Tie And Jeans

Archive for the tag “electronics”

Under the Kit Frosting

Kits promise a grimoire for electronics. Make these symbols, perform these rites, and transform yourself into someone with the understanding you desire. Fundamentally, this is a pattern I believe in, for electronics and all other craft learning. Grimoire learning demands that novices become careful observers, that they seek out and identify useful patterns & sub-patterns. I’m frustrated when features of the kits obscure the subject itself.

To be commercially viable, kits need to present an attractive picture for parents/teachers/schools. These are a distinct market that I think of “aspirational purchasers,” and they’ve always been a central target for edu-tainment material. To convince Aunt Ermeline to buy an electronics kit, it needs to present as both approachable and transformative. Since the purchaser will likely never even open the kit, functionality and component quality become very minor factors.

Successful kits promise to solve two problems for electronics novices; acquiring materials and connecting parts. Honestly, these are real bugbears. Have you spent time on digikey? Soldering irons connote dangerous and breadboards are confusing. Even a bag of through-hole LEDs can trigger Aunt Ermeline’s “chocking hazard” instincts. Aspirational purchasers, who are already imagining how this kit will transform a surly adolescent into a world-changing savant, are also primed for catastrophic “what if?” scenarios. To signify as “novice friendly,” electronics kits entomb components in custom plastic housings that snap, click, stack, in familiar toy-like ways.

This process adds costs and creates a perception of incompatibility between kits. These component wrappers can also hide the real information about a component or circuit. As the complexity of each component goes up, so does the possibility of errors and component failures that occur below the kit’s “user interface” level.

 

IMG_5102

I found a stack of these electronics kits outside my classroom one morning. I assume they were unearthed from a science closet and delivered to the Makers Space instead of the recycling bin. All the Makers students who played with them remarked that they had kits “just like this, but different” in a closet at home. They’d flip to a page in the back half of the pictorial instruction book, and start to build. Ten minutes later, the comments shifted to talking about how the kits they had “were just like this, but better.” Ten minutes after that, it was “kits just like this, but those worked!”

“It doesn’t work” is a trigger phrase for me in the Makers Space, right up there with “Am I done?” So we dug in, and built a single loop from battery to switch to light bulb. Nothing.

What do you do when the rituals don’t work? In the Makers Space we could pull out a multimeter and confirm that the batteries were good. Then we could identify the one battery holder that didn’t actually connect the two AAs. We could trace through the stacks of plastic and find the parts that were touching but not connected. In short, we needed electronics knowledge from outside the kit’s domain to troubleshoot both students’ circuits and the kit’s components.

Here’s what we found. This incandescent bulb block was designed so that the wires connected to the small metal Edison thread housing, so that end users could easily replace the individual bulbs. With the bulb in place, everything about the light block looks fine. On closer inspection, you can see the loose wires and the bare solder spots on the socket.

Snap_lights

 

It’s easy to see this as just a broken component, but i think that’s missing the larger problem. I draw on a mess of experience to assemble a few bits of data into the story of a broken component. But for novices, simple isn’t obvious.

After this, kids filtered out all circuits that involved the incandescent bulbs.  Then they hit this “door bell” design.

two_snap_circuits

Yes, despite my best efforts to double check student work, the top circuit (which looks correct!) does not produce sound. In frustration, a student started to slap extra wires across his setup, and got sound by bridging two other pins on the music IC. In the bottom picture, this bridge is yellow pushbotton block. Now instead of prompting careful observation, the grimore was rewarding Sorcerer’s Apprentice behavior.

Frustrated, another student popped the plastic casing to investigate how those pins were connected. This is what she found.

IMG_5106

Vanishing electronics. Hidden in the middle of an electronics kit is a tiny microcontroller, sealed under a glob of epoxy.

Better engineering can mitigate these problems., but the trade-off is fundamental to the component packaging approach adopted by electronics kits. It’s not easy to dislodge a SMD resistor from a LittleBits block, but when it happens it will create mountains of frustration with no obvious cause.

Every packed kit presents a simulacra of “real” electronics, elegant and convincing so long as you stand in the right place. Poking at the seams in several kit systems, I have a much deeper appreciation for the Minecraft’s redstone microworld. At some level, the kits will always suffer because they have to build a simplified model of electronics out of actual electricity.

Electronics Kits and Bits

Kits make electronics accessible by simplifying along two distinct axis.

