supply.

The articles assumes you have access to some fairly high end equipment (a network analyser, etc.), but it’s an interesting read just to see some of the real world considerations that you need to keep in mind with commercial design and how much testing really goes on (or should go on) in the manufacturing process.

Some manufacturers do this better than others.  I have vivid memories of working as an apps engineer being tasked with helping a Japanese company using a brand new and complex MCU in their product.  Every time I saw an email from them in my inbox (daily) I winced because I knew it would be another incredibly precise question involving very specific values or test conditions and I’d have to call up the chip design team to get the test parameters and see what was up … it was never much fun, but it did instill in me a lot of respect for the testing efforts they went through, and sometimes they were right and identified one or two issues that got past the validation team MCU side.  It was a wakeup call for everyone, but a good one.  Testing matters, and often testing can be as much or more effort than initial product development.

Have any horror stories where you wish you had put more effort into testing yourself or had to compromise because of deadlines and it ended up biting you in the back-end?  Post them up in the comments below!

Adding a LIPO battery charger to your latest project?  You might find AN10910 worth reading from NXP, which presents some strategies to reduce the risk of working with LIPO cells and battery charging in general.

Crimping cables kind of sucks, but it sucks even more when your cables are poorly crimped and fail in subtle (or not so subtle) ways as soon as you finish screwing everything together.  Did you just order a couple thousand crimped cables from China and want to know if they’re well made?  Are you trying to make sure your own crimped cables are OK?  Molex has a great guide called the QUALITY CRIMPING HANDBOOK that sheds some light on what makes a good crimp.7

Want that in poster format to slap up on the walls of your lab or hacker-space?  Molex has got you covered there as well! English EU Metric version, for example.

It’s been a while since I mentioned any books on EE Bookshelf, but I’ve picked up quite a few this past year and thought it might be nice to mention some of them.

While my big New Year’s resolution for 2013 is to finally step into the 19th century, following up on my 2012 resolution I’ve been really trying to make an effort to integrate better testing into my code and projects, and part of that is recently playing with Unity, which is a unit testing framework for C that works for embedded projects.

Unit tests and test driven development are familiar terms to most PC developers, but they still haven’t made major inroads into the very traditional embedded space.  James W. Grenning’s book ‘Test Driven Development for Embedded C‘ aims to change that, and make test driven development far more accessible to embedded engineers.

If you’re serious about improving the quality of your code and test procedures grab a copy of the book for a meager $22 (DRM-free e-book) or in print off Amazon, and have a look at why test driven development is at least as relevant to embedded developers as anyone else.  Aside from being a veteran embedded developer, Grenning has been part of the test driven development movement since the very beginning, and he knows what he’s talking about!

Have any experience with Unity or another testing framework, or test driven development in the embedded sphere?  Post about it in the comments below!

I wasn’t aware of this before, but Google has a decent style guide for C++ development.  There are some interesting points in here that are very relevant to C development as well, and I always find this kind of stuff really interesting to read.

Thanks to Ian over at Dangerous Prototypes for the heads up on this app note from Maxim: Successful PCB Grounding with Mixed-Signal Chips.  Have a good GND plane and return path is critical for high speed or mixed signal layout, but it’s often overlooked, and there is precious little information out there that’s accessible to non-specialists.  It’s always nice to see app notes from manufacturers trying to make this somewhat opaque area of interest a bit clearer!

A big thanks to @SiliconFarmer for the heads up on this interesting article on ceramic capacitors and voltage variation.  I switched to exclusively using ceramics a while back, except where there were specific circumstances that made a tantalum or electrolytic a more sensible choice.  They’re small, they’re affordable, and they have no polarity issue.  This great article from Maxim made me pull out some datasheets, though, and take another look at something I’ve just been adding and ignoring for ages: Temperature and Voltage Variation of Ceramic Capacitors, or Why Your 4.7µF Capacitor Becomes a 0.33µF Capacitor.

Update: The most common large ceramic caps I use are some 10µF 0805 16V X5R ceramics from AVX.  No mention whatsoever of capacitance loss over voltage in the 3 page datasheet, and you have to dig down to page 83 of the generic information for their entire family to find a single chart on this (shame on AVX) for such significant information and they only discuss AC, without going into any detail over package sizes and with DC voltage, etc. … though perhaps I just missed something?  Seems like a good experiment to pull some caps out and check the numbers myself!

