So in anticipation of the new board coming in a few weeks, I used the last prototype to finish off a final test station. My last one was pogo-pin based, and as “cool” as that one was, it just wasn’t efficient enough. You’re probably thinking “well all you have to do is press it down on the pins and turn it on, that’s about as simple as it gets.” Well, that’s true, and not true.

The last fixture had flaws, one of which is that aligning the pins for every board I’m testing can be a huge pain in the ass, especially when the holes are hand drilled. They’re close to being dead on, but not close enough for me to be happy about having to look under the board to double check if I’m on correctly or not. The second flaw is the testing itself, which it really only allowed me to power it up, and of course update the firmware via USB. The issue I have with this is that:

1. It didn’t allow me actually program the bootloader into the on board MCU very well since that meant an extra step prior wherein I had to power the board while using the other hand to click in windows, which doesn’t work well when I’m trying to push the board down on to pins.

2. It didn’t allow me to test any of the signals, only whether there was a blatant malfunction.

To fix this, I did a quick re-design and approached it a different way. My new board supports a 20 position breakout header for signals, while retaining a terminal for the main power and USB signals. Using these properties I wanted to make this one have a few options.

  • Modular like the last one (USB plug)
  • 20 pin signal access
  • The ability to not only control the power and main function lines, but also actually test the individual signals as well as a final run-through

So I ran to radio shack and picked up a few things, one of which included 30 gauge kynar wire. I’ve used it at work before, and it’s helpful in some ways but really irritating in others. It saves a boatload of space if you’re planning to route a lot of wires in a very small area, but stripping it requires a bit of finesse. I’ve heard of people just prepping the pads and allowing the backshrink of the insulation to kick in while soldering, making the stripping process not needed. I didn’t have much luck in that, so I just stripped every end.

Eventually I laid out how I wanted things and soldered it all together, I ended up with this:

Looks pretty snazzy on top, but that’s because I wanted to keep all the wires underneath as much as possible. Check out the ugly truth to this little guy.

Power and Ground routed with 26 gauge, the rest is 30 gauge. The 3 toggle switches are for the main power control and power up functionality of the new board. The breakout is for the signals, and the little button on the right is connected to the Test MCU that runs a program which cycles through all of the connected signals, allowing me to pull up the controllers applet in Windows and get immediate feedback on all lines. The 2 small outgoing wires head to the terminal of the new board for pass through of the USB data lines.

One huge issue I had with this is that the large mess under the board left this thing kind of fragile, so I added one final requirement…

  • Make it rugged enough to endure the hectic nature of testing a crapload of boards in a single sitting.

So another trip to radio shack, some nylon stand offs I had laying around, and about an hour of dremel work landed me this guy. When I step back from it I’m actually pretty proud of it. For a test fixture it looks pretty snazzy and pretty god damn rugged too. I’m not going to accidentally tear any wires, hell I could throw it against a wall and I think it would survive.

When the new boards come in I’ll be ready for them now, and should be able to run through them (testing wise) pretty damn quickly now.