Some things are inevitable…

I taught undergraduate and graduate hardware design for many years. It was always a challenge to give students hands-on experience with digital logic circuits, especially for large-scale digital systems. Gosh, I could write several long blog posts that recount the various breadboarding, prototyping and simulation approaches that my students used in the lab. Along with the achievements came many horrors.

So, picking up littleBits was inevitable.

littleBits are so cute. Just in case you’ve been in an isolated cave, littleBits modules are small building blocks with magnetic “bitSnaps” at opposing ends. The magnetic bitSnaps sort the ports into inputs and outputs, and enforce an output-to-input connection rule. Each bitSnap carries the positive voltage rail, the ground rail and a signal. (Click images to get full resolution.)

The littleBits range covers basic digital logic and analog building blocks including audio. The Synth Kit — designed in collaboration with Korg — has perhaps done the most to promote widespread discovery and use of littleBits. The Synth Kit certainly placed littleBits in relatively non-traditional hands and sources, namely, musicians and musical instrument retailers. Before their recent bankruptcy proceedings, one could find individual ‘Bits in Radio Shack stores.

According to the manager of a local Radio Shack, the littleBits suffered a very high theft rate. Shame on anyone who unlawfully filched these!

Most retailers today (like Barnes and Noble bookstores) sell littleBits kits. The kits make it easier for people to get started and it reduces the number of separate SKUs at point-of-sale. With Radio Shack’s reorganization, these days, it is more difficult to buy individual modules a la carte. The littleBits on-line shop carries the full range, of course. Still, one rarely finds ‘Bits or kits at a discount.

The kits are also a little less easy to steal although a Barnes and Noble shopkeeper acknowledged high theft potential, too. Stop stealing!

One reason why people may feel entitled to boost littleBits with sticky fingers is their perceived expense. An LED at $7.95 may seem grossly overpriced. However, each individual module has at least two bitSnaps and there is value-added design and electronics on board. The LED module, for example, has a voltage protected input and a driver. In fact, all signals in the littleBits series are driven and normalized; that’s what makes them special. If all you want is a naked LED with two leads, then by all means pass up littleBits and try something else!

Signal standardization is what makes littleBits what they are. littleBits are very close to the plug-and-play electronic LEGO™ blocks that educators have wanted for a zillion years.

I decided to forego prepackaged kits and buy a la carte. The kits usually have modules which I probably won’t use and it just seems like a waste to buy them. That includes the Synth Kit, too. (Are you surprised?) I don’t have any interest in the tiny keyboard or the sequencer. I’m not that interested in the oscillators for that matter. I want to drive my synth and other experiments with Arduino. Thus, the only kit that I have purchased is the Arduino Coding Kit. Everything in this kit is quite useable with the exception of the servo module. (Although a Neil Young Whizzer would be a unique project!)

The littleBits Arduino board is a good example of value-added engineering. Sure, the Arduino ‘Bit doesn’t provide all of the ins and outs of a stock Arduino UNO, for example. However, the bitSnap signals are standardized with protection on the inputs and (switchable) low pass filters on the PWM outputs. The serial transmit (TX) and receive (RX) signals are made available through bitSnaps, making MIDI IN and OUT a future possibility. Programming is supported through either USB or ICSP. You’ll need to add header pins for ICSP, but if you’re using ICSP, you already ride tall in the saddle.

I intend to whip up MIDI IN and MIDI OUT circuits using PROTO modules. I thought long and hard about the options. The Hardware Development Kit contains two PROTO modules, one PERF module and 12 bitSnaps (6 male, 6 female). I am not a fan of pad-per-hole perf board, so the PERF module is not that attractive to me. I much prefer “DIP style” prototyping boards that accomodate full size, through hole DIP ICs. This is sooooo much easier to solder. For similar reasons, I’m not sure that I want to fiddle around crafting circuit boards to conform to the bitSnaps. Overall, I will probably buy two PROTO modules a la carte.

The littleBits web site is a little sketchy on certain vital details:

• The Arduino board USB port is a micro USB port.
• The power module can be driven by a 9V, center positive, 2.1mm power adapter — the same as an Arduino UNO.
• The power module can drive any input bitSnap (any port in a storm).
• The Arduino IDE should be configured for the Leonardo board.

I don’t like hunting around for this kind of information and I especially do not want to watch videos to find it! In fact this information should be printed on the plastic purple baggy containing the Arduino board. (Also, the link to Arduino sketches on the current baggy is dead.)

Thank goodness that the Arduino Development Kit comes with two mounting boards. A mounting board holds a littleBits circuit in place and makes connections reliable. I whipped together a simple circuit without using a mounting board and the relatively heavy adapter cable kept pulling the power module away from the other boards. The bitSnap magnets are simply not that big and strong. Programming and using an Arduino module without a mounting board would be a nightmare.

So far, a favorable, somewhat vanilla review. Here’s a bit of my own value-added as a computer science educator.

When I first powered up the pulse module — a 555 timer-based circuit that generates a regular stream of digital pulses — I didn’t necessarily see a clock. I saw a token generator. I thought, “Given the current design of the logic elements, wouldn’t it be neat to build and visualize a one-hot state machine (controller)?” If each latch module had a built-in LED showing the state of the latch, this application would be a natural for littleBits.

Even better, why not bag synchronous control altogether and go asynchronous, self-timed?

Ivan E. Sutherland, Turing Award paper, Communications of the ACM, Volume 32, Number 6, June 1989, pages 720-738. And don’t forget everything written by Chuck Seitz, too!

This immediately raises the question of what we want students to learn from their experience with littleBits. I want students to learn that:

Electronic systems are inherently parallel.

Students need to be exposed to concurrency and parallelism before their minds are corrupted by the sequential, synchronous paradigm in conventional programming languages.

The other question that pops up is scalability. littleBits logic elements currently operate at the bit level. However, I’d love to have modules at the bus level such that students could build their own CPUs, GPUs, whatever. Plug-and-play busses are a tougher design challenge electrically and mechanically. One possible solution is to transmit data words serially from module to module. Pass the word size along with the data in order to simulate any (reasonable) bus width. littleBits could then keep the current bitSnap form factor.

Finally, I would like to mention the module technology that gave me a start in digital electronics: DEC Register Transfer Modules, also called the “PDP-16” or “RTMs.” RTMs use a flowchart-like notation for control design. This approach opened a whole world to me including ISP, Petri Nets and eventually dataflow computing. Most importantly, RTMs taught my 19 year-old mind that computation is fundamentally parallel and we should design for and exploit concurrency right from the get-go. Nico Habermann, God bless him, did the same in the sophomore programming course at C-MU. This point of view influenced my entire professional life!

If you would like to know more about RTMs, then read “Designing Computers and Digital Systems Using PDP-16 Register Transfer Modules,” by C. Gordon Bell, John Grason and Allen Newell, 1972. I was one of the first guinea pigs to use RTMs in the computer design course taught by John Grason. Thanks, John!

See also “The Description and Use of Register-Transfer Modules,” C. Gordon Bell, J.L. Eggbert, J. Grason, and P. Williams, IEEE Transactions on Computers, Volume C-21, Number 5, May 1972, pages 495-500.