I recently found myself having to make a footprint for a BGA/CSP-style part with 0.5mm pitch (from the center of one ball to the center of the next) and 0.3mm ball diameter.  That doesn’t leave a lot of room, and you need to be extremely careful with your footprint design if you want to have decent yields, which is where good documentation and manufacturing suggestions come in handy.  The chip maker behind the chip I was using didn’t see fit to provide any suggestions (shame on them), but I remembered an excellent app note from NXP on designing footprints for their BGA chips and found one with identical pitch and ball dimensions (TFBGA180 in this case).  If you’re ever wondering how to properly design a BGA footprint, AN10778 is one of the better starting points out there.  (Homework assignment: some of the FPGA vendors also have good app notes on this if you want to do some digging.)

Planning on powering that next über-low-power board of yours from a measly CR2032 coin cell?  All the power to you (ug), but you might find this helpful app note from TI worth reading to understand exactly what the limitations of coin cells are, I.e.:

When designing a small wireless sensor node to be powered by the popular CR2032 coin cell, some sources claim there is a 15mA “limit” and that drawing more current is not possible or will “damage” the battery. This may give the impression that at 15mA everything works perfectly and battery capacity is great, while at 16mA nothing works. There is little public information available to explain why such a limit exists (if it indeed does exist), and little information explaining why 15mA would be a “magic number”.

If you order a single PCB from one of the many pooling services out there (shout out to Laen’s OSH Park, for example, which I recently tested  myself) … you usually get back individually cut out boards that follow the dimension layer of your submitted PCB or gerber files.  That’s great if you’re assembly them by hand … but if that amazing prototype turns out to be a gold-mine you’re going to need to deal with machine assembly using a pick a place, and those tiny little one-off boards aren’t going to get you very far.

Pick and place machines place multiple boards in one go via something called panelization (panelisation for some of us) … essentially, taking one design and laying out multiple copies of it in rows and columns, then putting a frame around the panel and inserting some drill holes or V-scoring (or both) to later break the individual boards out from the support frame.  Sounds easy … but it’s a lot more challenging than you might expect to take that one board and design a panel properly, and there’s a but of inside knowledge required to do it properly.

Enter today’s EE Bookshelf entry, courtesy Tom Hausherr (who’s amazing blog we highlighted not too long ago).  There’s precious little information out there on proper panelization — what size for the inner cutouts between the frame and the boards, what drill holes for the breakaway tabs, and don’t forget the fiducials and mounting holes on the frame, etc..  This blog entry is definitely the best I’ve come across on the subject (and I’ve picked up a lot of books over the years): PCB Design Perfection Starts in the CAD Library Part 19:  PCB Breakaway Panels.  The information is SW neutral, and it gives you all the key measurements you need for your first panel to be a success!

Have some tips on doing this the right way yourself?  Post them up in the comments below!

Here’s the first of a series of posts where Andrew Gibiansky will address the physical principles underlying the abstractions that allow complex modern computers to function using detailed discussions and infographics such as the above to help connect the physical components/EE principles to where they reach over to assembly language and the basis for computer programming:

For many years, I studied computers without ever understanding how they work. On the inside, a computer is a monstrously complex beast, with layers upon layers of abstraction which ultimately boil down to electrons running through silicon, obeying the fundamental laws of physics. We’ve built up so many layers of abstraction that the vast majority of people using computers – even the vast majority of highly technical programmers – don’t know (and don’t need to know!) how it all works on the inside. But while understanding every single layer of abstraction to its fullest extent is practically impossible, it’s incredibly fascinating how modern computers are built and what physical principles allow them to function.

In this series of blog posts, I’d like to introduce you to many of the layers of abstraction bridging the gap between the laws of physics and assembly language.  Given the rather large scope, I’m going to end up leaving out a lot of information about every topic I discuss. Just note that every topic I mention has, essentially, a field and a half solely devoted to it. With that said, let’s begin with circuits.

Read More.

The always knowledgeable Chris Gammell (of The Amp Hour fame) pointed me to one of the best CNC and mold making resources I’ve ever seen online, and while this may be old news to some people there’s an incredible collection of advice holed up in this article book: The Guerilla Guide to CNC Machining, Mold Making and Resin Casting by Michal Zalewski.  If you’ve ever wondered how you can use a CNC mill to make custom gears, or just what the limits of a CNC mill are, this is probably the best online resource you’re likely to come across (note: feel free to prove me wrong in the comments below!).

