Pocobor.

I came across a blog post about the Arduino yesterday on an industrial design blog, Core77. I was definitely surprised to see an entire post dedicated to the Arduino, an open-source electronics and embedded software platform which targets DIY’ers and non-engineers who want to build/hack smart products, on a blog that focuses primarily on product and industrial design. The Arduino is getting great exposure and keeps popping up in places I wouldn’t expect. I’m excited the conversation about and accessibility of smart product design is spreading.

Why I’m Excited

The Arduino provides scaffolding for outsiders and non-embedded system designers, to understand and explore smart product design. It doesn’t matter what your experience or skill level is, Arduino provides an extremely accessible interface for people to get started, from both a hardware perspective and a software perspective. The electronics come packaged and ready to go, with easy to use connectors and easy to understand labels. Several vendors even provide drop-in electronics, called shields (click for a list of shields), which provide specific functionality (ie motor control) to the user with little effort. The free software interface provides a level of separation and simplification from the Microcontroller (MCU). Users have access to easy-to-understand functions and don’t have to familiarize themselves with specific registers and modules of the MCU.

People you wouldn’t expect are getting their hands dirty and cool things are happening. The internet is ripe with cool projects people have put together on their own and there are a ton of project examples and project guides to get people involved. Everyday people are building their own smart products!

Why This Matters To Me

Most importantly, the discussion is finally spreading to people in different walks of life! The exposure allows people who aren’t necessarily engineers to see the possibilities available in smart product design. Different perspectives can easily join the brainstorm. I’m convinced more wild and crazy ideas will be born, not only in garages but also in the office. Ultimately, better products will be designed.

And hopefully the realization of what is possible with a simple open source tool will lead people to imagine what is possible from a professional service firm (ahem Pocobor) and the value we offer. If nothing else, it helps me describe what I do and how technology is being incorporated into new products we use in our everyday lives.

Get Involved

We’ve even put the board in an open source project, called PedalOn, we’re completing for a client to allow customers to modify or rewrite the system software. We’ll talk more about this project in the coming weeks.

An Arduino is even inside PedalOn, a Pocobor project.

I encourage anyone not directly involved with smart product design to get their hands on one of these and start playing. The barrier to entry is low; you can get one for less than $30 from Sparkfun. Or try another distributor - for a full list of distributors look here.

Recently I have been using LEGOs to prototype a gearbox for one of our clients. Every time I dig into a box of LEGO parts and hear that familiar rustle of plastic components, I am struck by how useful LEGOs are for creating rapid models of mechanical systems. And while I normally use LEGOs to build smaller-than-life models of larger systems, a group in England is constructing a full-size house, complete with LEGO shower and toilet, from millions of LEGO blocks! Seriously! Check out more here.

I am very interested to see how they incorporate LEGO’s “Mindstorms” robotic technology into the house. I can imagine motion sensing light control and a LEGO HVAC system powered by a Mindstorms computer. Of course, they may have bigger issues to worry about, such as waterproofing the roof.

The MICROFACTORY MOW

I came across a cool post about the MICROFACTORY MOW by DaeKyung Ahn on Core77 yesterday that really got me excited. DK’s MOW is an in-home manufacturing machine to produce products from scrap flat stock you may have lying around your house, like the brown cardboard/corrugate that lies around after the purchase of any new product. The concept relies on a community open source web-portal where users freely share/upload their custom product ideas so others can search and download the plans to their personal machine for home machining. Although it wasn’t clear to me how the machine instructions would be translated from the ideas, I can easily imagine a simple sketch pad interface on the website to translate a pattern to control code that users with basic geometry skills could master.

MOW In Action

I am really excited about empowering everyday people to explore design in their own homes with simple manufacturing tools and by utilizing waste material. (Brian discussed this in his Democratization of Design post back in January related specifically to the RepRap machine.) So many people have amazing ideas that remain dormant because they don’t have access to the tools or don’t realize how easy it can be to prototype those ideas - maybe MOW can help bridge that gap. And the open source web-portal provides an extremely low barrier to entry; you can try out existing proven designs before experimenting on your own. Or maybe you just want to turn your garbage, bound for a landfill, into something valuable.

The video captures the iterative design process and a demo of the finished prototype. Check it out:

Quad Flat No-Lead Chip

This post builds on my previous post, which discussed hand soldering surface mount passives and QFP chips. As I mentioned last time, these are vital skills for any circuit board designer – the ability to modify or rework a board is integral to a time-effective development process.

