In this case, here are the clues she used to reverse-engineer the wireless protocol:

1. The specs for the radio chip are easily available, so she could identify the actual packets from looking at the voltages on four wires.

2. Each packet had two identical bytes that went up when she turned the remote up, and down when she turned it down.

3. There were some bytes which were always constant, and didn't seem to do anything.

From there, the code ended up being fairly simple:

https://github.com/scanlime/arduino-lelo-remote/blob/master/...

On power-up, it sends an initialization sequence, copied from the trace from the original remote. This has a bunch of settings that you can look up in the manual. Then for each packet, it sends a start sequence (essentially, "finish sending anything queued, and stand by"), puts a packet into the transmit queue (packet = 0x0100a5[power][power]00000005, where power is the power level byte), and then tells the radio to transmit.

And then of course you can debug this by comparing the traces it generates to the traces from the original remote.

> 1. The specs for the radio chip are easily available, so she could identify the actual packets from looking at the voltages on four wires.

How does one go about doing this? Is this done on hardware or some wireshark-esque software tool?

My favorite tool for this is the Saleae Logic[1], but I've had mixed success with the cheaper Bus Pirate[2] as well.

[1] http://www.saleae.com/logic [2] http://dangerousprototypes.com/docs/Bus_Pirate

I can't wait for easily and cheaply 3d-printed electronics. Just the thought of buying the parts and plastic stuff, putting them in a machine, and letting it work, building a complete device.

I can dream.

Well, the more I looked into this possibility, the more I think it's not going to look like traditional electronics that we know. Chances are, somebody's working on this right now, and we just don't know about it.

There's been some promising blips here and there with the reprappers:

http://blog.reprap.org/2011/06/new-approach-to-printing-meta...

http://blog.reprap.org/2012/04/some-more-printed-circuitry.h...

Some recent discussion on it and more up-to-date info.

http://forums.reprap.org/read.php?2,154755,154835,quote=1

I wonder if there is some order-to-3dprint business out there?

There are, but none of them print circuits.

http://www.shapeways.com

http://www.ponoko.com

http://imaterialize.com/

But there are plenty of 1-off or small run PCB manufacturers, which will surely be higher quality than anything produced by some 3d printing experiments.

It's true for now, and for the near term (read: couple of years). The trend for the 3D printers is that they'll both get faster and use different materials (or same material with different properties). That way, you can print everything in-place, from batteries, the board, and enclosure. You won't be able to replicate chips for a long while, besides really simple electronics, so it will probably be pick and place.

How about a web-based design tool that lets you make a custom case, which will then be 3d-printed on demand? Alternatively, you could accept customer uploads of certain 3D design formats.

There's gotta be a market for this.

Well, I meant more of an all-in-one product: the company would provide a few different options of base vibrator units with minimum case size/dimension requirements, and the customer creates the custom case; the product would be shipped fully assembled. I suppose they could outsource the printing to Ponoko, though.