I’ve been on the fence for years between buying a mid-range 3D printer or a mid-range CNC mill.  They’re different solutions to (in my case) similar problems, but their respective advantages and disadvantages come close to cancelling each other out, which is why I still haven’t taken the plunge on either.  My only interest is create prototypes of enclosures, or small mechanical parts like gears, and molds for silicon keypads or resin prototype parts, etc.  A CNC mill is more flexible since a mid-range mill can also cut soft metals, and has a better finish than many affordable 3D printers, but you’re also more limited with the shape of the interior area compared to 3D printing.

If anyone out there has anything to contribute on the debate (in my head) between 3D printers and CNC mills, my guess is I’m not the only ‘maker’ in this same boat … where finish quality and price are important and it’s probably not realistic to just ‘buy both’.  Feel free to chime in below with your own thoughts and experience!

I don’t remember how I originally stumbled across this, but looking through my ‘Embedded’ bookmarks I was reminded of this interesting and relatively accessible introduction to basic sensors.  Rather than focusing on sensors, though (temperature, humidity, etc.), it focuses on something much more interesting … the ways that many sensors actually measure things (via resistance, capacitance, inductance, etc.).  If you’re looking for a basic primer of the advantage and disadvantages of different ways of measuring the physical world around us (visible or not), this is as good a starting point as any!  Head over to Introduction to Sensors by the University of Exeter for more information.

If you’re really interesting in the many details that separate functional PCB designs from professionally designed boards, you can’t do much better in the free department than reading Tom Hausherr’s Blog (Mentor Graphics).  There’s some exceptionally good info in there.  Find any favorites … please post them up in the comments below for other people to benefit from them!

A while ago I backed the Solar Factory campaign on kickstarter, and I’ve been following the Solar Cells, Fuel Cells, & Batteries course from Stanford, so this article from Spectra-Physics really interested me when I came across it: Manufacturing c-Si Solar Cells with Lasers.  It’s a very accessible overview of some manufacturing options to produce solar cells, though it’s obviously still a growing field with a lot of new, interesting research going on.

If I had to come up with a list of the top 5 technical resources ever, the CC-Antenna-DK kit from TI (formerly Chipcon) would definitely place pretty high up there.  Antenna design is a time consuming, repetitive task requiring an excellent attention to detail and understanding of the materials you’re working with, more than a few calculations, and a dash of trial and error.  $49.00 for more than a dozen pre-rolled, easy to test, easy (and legal!) to steal antennas is seriously one of the best deals I’m yet to come across.  Heck, they even throw in short, load and open boards if you’re in a pinch and need a cheap way to calibrate your VNA!  The frequencies are a bit biased towards Europe with a lot of 868MHz boards (no surprise since Chipcon was Norwegian), but if you’re looking for an excellent primer in antenna design, including complicated ones like mixing 868MHz and 2.4GHz in the same antenna, you’ll definately benefit from reading some details on this excellent board.  See Design Note DN031: CC-Antenna-DK and Antenna Measurements Summary.  And if you’re interested in RF at all, seriously … buy one now before someone at TI wises up and starts charging the $495.00 this board is really worth!

Ever wondered what that 300th variation of a SOT-package actually looks like and what’s different about it?  These package posters from NXP might shed some light on some of the more common variations!

• General Discrete Package Poster
• Discrete Flat No-Leads (DFN) ICs

There’s nothing particularly fancy about trimpots … but my weakness for hand-drawn books and illustrations kind of forced me to publish Bourn’s “Best of the Trimmer Primers“.  Need a trimpot with a safari hat and a gun?  Bourns has got you covered!

While I rarely work with signals much beyond 100MHz (SDRAM, etc., usually being the limit), it never hurts to try to improve your understanding of high speed layout.  By far the best book you can buy on the subject is High Speed Digital Design: A Handbook of Black Magic by Howard Johnson and Martin Graham.  That said, I found myself routing some USB signals that I wanted to have matched since the USB connection is high speed, and after routing the board I took a look around to see what advice I could find before signing off on that part of the board.  There are some excellent replies over on stackexchange to “How should I lay out timing matched traces“, with a valuable reminder to step back and consider the scale of your board, and that 1mm length on your PCB probably equals about 5 picoseconds in reality!.  Sometimes is helps to just zoom out, look at something at life size, and realize how small that little green board really is!  The other good source of information I found was  Board Design Guidelines for PCI Express Architecture.  Some very good tips on layout and real-world technical considerations that aren’t always cleared explained in more academic texts.  Any suggestions yourself?  Feel free to post them in the comments below.  I’m as happy to find new sources of expert advice as anyone!

