We’ve Gone Nuclear

The return of the pink badge! We created another super simple, entry level badge for first timers at DEF CON. There’s not even a microcontroller! Just glue logic tied together to make a cool LED pattern

Designing with Glue Logic

You will be hard pressed to find a badge in this day and age that isn’t based around some kind of microcontroller. That is the easiest way to do it, and allows for a bunch of flexibility in your design. It’s also gotten extremely cheap to do, especially if you design everything around some nonstandard chinese micro.

I find it much more interesting to design blinky badges with glue logic. There are tons of special purpose chips out there made for interfacing more complex systems together, and all you have to do is string them together correctly to get some interesting emergent behavior.

The DCPunks Nuclear badge has just a couple chips: two shift registers and a 556 timer. I used these chips to design an LED pattern that swirled around the nuclear symbol. The 556 timer is split in half: one half generates the clock to feed bits into the shift registers, and the other one determines if the bit we are feeding is is an on LED or an off LED. This creates a snake trail of LEDs that move around the badge in a deterministic pattern. And you don’t even have to program anything! It’s a single purpose machine all designed in hardware.

Manufacturing Pains

It is absolutely imperative that these badges be pink. Pink is cool and unique and nobody does it. The place That I order pink PCBs from (PCBWay) does not offer an easy method for assembly. That means we gotta do it all by hand again!

We were a little more organized this year. We set up our assembly line in the living room, had someone squeegee on solder paste, someone else place components, and finally someone on oven duty to make sure they all baked correctly in the reflow oven. We were able to blow through around 30 or so badges in a couple of hours.

Cleanup was a pain though. The soldering squeegee was not perfect and many of our parts came out with bridges that needed some cleanup under the rework microscope. A fair amount of components were also not placed correctly migrated down a pad, which required a hot air rework.

These SMD resistor arrays were a HUGE mistake. My original though was “hey, instead of placing a bunch of resitors, I can just place one array and be done with it!”. It didn’t work like that. Every single one of these stupid things had a bunch of bridges that required microscope rework. It took way longer than if we had just placed a bunch of 0805 resistors by hand. Assembly would have taken longer, but there would have been pretty much no cleanup needed

Also, WHY did I design this with 0603 parts? There was plenty of room! Another needless headache that I could have avoided with a but of forethought. Oh well.