Today I’m going to talk a little about hand soldering QFN chips – these chips are extremely small, which can be very useful for tiny, densely packed boards but makes them a little difficult to work with by hand. As a reminder, these posts are building on some excellent tutorials created by Curious Inventor. The relevant information for QFN chips can be found here.

Quad Flat No-lead (QFN) chips are like QFP chips but don’t have leads extending beyond the sides of the chip; instead, they have pads on the bottom of the chip. For these types of chips, however, my experience diverges from the Curious Inventor tutorial a bit. Their method certainly works but I have found it easier and more effective to use solder paste with the hot-air station as opposed to tinning the pads with actual solder. In terms of equipment, you will need good tweezers, a hot air source (either hot air gun or dedicated soldering rework station), solder paste and solder flux.

The technique that we have found works best involves the following steps:

1. Clean the chip and circuit board pads and make sure that the chip can lie flush on the board.
2. Apply solder paste on all of the circuit board pads using an applicator syringe (below). Make sure there is paste on all pads but use sparingly – too much could lead to shorts.



Solder Paste

3. Place the chip in the correct position on the board using tweezers. The solder paste may make it difficult to see if the chip is correctly aligned with the pads but use silkscreen markings on the PCB to index the chip’s location. Fortunately, surface tension effects with the large center pad will help the chip self-align as long as it is close to being in the correct position. Make sure that the chip is firmly pressed down on the board and is flush or as close as possible to it.
4. Preheat the chip and board using the hot air gun to about 200 degrees Fahrenheit. Depending on your setup, this can be accomplished by varying the heat and flow rate of the gun or changing the distance between the chip and the hot air nozzle. The chip should remain in place without fixturing as long as the board is horizontal and the hot air is coming orthogonally from above.
5. Turn up the heat to about 350 or 400 degrees. Once you see the solder melt, continue applying heat for up to 15 seconds more. You want to be sure that the solder has fully melted and reflowed to all of the pads but too long and you can overheat and break the chip and / or board.
6. There will probably be excess solder along the sides of the chip. Apply flux and use a soldering iron and wick to remove this solder. Try to remove all accessible solder, not just visible accumulations.
7. Verify that no shorts are visible with a loupe or microscope. Professional shops sometimes use x-rays to check underneath the chip and if this kind of equipment is available, it can be very helpful. However, in our experience it is not necessary.

QFN chip mounted on a PCB

I’d love to hear what works (or doesn’t) for others so feel free to add your two cents in the comments.

Happy Soldering!

Context

I originally set out this week to put together a tutorial on hand soldering surface mount components. Then, I realized that there are some excellent resources already out there and decided instead to write a post supplementing one of the existing tutorials with some things I have found based on our experience at Pocobor. Accordingly, this post builds on some excellent tutorials from the Curious Inventor website:

Why Should You Care?

If you aren’t familiar with how to hand solder surface mount components and you are reading this blog (which potentially indicates some level of interest in electronics), you should definitely think about trying it out. It is a very valuable skill for several reasons: (1) assembling your own board can be cheaper than outsourcing assembly (although it can also be fairly time-consuming), (2) being able to modify or rework a board is integral to the development process. Prototyping inherently involves some trial and error / experimentation and the ability to perform a little circuit board surgery can save a ton of time and money on new board iterations.

Passives

I don’t have much to add to the Curious Inventor’s take on this – the two things I would just emphasize are that (1) I generally find that flux is unnecessary for components that are 0805 or larger, and (2) very little solder is needed on the pad to tack down the first side of the component (too much solder actually makes it harder to align the component and get a good joint). In addition, I would point out that I generally find that a very fine-tipped soldering iron is more effective for most operations than a blunt or dull tip.

QFP Chips

Quad Flat Pack (QFP) chips are the little chips with legs coming off the sides that look like little spiders or insects. I endorse most of the points made in the tutorial and have a few of my own to add:

Happy Soldering!

What can we accomplish in a day? Can it prove our value to a dream client? We had the opportunity to find out.

We were referred to a project manager of a dream client about a potential project. Once we heard a few minor details, our minds started circling around the problem and how we could quickly put something together to fill their need. When the project didn’t materialize as we had hoped, we decided to turn it into a 24 Hour Design Challenge. We took what little we knew about the project and built something in the next day that would showcase the value we could create in a short amount of time. 24 hours later, this was the result (click the detail photos below to enlarge):

It’s what we do best: integrate electronics, mechanical systems, and computer intelligence to create modern interactive products … all in a day’s work.

Most of Pocobor’s projects involve integration of electronics and software into a physical, mechanical system. Any single weak point in this integration can jeopardize the entire design; with the individual subsystems already complicated enough, we can’t afford to have weak interconnections between them. Of all the possible failure points, there is one that stands out to me as being both particularly devious and simple to mitigate: loose wires. A loose electrical connection can be extremely difficult to find and can wreak havoc on any electrical system. But there is hope for all those loose wires out there — strain relief!