As a sidenote, the new Meander tool in Eagle 6 is very useful for this.  You can use it to click on a trace and it will tell you the exact length, which makes it much easier than having to type ‘run length-freq-ri.ulp’ in Eagle 5 and try to find your trace in the other 300 listed by name!

Limor sent this wonderful little document my way after we had some problems with a 0.4mm pitch footprint I made, just to see if we could improve the paste layer and reduce the bridging a bit: Root Cause Failure Analysis of Printed Circuit Board Assemblies. Despite the organization not being the best — a clearer separation of problems from potential solutions/recommendations would have been helpful — there’s a lot of useful information in here.  It’s often easier to understand the kinds of problems you encounter working with fine-pitch devices looking at photos.  Unfortunately, X-ray images of things like ‘ball-drop’ with BGA reflow aren’t easy to come by, nor post-mortem, sawed-in-half BGA packages, etc.  If you’re fascinated by the many ways boards can die a premature death, this is a goldmine of info.

It’s a while since I’ve included a book in EE Bookshelf, but I’ve purchased a lot of books lately on the manufacturing process — specifically material science and manufacturing techniques — so I figured a few book suggestions are long overdue.  There are a decent number of good books out there on manufacturing and industrial design (I picked up six after collecting some recommendations), but by far the best book of the bunch to me was Manufacturing Processes for Design Professionals by Rob Thompson.  It wasn’t the cheapest book of the bunch, but it’s an exceptional value and a fascinating read if you have any interest in moving from a populated PCB to a finished product, but don’t know what options are out there to wrap around your PCB.  Thompson not only goes into sufficient detail on many modern manufacturing processes — describing their relative strengths, weakness, associated costs, etc. — but the book relies heavily on invaluable technical illustrations and photographs, including some inspiring examples of products that were manufactured with the said process.  If you’re interested in getting inside factories around the world, this is the cheapest quality tour you’ll find!

I’m not sure how this fits into the usual flow of things with EE Bookshelf, but given all the interest in the Raspberry Pi, I figured there are probably a lot of people out there for whom this might be their first foray in Linux.  While there are a lot of good books and resources out there for Linux, it can be tough to wrap your head around which commands are available from the console, etc.  The GNU Coreutils documentation does a decent job of showing what should be included in any distribution, and will hopefully help people get a bit more familiar with the command line.  There’s also a PDF version for offline browsing.

If you’re just looking for a concise cheatsheet, there’s lots out there, but this one from FOSSWire should get you started pretty quickly.

Every question, curiousity and doubt anyone has ever had about Peltier cells seems to be addressed in this FAQ from Tellurex.  I stumbled across it over the weekend, and yes …  it’s marketting material, but there’s still a lot of good information in there if you want to keep your favorite that beverage nice and chilly, or scavanage free energy using a heat differential.  If you didn’t know about Peltier cells, now’s a great time to look them up and figure out some fun things you can do with them (post the details below … I love these things)!

Since we’ve been busy adding quite a few I2C sensors and breakouts lately, I thought this technical overview of the 2-wire “Inter-Integrated Circuit” bus might be handy.  I2C isn’t fast (typically limited to 400kHz in most real-world situations), but it’s convenient since it only requires two pins and more than 120 devices can be connected on the same bus, address space permitting.  For low-pin count devices, it can be a real life-saver since you can hook an OLED display, a DAC, a 7-segment display and 16 servo motors up to your Arduino with a measley two pins and some careful coding!  The full bus specification is available from NXP in UM10204 – the bus was created by Philips, whose semiconductor branch later became NXP — but the more concise information from Embedded Systems Academy might be easier to digest as a starting point.  The FAQ has some very good information in it.

Filter design is one of the most interesting aspects of electrical engineering. This is in no small part due to the wide variety of topologies and configurations available for each kind of filter. In the early days, filters were passive — made exclusively of resistors, caps and inductors — but since the advent of low-cost, low-distortion op-amps, active filters have become much more prominent. One of the greatest benefits of active filters is that they don’t require inductors, so size and weight are kept to a minimum.

App note 1762 from Maxim is a good survey of common active filter topos, and explains the strengths and weaknesses of each one. The basic topologies can be extended with more advanced techniques, such as voltage control or switched capacitors, but the underlying configurations are the same.