The first is component selection. Kits offer a cultivated garden of parts that all work properly with each other, avoiding the bewildering Digikey page or the haphazard component drawers at RadioShack. Different systems stake out different positions along this axis. Squishy Circuits (to the extent that it’s a kit at all) simply offers generic components with fat dough-friendly leads. Hummingbird Robotics boards have easy, labeled terminal blocks and come bundled with commodity parts. At the low end of the spectrum, kits include parts to indicate a stable starting place, but they don’t build fences. Systems that aim for openness create a substrate that still works with plain/“bare wire” components. This covers everything from a simple breadboard to conductive dough and the unpattentable springs of 101 Experiment kids. Forest Mims’ excellent Electronic Learning Lab offers all of these, along with integrated LEDs and potentiometers. The Electronics Learning Lab was a longtime Radioshack stalwart, but is currently, lamentably, out of print.

What restricts learners from selecting or making use of other parts is the way a kit/system decides to simplify connectivity. I suspect that this is also the major area where designers can apply trademark and patent restrictions to their systems in an attempt to prohibit “knock-off” competitors.

Closed systems encapsulate individual components, which improves users’ first circuit experience and ensures that all parts are reusable. However, anyone looking ot replace a missing ingredient or add new ones are locked in to the kit vendor’s store.

Circuit Stickers from Chibitronics occupy an interesting position in this quadrant.
Individual components come on sturdy but flexible pcbs, backed by z-tape reusable adhesive. The basic kit has a wonderfully designed workbook and includes a roll of copper tape on paper, but nothing about the component design restricts how they can be used. Each tiny sticker even has wide solder pads! Chibitronics even includes a spare sheet of conductive backing for components that lose stickiness over time. Jie and Bunny’s design is a marvel of simplicity and economy. Currently, the Chibitronics website offers 24 packs of LED sticks for fifteen bucks. That’s a price lower than the horrific RadioShack markup on standard through hole 5mm LEDs, which no one should ever ever pay, but we often do. The same pricing extends across the Chibitronics line, from the programable Attiny stickers to sensors. Chibitronics offer custom components that are simple to connect without locking the user in an exorbitantly expensive walled garden.

Which brings me to LittleBits.

I’ve been deeply intrigued by and frustrated with LittleBits for about two years. At this point, I think the Reggie Watts commercial for the SynthKit does the best job of capturing the excitement “snap together amazing things!”


Stoked? Great, now look at this.

Screen Shot 2014-05-21 at 10.53.52 PM

I’ve heard LittleBits described as a “learning electronics” kit, and never felt it was accurate. Their proprietary magnet-snap connections pushed both simplification factors so far that they seemed to rocket out of the electronics kit quadrant. Components were so simplified that few productive mistakes remained for the user. There’s nothing to observe in a LittleBits creation about how circuits are designed. Then there’s the price. A “class set” of LittleBits rings in at several thousand dollars. I’d seen students enjoy the momentary experience of playing with LittleBits, but since the high price-tag meant that all of the parts went back in the bin at the end of every class, I never saw work pass beyond the “kinda neat” stage.

With the release of the Arduino Bit, I feel like LittleBits has fully abandoned the “learn electronics!” category,or maybe they’ve just hit escape velocity. Instead of a over-engineered electronics kit, they’re offering a portal into psychical computing in a radically divergent form factor. LittleBits offers the same kit benefits (connectivity, part selection) and provides and access ramp to the Arduino “prototyping platform.” LittleBits allows users to focus on the hard fun of designing physical computing systems to make use sensor data, wether through simple logic gates or a full Arduino. Like all other micro-controllers, there’s electricity at work, but the user is hidden away from almost all of the electrical engineering.

Under the Hood

Laura is a smart cookie.  The voltage frustration that emerged from the intersection of batteries and LED series is entirely a matter of exposure.

The problems she details in this thread exist at a level of electrical detail that she’s never needed across years of learning to code, teaching CS, diving into and teaching physical computing, coaching a whole variety of robotics teams, and renovating a house! Big tools or appliances might draw to much current and blow a fuse. Robotics are designed for and often ship with specific batteries and polarized plugs. All designed, commercial systems are careful abstractions that do their best to constrain the users exposure to the confounding detail and complexity. Smart engineers have gone to incredible efforts to make the user experience feel closer to redstone than 6.002.

A major contributor to this layer of abstraction is the cheap, accessible ubiquity of micro-controllers and other discrete logic systems. This creates unified circuits that use electricity as bounded signals. In these circuits, failure occurs when signals are blocked or interrupted, but rarely every as a result of signals that fall outside a component’s expected/acceptable range.

The heart of Laura’s LED problem was a circuit that could be closed and complete, but still not function. It’s not that this is complicated, but that behavior is orthogonal to the “blink sketch” mindset.

So that’s it, right? Here’s the great example of why #makered novices need some “real” electronics knowledge.