I like to think of strain relief as a mentorship program for loose wires. Without a sturdy companion to provide guidance and assistance, a loose wire may find itself getting caught up in undesirable activities (maybe a rotating motor shaft or a passing foot). Sometimes this mentor is another, larger wire — there’s safety in numbers — or sometimes it’s a sturdy mechanical component in the system. A link is made between the loose wire and the mentor, and future snags are much less-likely to cause significant damage. Sure, our little loose wire might get pulled in dangerous directions, but his mentor will be right there beside him to provide support.

Three simple ways to add effective strain relief to essential wire connections in rough prototypes: zip ties, heat shrink, and hot glue. I’m focusing on rough, quick prototypes in this post (hot glue probably isn’t the best solution for manufactured products), but many of these techniques can apply for finished products as well.

Zip ties can significantly reduce the forces on electrical connections. Here, the red zip tie is linking the large and small black wires together well below the electrical connection within the red wire nut.

The yellow zip tie shows the strain relief for the wire bundle leaving the breadboard. In this case, I’ve zip-tied the bundle to a hole drilled in the edge of the breadboard.

The yellow heat-shrink joins the small black wire to the larger grey wire to reduce forces on the electrical connection. Heat shrink used in this way also reduces the chance of an electrical short circuit.

This very rough prototype shows the value of hot glue for strain relief. In this example, I’ve soldered small wires (green) to a surface-mount sensor (small and delicate connections!) and have used hot glue to keep them from moving around and breaking loose. You can also see the strain relief for the green wires’ heat-shrink connection to the wires leaving the board. Those (larger) black, white, and teal wires are going to protect the delicate connections if they ever get yanked.

Loose wires can get snagged, pulled, bent, melted, or trampled. Whether you’re connecting to a breadboard or a PCB, wires that enter or leave the electronics should be properly strain relieved. The small amount of time it takes to strain relieve a wire can save hours of stressful troubleshooting and/or repairs down the road. The important thing to remember is that strain relieving wires is not necessarily going to prevent them from getting yanked or twisted or snagged, but it will prevent the force of that snag from reaching a critical electric connection in your design.

As Engineers, there are many things that we can’t easily control (stray electro-magnetic fields, the weather, apostrophe usage). Therefore it is essential that we take relentless control of the variables within our grasp. The universe tends to favor the chaotic monotony of loose, poorly-managed wires — fight back with strain relief!

The internet democratized information by making it orders of magnitude more affordable and accessible. New technology like the RepRap system shown above (http://reprap.org) has the potential to enable a similar revolution for hardware development by bringing similar scale improvements to the accessibility and affordability of mechanical design and prototyping.

RepRap is an open-source self-REPlicating RAPid prototyping machine. If you’re not familiar with this type of device, picture a printer that creates 3-dimensional objects instead of 2-dimensional pictures. Instead of depositing ink on a sheet of paper, 3-D printers deposit layer upon layer of material to make robust, 3-D parts. They are an invaluable development tool for prototyping and mechanical design because they allow a designer to bring his or her vision into the real world extremely quickly and cheaply so it can be tested and iterated on. Furthermore, they enable intricate, customized designs and shapes that can be created in a fraction of the time and without the labor required for traditional machining techniques.

Traditionally, rapid prototyping setups (e.g. FDM or STL) can cost hundreds of thousands of dollars or more. Engineers and designers have been restricted to using specialized prototyping job shops, who own these setups, to create rapid prototypes of their part designs. However, the design for the RepRap system is freely distributed under the GNU General Public License and (this is the really cool part) the system has been designed and developed so that it can create all of the non-standard parts needed to build a copy of itself. Add this to a cheap Bill of Materials due to some great design work by the people who developed the system and you have a 3-D printer that is relatively easy and inexpensive (on the order of hundreds of dollars) to obtain and use.

What does this mean? For starters, this means that thousands of designers who previously couldn’t afford rapid prototyping systems, both in the US and especially in other parts of the world, can now add small scale 3-D printers to their labs. Moving the systems in-house drastically increases productivity compared to dealing with external job shops and the associated details and headaches. It also creates the potential for a leap forward in the reach of the field of mechanical design. With easy access to sophisticated and powerful tools, engineers and designers will be able to utilize customized and optimized parts and systems instead of being restricted to readily available but non-ideal components.

Personally, I can’t wait to see what people start doing with these and to get one and try it out myself!