If you ever have to move from a single prototype to machine assembly using automatic pick and place machines, you’re most likely going to have to deal with panelizing your PCBs.  This just means taking your basic PCB design, and multiplying it a certain numbers of time in an X/Y pattern.  What happens is the machine operator will program one board in, and they can simply step and repeat across all the other boards on the panel, significantly speeding up assembly since solder paste can be applied to many boards at once, and minimizing time lost shifting boards in and out of the pick and place.  While there’s some information out there on the panelization process, it’s not spoken about nearly as much as it could be, which is why I was glad to come across this article (from the very read worthy Printed Circuit Design & Fab): The PCB Array, and Why We Use It.  Definately do some more digging around on that site if you’re into PCB design, since there’s a lot of great stuff and you can browse back issues online for free!

If you’ve ever been curious about how to properly define BGA footprints or what patterns to use for BGA ‘fan-out’ (breaking the signals out from underneath the chip), AN10778 from NXP provides a lot of helpful information on some common BGA packages.  1.0mm and 0.8mm pitch BGA actually aren’t that bad to work with and you don’t need to pay for super-exotic tolerances from your board house, plus bridges underneath are reasonably rare if you have a good paste layer and footprint.  It does get a lot more challenging and expensive from 0.65mm and lower, but NXP provides some good suggestions and clear numbers on how to handle both situations.

This is one of those things that just seems to be so stunningly obvious that you should have thought of it yourself every time you make a dense footprint (say 0.65mm BGA or less) … but I’m ashamed to say it never even crossed mine, despite having made several BGA and CSP footprints and boards.  Trying to make some tight BGA pads with a specific diameter?  0.3mm in Eagle may not turn out to be the 0.3mm you’re expecting on your PCB, and Andrew Zonenberg explains why in PCB Fab Characterization on his excellent blog Silicon Exposed.

I came across this presentation from Embedded World today that gives a brief summary of the various ARM Cortex chips (and there are a lot of them in the Cortex M, R and A families!), but specifically the ARM Cortex M series.  One of the biggest advantages of using ARM is that everything tends to be forward compatible, meaning that every instruction in the teeny-tiny Cortex-M0 core is supported in the larger Cortex-M3, and all the M3 instructions are included in the even larger M4 (which adds single-precision floating point acceleration, and some basic DSP-type commands), and so on.  If ARM has really taken off in recent years, this easy migration and transfer of knowledge is a big part of it, combined with the huge number of chips available from a half-dozen manufacturers (TI, NXP, Atmel, ST, Freescale, Energy Micro, etc., etc.).

I’m kind of curious how much interest there really is in low-end ARM chips though — essentially the Cortex M family, since the A-series is far more complicated to work with and isn’t aimed at solving the kinds of problems most hobbyists are working on.  I learned embedded development on ARM and it’s what I’ve always used, so I always have a hard time judging how accessible people find it, and what they’re looking to learn to get started with ARM.  I’m convinced the low-end ARM chips are the best value on the market right now — just look at the specs for dirt-cheap chips like the Cortex-M0 LPC1114 at $1 (50MHz, 32kB flash, 8kB SRAM) or the Cortex-M3 LPC1343 (72MHz, USB, faster code execution thanks to three pipeline architecture versus single-pipeline on the M0).

Let me know in the comments below what you’d like to see with ARM, or if you’d rather see things focused in a different direction (like focusing EE Bookshelf more on PCB and HW design, or general embedded SW development best-practices).  I’m all out of favorite datasheets and books, and wondering what people want to see more of moving forward!

It’s admittedly pretty specialized, but there’s very little accessible information out there about ARM assembly in general (documentation from ARM always feels  sadly neglected and half-hearted to me), but I came across this wonderfully precise and accessible summary of inline ARM assembly with GCC that really surprised me.  It’s not something you’ll use all the time even if you do a fair amount of development on ARM, but it’s a treasure chest of information if you’re trying to figure out how to optimize some particularly picky function: ARM GCC Inline Assembler Cookbook.

While trying to reduce the overall noise level on some boards I’ve been working with, I was looking for a reliable way to first measure the noise level and figure out how best to address the problem.  If you can’t measure or visualize it, it’s hard to fully understand it and reliably fix it!   Looking for a solution to measure the quality of the GND planes, etc., I came across this helpful appnote from Analog Devices: Measuring Ground Noise.  They present a relatively straight-forward setup based on an AD620 instrumentation amplifier that can be used to characterize the GND plane on your boards.  The article is worth reading simply because it brings up something a lot of people don’t think about doing boards design: GND.  Particularly with high-speed design (anything greater than a few MHz), or with switching-mode power supplies, etc.,  it’s important to think about the return paths on your boards in addition to the signal transmission lines and part placement.  Howard Johnson’s High Speed Digital Design is probably the best book (that I know of!) on the subject, but I haven’t seen very many circuits showing how you can measure noise on the gnd plane, so I thought it was worth sharing the above PDF.