I’m still skeptical. Clearly, there’s a good lesson lurking in here that could be packaged up neatly into a bag with LEDs of various Forward Voltages and a small pile of batteries. Blow a single LED with a 9v (actually a step in the great Make: Electronics book), have a 3v coin cell power two small red LEDs but fail with two blue ones, compare series vs parallel circuits with copper tape. I think that’s probably enough hands on experience to convince a novice to check Adafruit’s Circuit Playground app when a LED circuit doesn’t behave as expected.

My skepticism about the value of “real” electronics knowledge for novices in 2014 comes from how complex the picture becomes when we look under the hood of digital logic assumptions.

Vanishing Electronics

In this course, students discover the basics of electronics design and assembly. They use this knowledge to build their own simple flashing LED using solder-free breadboards. By diving into the assembly of these projects, students learn about impedance, resistance, conductivity, and circuit design through personal, hands-on engagement, opening up incredible possibilities for creative projects.

That was the first description I wrote for what became our Makers program. This was a perfunctory bit of text written a year before our first class, tossed onto a Google form and into oblivion. The audience was exclusively parents, crassly intended to trigger connotations of learning and complexity for an unproved course.

In that first year we did some of what that blurb promised, but it wasn’t the focus of the course at any point beyond the second week. We built Squishy circuits and made 555-based projects, but I guarantee I never used the word impedance in class.

That same year I took the first MOOC version of MIT’s 6.002. Well before we hit the midterm, I was struggling to keep up with increasingly complex circuits, dusting off my integral calc skills, and churning through paper at a fantastic rate. Coming home from teaching Makers to a new lecture or problem set triggered new waves of teacher-panic. Had I really promised to teach this material to 7th graders?

As part of the FabLearn cohort, I’m exploring broad plain of electronics with an eye on how it fits into the modern/developing #makered landscape. It’s got to be central, right? Even though we advocate for the value of cardboard prototyping and physical construction, the big name tools are all complex electronic systems.

Well, maybe not.

From the 70s into the 90s, there was a rich field of interesting projects and creative experimentation that was only open to people with a functional literacy in electronics and electrical engineering. My sense is that in the last decade, the growth of cheap, flexible, accessible micro-controllers has taken control of the sophisticated projects that used to drive students deeper into electronics. While that trend leaves the electronics domain with fewer “exclusive” projects, the hurdles facing a novice haven’t changed much. While the internet makes it easier to share schematics and video tutorials, electronics still requires parts and precision. I can download the schematics for a transistor radio or a preconfigured disk image that transforms the RaspberryPi into an FM Transmitter.

Deep electronics knowledge opens up incredible possibilities. I’ll submit as evidence any of ch00ftech’s posts or Jeri Ellsworth’s Short Circuits. But in an Arduino-rich world, there’s far fewer low-end projects that require those skills. Instead, those electronics skills become mandatory when a project needs to exceed the constraints of what’s possible with a micro-controller. When a project needs to use less power, take up less space, respond with less lag, scale out at with less cost, then you’ll need the skills to design and build complex, task-optimized circuits.

Here’s the chicken-egg of learning electronics in 2013. Every basic circuit in a Forrest Mims notebookcan be duplicated with an Arduino and a tiny selection of components, using copy-pasta code and breadboard illustrations. In that world, how much discrete electronics knowledge does a novice need?

In the months I’ve spent looking at this problem, I’ve slammed repeatedly against a cognitive wall. I examine my practice, the projects my students pursue, the projects shared throughout the wider #makered community, and I see a role for electronics that’s smaller, more constrained and highly task specific. When I interrogate those findings, I keep coming back to a enduring conflict. Are my observations accurate, or does my weak understanding of electronics obscure a larger and more complicated story?

I can’t be sure. I’ll be sharing my look into the current market of “learn electronics!” kits in a later post, along with exploring the uneasy border between circuit simulators and Minecraft. But throughout those, know that this question – am I missing the real story? – lies under every observation.

Makers? How can making be a class?

I’m still not sure if it’s a great class, but it’s certainly a standout hour of my day. Here’s today’s class compressed into about 2 minutes. Yes, I cheated and used audio to elicit an emotional response. I also captured spontaneous student emotion and cuteness.

Projects being worked on in the video include:
Sparkfun TriColor LED Breakout
First Arduino sketches
Testing power supplies
MintyPOV3
mini fume extractor using 120mm case fan
3 digit 7-segment display counter from the Make: Electronics text
An APC-esque noise maker using two 555s and a pair of transistors
parts shopping for ELWrire glasses, binary clocks, and others

As an unexpected benefit, I learned far more than I expected about my class by watching the timelapse three times. J-, who is seemingly sedentary through most of the class, is actually deeply engrossed in inventory and setting up the soldering for his MintyPOV. A-, who moves around far more, is clearly hopping from one conversation to the next, wandering off when the person he’s talking to gets involved with some tools and stops answering back. That boy needs a new project.

Post Navigation