Bi-color LED Matrix Clock Uses Point-to-point Construction

[Daniel] wanted his child to stay in bed until a semi-decent time each morning. The problem is the kid doesn’t know how to read a clock, so [Daniel] built him a clock. Yeah, doesn’t make much sense to us either, but we’ve used our own shaky premises for projects so who are we do judge?

He used a bi-color 8×8 LED matrix as the clock display. What caught our eye is the point-to-point soldering he used for the three strip boards that make up the device. Note the use of a drill-bit to break the traces when needed. Each board has its own purpose; the matrix drive, the logic board, and the power board. A PIC 18F4550 lets [Daniel] control the clock via USB, and takes care of lighting up the hour as a red number when it’s time to sleep, and a green one when it’s okay to arise. There’s a flashing pixel for seconds, and a binary readout of minutes along the bottom.

We’ve asked [Daniel] to post a schematic and an image of the clock face when displaying the time. No word yet but we’ll keep our eye on it. In the mean-time, check out this clock that uses an RGB 8×8 LED matrix.

15 thoughts on “Bi-color LED Matrix Clock Uses Point-to-point Construction

  1. The way it shows the time is quite cool and the project is well made.

    But no harm can come from teaching your kid how to read the clock, right?
    I mean, it could be as simple as: ‘don’t get out of bed until the little hand gets here’.

  2. If you want to see some impressive point-to-point wiring, Google “wiring pen”. You can buy a pen, or easily build one from a mechanical pencil like I did. Load it with wire-wrap or magnet wire.

    It’s fast. No stripping wire because the heat of the soldering iron burns through the insulation. No cutting wire on through-hole components, because once you make the final connection for a wire, you can just pull the wire taut and work it back and forth a couple of times; the excess snaps right off and you’re ready to make the next connection.

    Usually the board is laid out so there are horizontal/vertical “channels” between ICs and other major components, and wires are routed in those channels. Some people use mostly the 0.2″ space between rows on DIP ICs for the channels, leading to very small and neat layouts. I usually leave 0.2″ space between ICs as well for an additional channel. “Wiring combs” can be used which form physical channels that hold the wire in them, but they’re a bit pricey and I’ve never found them necessary. Instead, I just hold the wire down to the board and make a 90° bend, the wire will stay that way unless you tug on it too hard.

    Some people even do SMD this way, here’s an impressive example:

  3. Green incandescent bulb in a lamp, and plug in timer, any one can duplicate the anyone can duplicate the; you get out of bed now concept easily. I can hear it now Dad the light didn’t come on, so I thought it was burned out.

    @jc All Mike said was that the wiring caught their eye, that falls several degrees of amazement IMO. Few project employ very much point to point on a major amount anymore, so when does come across one, it will catch your eye

  4. @Chris I love that SMD p2p link, that’s impressive and insane really.

    As for wire-wrap, I’ve only come across that a few times in radios from the 40’s and before, which shows it works pretty well since often those radios are still functional.
    But with modern devices it’s not just the size but also the high frequencies everything uses that makes that a bit of a hard thing. to use.

    There are incidentally pages at NASA with instructions on it, shows how effective it can be.

    However I read this: “Although the NASA STANDARD for Wire Wrap has been cancelled. JPL does offer a Wire Wrap class for those in industry who are still using this process”

    Example link:

    And classical soldering:

    And here’s a selection page:

    Hackaday might need to do an article on those NASA standards incidentally, it’s very tutorial to learn what can go wrong.


    That’s amazing work, but imagine troubleshooting it… What if the insulation of wires is damaged as they make the sharp bend into the protoboard’s holes? It would be a nightmare to figure out

    Since I’ve started to etch my own boards I haven’t looked back to protoboards and P2P wiring. When you run a decent design rules check on your PCBs when you draw them in e.g. Eagle, the risk of problems is minimized very effectively.

  6. When I submitted my Boston Bomb to Silicon Chip (a random electronics paper publication) here in Australia, they declined to publish it, because it used Veroboard, and “our readers do not like Verobaord).

    But – and this is probably more important – has anybody seen that he wrote the code in Forth. Thats pretty neat (duck) :-)

  7. I wonder how they determined their readers don’t like veroboard doug, sounds like they might assume a lot.
    Lol on the forth btw, amusing catch, forth is so rare that if I didn’t find a wikipedia entry on it I would not be 100% surprised, just 98% :)

  8. The only reason I spotted it, was that I used to program in it in the late 80’s.

    LMR Forth for the 8080, with a cross compiler. You knew exactly what the hardware was doing, because you made it do that…. None of these fancy abstraction layers…. Those were the days ;-)

  9. @John Doe: Normally you stay on one side of the protoboard with wiring pen technique, going through holes is an extra step that slows you down. But you’re right about damaged insulation, although it happens fairly rarely; I can only remember two times it’s happened to me. Once because I was in a hurry and being incredibly sloppy, and another because I had to rewire the board several times.

    @Doug/Whatnot: He didn’t just code the program in Forth. He also wrote the Forth compiler and runtime environment, and Forth is a natural choice for that because it’s so easy to write.

    Some years ago, I created something like the currently popular Bus Pirate. It was an ATTiny with a simple tokenized Forth engine, with commands added for digital I/O. A program on the host computer would accept short Forth programs, tokenize them, and send them to the MCU over the serial port for execution. The program could be as simple as reading/writing a pin. Or as complex as bit-banging a whole I2C operation; the MCU would execute it fully and then return the results without any further host intervention, and without having that protocol hard-coded in its Flash or using a dedicated I2C peripheral. Pretty impressive for 2k of Flash and 128 bytes of RAM, although the memory was limiting for anything more complex. I always liked the flexibility of this approach, maybe one day I’ll update it for a current MCU and release it.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.