I’ve been doing a lot of power supply testing lately, using both switching a linear supplies.  Since it’s not something I’ve had to do often in the past (I’m an apps person, not so much a test engineer) … I thought it was worthwhile to spend a bit of time digging around for app notes on accurately characterising linear and switching supplies.  Long story short: AN 372-1 from Agilent was one of the more useful ones I found (though I’m sure there are dozens out there). Short story long: read on …

I recently picked up a programmable load to make the measurements a bit less tedious (a B&K precision 8540 if you’re curious), which is an expensive tool for what’s in the box and perfectly doable for $50 as a DIY project, but a wide-ranging professionally built unit becomes your bench-top BFF when testing power supplies with it’s convenient constant voltage modes, etc., and direct visual feedback on supply voltages.

Armed with a good 6 1/2 DMM (an Agilent 34410A courtesy the ever generous Adafruit Elves last Christmas!), the B&K 8540, my trusty oscilloscope (Agilent MSO-X 2024A), and a DC power supply (Rigol DP1308A) I was able to quickly measure the efficiency of switching mode supplies across a variety of loads and with varying input voltages.

I’d actually like to put a quick tutorial together of getting basic measurements with a linear or switching supply, but if (like me a month or so back) you were wondering about the best setup to accurately measure certain characteristics of your power supply, the app note above from Agilent is worth the time, and include clears explanations of the benefits of different types of test equipment (depending on your budget), and what tests are available to you with the equipment you have.

I’ll try to flesh this out a bit more in the future, and take a couple photos of the setup I’ve been using (gear pr0n), but if anyone has any other suggestions on related app notes or white papers, feel free to post the in the comments section below!

Having a hard time trying to figure out whether that FET can handle enough current for your project?  AN11158 from NXP might help clarify some of the many parameters that you need to take into account that are often overlooked.  The Safe operating area, for example, is an important one that often gets skipped and people just look at the best-case scenario marketing numbers on the front page of the datasheet: “The Safe Operating Area (SOA) curves are some of the most important on the data sheet. The SOA curves show the voltage allowed, the current and time envelope of operation for the MOSFET. These values are for an initial Tmb of 25°C and a single current pulse. This is a complex subject which is further discussed in the appendix (Section 3.1).”

It’s pretty rare that I come across a free book with so much good information in it, but Analog SEEKrets is definately worth a few days of any budding young engineers time.  Written by someone who clearly has enough years experience behind them to know what they’re talking about, this book — freely downloadable in PDF format — contains a lot of excellent real-world information that anyone working in the commercial world will either recognize or appreciate.  Some of the practical things I liked in the book were checklists you can give to any new team member, such as the Pre-Supervisor Checklist (p.19) to make sure someone did their homework before pushing problems up the food-chain.  (Helpful Hint:  If you want to fit in well in any engineering team and get the help you want when you need it, it’s important to make sure you’re not asking questions you could easily have solved yourself, or at least be able to demonstrate that you’ve clearly made some effort first.  I wish I would have thought of making a list like this myself working with interns, etc., as they tried to learn the ropes to make sure everyone is getting the most out of the time available in the average working day.)

There’s a lot of excellent information in this book whatever your experience level, and it’s well worth downloading and reading.  If you find it valuable, definately consider making a contribution to the author to thank him for making the book freely available online now that it’s no longer in print.

Have you ever wondered how complex ICs are really made, and how you get from raw silicon to finished chips?  I was lucky enough to spend a short but happy part of my career working at a fairly large chip fab (though as an apps engineer, nothing to do with manufacturing), but still find myself with more questions than answers on the gritty details of it all.  It was an amazing eye-opener for me to see first-hand the kind of mastery and machinery that it takes to produce finished chips, going through design, processing, testing, packaging, characterization, etc.  I think the only thing as humbling has to be seeing something like the space programs at NASA or something similar.  You can only be in awe of the level of mastery required across so many disciplines to pull something like this off.

