There's a side effect to the multiplexing: in the sample code, if down is pressed between the if (!digitalRead(p53)) return 0x01; // Down and if (!digitalRead(p53)) return 0x08; // Right sections, the code thinks right was pressed instead. It happened to me quite often (I even got it to think I was pressing left a couple of times).

While I know this part was retired a long time ago, I have two of them and for as long as they keep kicking I'll use them in my various wearable art projects. I for one would love to see you resurrect this part. I would buy ten of them as a hedge against future retirement because they are exactly what I need for my projects. (http://anthrolume.com)

Have a look there: https://forum.sparkfun.com/viewtopic.php?f=14&t=29321

With detection of short, long, double click.

While this product did have its drawbacks, I have two of them and they worked great for what I was using them for (http://anthrolume.com). I'm quite sorry to see the product go away - the form factor was really perfect, and the thing was weather and dust-proof. Please replace this product with something that has the same form factor and seven wires without the weird multiplexing.

I have to say that I'm not pleased with this product. While I understand the desire to save I/O pins, increasing the number of leads from 5 to 7 would eliminate the need for the diodes and confusing wiring scheme (and maybe reduce cost!).

The polling currently required is annoying. Not only does it mean running the poll code when the processor could be doing something else, it is unreliable, as stated elsewhere. The need to do multiple polls to ensure an accurate reading means more programmer headache. It isn't worth 2 pins!!

Buttons are made for interrupt-driven code. AVR microprocessors' pin-change interrupts (and Arduino libraries supporting them) are easy to use and make sense for many applications. While you can set up interrups for the up/down/logo buttons, left/right will still require manual polling.

On mine, the wire order was inverted !!! The LED was at wires 4 and 5 (not 2 and 3 as per PDF).

It was easy to test, put a 3.3V between wires 2(+) and 3(G) or 4(G) and 5(+). The LED will go on and you will know the wire order.

I got the LED to work but I can't get any of the buttons to function. Where is page 2 of the datasheet?

The schematic on this little gizmo is kinda screwy in my opinion. It works, but is somewhat messy in the software logic. I get that the designer was going for economy of port pins (only 3 is good), but with most *duino (or other MCU) boards, port pins are NOT in short supply.

I much prefer a clean row/column schematic, which helps the software weenies get their logic orderly. So in this product, a row/column scheme would have yielded 5 port pins (2 rows x 3 columns) and more ordered code.

< / 2 cents>

It would be really nice to have either a mechanical drawing of this gadget, or views of the side and back with the little gray ruler.
It appears I could mount it behind a control panel, exposing only the inner square bevel. It would be great to be sure.

Tell the truth, did the decision to use five active wires have anything to do with the size of the Lilypad small prototype boards? Or is it just a fortunate coincidence? I had to add solid-core wire ends, but otherwise I think this works really well, thanks! (I used it to 'draw' on my LED-matrix top: chebe.dreamwidth.org/61563.html )

hmm... looks like a custom sparkfun logo in the middle :) could we ask who supplies these to you?

I wrote an object and demo to use this keypad with a Propeller chip.
http://forums.parallax.com/showthread.php?134434-Wearable-Keypad-Object
I used 10K Ohm pull-up resistors on the button lines.
As other's have noted, the wire order is reversed in the datasheet.
I measured 34mA current when the LEDs are powerd with 3.3V.

The wire pitch is less than 0.1", but I found that it was pretty easy to crimp an 8-pin IDC connector onto the ribbon cable if you split the conductors and align them carefully.
I too would love it if the LEDs (there are actually four inside the thing) were RGB.
I'm also puzzled by the multiplexing mechanism - all that to save one wire and two diodes? You could have used seven wires(GND, LED, Left, Right, Up, Down, Flame) and no multiplexing. Plus that way you'd be able to detect all the combinations of buttons. With the current schematic, it looks like only combinations of up, down, and flame are detectable.
Nevertheless, it's a cool gizmo and I'm using it for controlling my anthrolume project (http://anthrolume.wordpress.com).

the datasheet is backwards to me, anyone else?
Yes the datasheet wire ordering is backwards for me as well.

at first, my dpad just output "left" repeatedly with the example code using arudino022 w/ duemilanove. i reversed the order of the connections and now it works. the datasheet is backwards to me, anyone else?

Not that I haven't got it yet or anything, but it would be really helpful if you colored the last bit of wire or something somehow
I could technically do it, but it might be neat...

What type of buttons are being used under the pad? I understand if you can't tell me, but I need something like this, but in a different form factor, so unfortunately I can't use this keypad and need to construct my own solution.

Needs to have a blue LED.

Sounds usefull, but the 2nd image makes the connection to the pad look nasty. Is it poor quality, or just a bad image?

How long is the cable?

What are the dimensions ? The data sheet says page 1 of 2, but only 1 page in the PDF :(

Is it water resistant?
How strong is the area where the wire meets the housing?