Saturday, 2 July 2016

Reviving a Xyloband(TM)

Saw Coldplay recently, and have also been messing around with LED Matrices via the R-Pi, so thought I'd have a go at reviving the wristband.

These things are powered by 2xAAAA batteries, and have an led strip in the band consisting of 8 RGB LEDs. They light up via controls sent over radio, but since I don't have a transmitter or would know how to reverse engineer the protocol, I thought I'd see if I could just make one light up without any of that using a little switch arrangement.

Stripping it down was straightforward. One end of the LED strip has 4 contacts, which seem to correspond (left to right) as:

Pin 1: +ve contact
Pin 2: Blue -ve contact
Pin 3: Red -ve contact
Pin 4: Green -ve contact

(I may have +ve and -ve the wrong way around in that - the board itself uses a red wire for -ve and a black wire for +ve, so it's a bit confusing).

The other end of the Xyloband produced nothing. So obviously that's the "far" end of the strip.

Anyway, I was able to get it to light up, by touching a wire from the +ve and -ve end of the battery connectors to the various pins. Since this didn't actually blow the LEDs, I figured i could just wire it up with a little switch and I'd be done, and not have to worry about resistors. (slightly wrong on that, see later...)

I found a nice little 4-throw dip-switch in Maplin (code JH08) so I could control the red/green/blue independently.

Removing the LED strip from the existing radio board was easy using a soldering iron, just hold the iron to each of the 4 contacts in turn and the little ribbon connector pulls away. (I've chucked the radio board and its silly little speaker away now, to get more room in the casing).

Then I had some breadboard wires lying around, so soldered those up to the various pins of the little dip switch, one switch for red, one for blue, one for green... and cut a little hole in the side of the casing to wedge the switch into (be careful which side, or you'll touch part of the battery assembly)... superglued it in place, and hey presto!

Only problem is that you do actually need resistors because the red LED clearly has a different rating to the blue and green ones. So if you have two LEDs on in parallel, the red one will always win, but blue and green combine nicely. This prevents you getting purple, white and yellow colours out of the LEDs, so I need to get my multimeter out, and then look at relative voltage drops etc. (Not today though)

I'll update here when I know the right resistor values.

In the meantime, see video for the (low-tech) working band!!!

I may also see what I can do to make this a bit more dynamic... but there really isn't much room in the case, not even enough room for a PiZero... although that would be cool - but probably needs a battery upgrade as well.

I've got my eye on a BLE-Nano from RedBear, however that is going to be quite a bit of firmware development in C to get to work, but should, in theory allow you to have something you could drive over Bluetooth from a smartphone, which then gives rise a set of rather more interesting-looking applications for it rather than just something you can make light up, pretty though that is...

<to be continued>

No comments:

Post a Comment