Unfortunately, most of the technical documentation out there on this stuff is understandably written for and by people who are hopelessly above my pay grade, and I don’t understand a fraction of it (and I’m happy to admit it!).  If you’re really curious about getting an accessible look into chip design without the need for 10 years of training in semiconductor physics, though, Demystifying Chipmaking by Yanda, Heynes and Miller (published by Newnes) is an unusually accessible introduction to the topic.  (You can even get the first three chapters free in a preview off Amazon if you have a Kindle or download the Kindle SW.)  It’s an old book, but it’s still relevant today since it’s focused on CMOS, and even if they only talk a bit about the <100nm processes that are common today, the basics still apply.  Have a look at the free chapters, and see if the rest interests you!

Today’s bookshelf isn’t a datasheet or a book. It’s actually just a webpage (which I suppose would make it an ‘EE bookmark‘), but it’s one that I like a lot, so I figured I’d share it.

Which capacitors sound better than others? Or perhaps, which ones distort the least? This is a contentious topic among audiophiles, musicians, DIY synth builders and occasionally HAMs as well. I don’t actually want to wade into this argument. My main purpose in writing this post is to share with you this excellent webpage over at greygum.net. I keep this one near the top of my bookmarks folder, and make sure it stays there when I migrate to a new computer.

This page dates from 1999, but the information is still good. It’s also a great example of an awesome late-90′s webpage.

The author, Steve, performed some tests to find the D-E curve of various types of capacitors (his test set-up is described at the bottom of the page). The results can be eye-opening.

The two worst performers on that page were electrolytics (which exhibit marked hysteresis, particularly the tantalum, above), and ceramic monolithics, which have  very noticeable distortion. The one thing both electrolytics and monolithics have in common is their relatively compact size, achieved in part by use of a very thin dielectric. It’s interesting to see that the high-voltage monolithic has less distortion than it’s low-voltage counterpart, perhaps in part because the HV version has a thicker and/or different dielectric.

While I had known from long-time experience that ceramic and electrolytic caps can muck up the signal path, I never had any data (visual or otherwise) to back it up — until I read about Steve’s experiment a few years ago. As such, I tend to relegate these caps to decoupling and bypassing, and use something else for anything that touches signal, usually polypropylene film.

Check out Steve’s page and see for yourself, and be sure to check out his followup on electrolytics as well.

Admittedly, Interconnecting Smart Objects with IP [ISBN: 0123751659] by Jean-Philippe Vasseur and Adam Dunkels (of lwip and Contiki fame) deservedly made the rounds a while back, so we might seem a bit late on the draw pointing it out again here.  But while I haven’t been doing anything with wireless lately myself — that’s much more the domain of the always knowledgable Akiba over at Freaklabs, who I’m always happy to defer to – I pulled this book out again this month for something else.  It goes into a lot of valuable detail on what it takes to implement a complete wireless sensor network infrastructure and it’s well worth the investment if you have any interest in WSNs, but an unexpected bonus for me the first time I read the book is that it also gives you an excellent overview of all the little pieces that fit into IP-based networking of any type … and there are a lot of pieces in that puzzle!  If you’re interested in getting started with IPv4 or IPv6 and Ethernet, there is a lot of information out there, but I was surprised to find Vasseur and Dunkels’ summary one of the most readable and accessible introductions that I’ve encountered, giving you a good understanding of the glue that holds everything together.  Trying to add wireless OR wired network connectivity to your next great idea of project?  You could do worse than spend an evening with this book and a good highlighter.

Alright … I admit the title sounds repulsive and makes you want to click away, but being someone interested in design as well as engineering, I sometimes look at sites like smashingmagazine just to see if there’s anything worth procrastinating over for a few minutes.  The article above is well worth a read.

Why am I posting this here on Adafruit under EEBookshelf, though?  Because I see a lot of questions about how to run your own kit business, and in my opinion — and I know Phil and Limor share it — good, precise, expert information is the best marketing you can ever  provide.  Give people the information they want in a regular, concise and accessible way, and as long as your products are aligned with people’s needs and expectations, the customers will inevitably come to you.  Good information is, in my opinion, the key to running a successful business, though it’s obviously more complicated than that.  Why was I an Adafruit customer long before I was an Adafruit employee?  Because they also treat their customers, their employees and their business partners fairly, and they play by the unspoken rules that reflect the best of what open source hardware should be.

Working in the kit business is exceptionally hard.  The support, shipping, sourcing and stock maintenance, site development, and a lot of other S’s make for long days, but if you’re serious about getting started, the best thing you can do is focus on a specific area that isn’t being treated well, and provide the kind of expert information that makes people want to come to you.  There are some huge gaps in knowledge out there, and people are thirsty for information … the best way to succeed online is to feed that need in a regular, readable format.  This article from smashing makes some good points about getting started doing it well from day one.

If you’ve ever had any questions about how to efficiently add CRC (Cyclical Redundancy Check) to your code, this PDF from the author behind the excellent Hacker’s Delight (Henry S. Warren) is a great start.  If you’re interested in writing efficient code, the rest of the site is worth a browse as well: http://www.hackersdelight.org/

I’ve been having a lot of noise problems lately, both with boards and with equipment doing testing in the lab.  As a result, I spent a bit of time looking around at some solutions to help reduce the noise on existing boards (basically filtering since the noisy board itself is out of my control), and came across a good app-note from Freescale on noise reduction techniques.  A lot of this is focused on PCB design, and much of the information is available from other sources, but if you’re serious about designing good PCBs, it’s worth a read and is more accessible than some other documents out there on the subject.  A lot of noise reduction comes down to good grounding, and unfortunately it’s not something that’s always done half as well as it could be (and I’ve been as guilty of this as anyone, though I’m making a conscious effort to consistently improve).  High frequency return paths need to be kept in mind while routing and laying out your PCBs — not after the fact — since you can’t do much about it when you find out you have noise problems and poor performance once the PCBs come back and get assembled.  See AN1705: Noise Reduction Techniques for Microcontroller-Based Systems for some good tips and suggestions.

While this is neither and app-note nor a book, I’ve been interested in anti-aliasing for quite some times since A.) UI design is just kind of a fetish of mine, and B.) There’s almost no useful information out there on this stuff in a form that’s appropriate to low-cost embedded systems.  Which is where this weeks EEBookshelf entry comes in: a very old article (think Quick Basic, ya … that thing you used back in High School or earlier) explaining the basics of anti-aliasing pixels.  While the text is obscenely small, the explanation is very clear, and it’s easy to transfer this to your own firmware in C or whatever else.  Now to work this into some of my own projects and see how far I can stretch the 72MHz on the LPC1343 and the few hundred KBs I have left! See QB EXpress: Issue #22 for this article from Eclipzer.  My advice … if you find this useful save it to a PDF somewhere special.

Recently picked up an oscilloscope from Adafruit, or wondering what you might use one for if you did have one at your disposal?  Tektronix ‘XYZs of Oscilloscopes‘* gives a great introduction to the basics of scopes, and the content is all pretty much universal.  Having a scope opens up a world of debugging and learning opportunities (seeing is believing!), but the learning curve can be a bit steep if it’s your first time in front of one.  Know any other good tutorials to help people get started?  Feel free to post them in the comments below!

*Note: Registration is required to download any of the primers and guides from Tek, but it’s all free and there really is some good information in the different guides.

If you’re anything like me, your brain is decidedly more SRAM than it is EEPROM, and any time something interesting happens, the previous memory contents disappear and get replaced by something else.  Which is why I always keep a little 2-sheet-per-page double-sided copy of this in my backup brain engineering notebook: Reference Guide & Formula Sheet for Physics by “Dr. Hoselton & Mr. Price”.  I had to dig around to find the original source for this since I’ve had it so long and have no idea how I came across is (no doubt googling some seldom used formula), but Dr. Hoselton seems to have made this while working at Trinity Valley School (See Dr. Hoselton’s Physics Page).  Makes me wish I had the kind of teachers who cared enough about this stuff to put these kinds of things together, but that’s another post altogether.  It’s a nice little reminder of all that stuff you can never remember, and should make that next physics-based iPhone app a breeze.  There are a handful of these floating around (feel free to post others in the comments below!), but this is what ended up in my little notebook many years back.

Looking for a poor-man’s capacitive touch sensor using only the ADC on your MCU and a couple discrete components?  NXP shows how it can be done in AN11023: Capacitive Touch Sensing using the LPC11xx (here for .zip file with code).  (Note: While this app note is aimed at the ARM Cortex M0 LPC11xx series of chips, the information is easily transferable to any other MCU.)

Analog Devices (who have a lot of great app notes locked up in their stable) have made available in PDF format enough good information and reading material to keep you busy for the rest of the winter while you wait for warm enough weather that you want to step outside again.  Their Linear Circuit Design Handbook has a lot of excellent material, and can also be ordered in printed format if you’re still a fan of a good old highlighter and notes in the margin like me.