This Clock Is Hard: No Arduino Needed

You always hear that people talk about the weather. But it seems to us we see more clocks than we do weather stations. A case in point is [frank_scholl’s] clock made from an old hard drive. We found it interesting that the clock has no microcontroller at all. The custom PCB is all digital and uses the line frequency to drive counters which, in turn, drive the motors.

The one catch is that you have to have a hard drive that uses a very particular motor scheme for this to work. The platter rotation shows the hour and the head’s track position counts off the minutes from 0 to 59. Two buttons can speed up either rotation for the purpose of setting the clock. You can see it all in the video below.

The schematic is a little hard to follow since it doesn’t use conventional logic symbols, but you can probably figure it out. Obviously, it would need a little tweaking for 60 Hz and your motor connections are unlikely to be the same.

Even if you can’t find a hard drive that fits the bill, it would be easy enough to refit a modern drive with some motors. As cool as the logic circuit is, we would have been tempted to punt and do it with a microprocessor, too.

We like when we see technology gizmos repurposed like this. We’ve seen clocks made with meters, for example. We actually saw something very similar to this before, but it did use a CPU.

23 thoughts on “This Clock Is Hard: No Arduino Needed

    1. Quite the contrary. Look, they have the expressions written on them. And is a &, or is a 1, not is a circle at the pin, xor is =1. You can instantly begin reading. Now the weird arbitrary shapes for same symbols make zero sense.

      1. the shapes are far more intuitive than the boxes. i mean there’s only really 2 main shapes and some modifiers. D shape = and, a rounded arrow shape = or. add a small circle to the output line and it inverts those to nand and nor. add a line across the inputs of the or and you get an xor. add the circle to that and you get xnor. the only weird shape is the not, and thats just a triangle with the circle on it. plus the visual distinctiveness alone makes it easier to read. where the boxes look like the russian cursive meme.

        1. agreed, this is far easier to read and rembember any moderately complex logic schematic this way. A carry adder cell is immediately recognizable, but good luck with those square symbols on a tiny schematic in a research paper, where you cannot even read if it is >= 1 or =1.

          The ‘new’ way is barely used in VLSI papers anyway.

      2. Some people see shapes better, some see text better. I do see the shapes and colors better, as i am a visual thinker. For me the old shapes are way more easy to read than the new ones. My wife has the oposite. She once gave me a bottle of beer with this label (warning: nsfw)
        https://www.brouwerijhetij.nl/wp-content/uploads/2015/12/brouwerijhetij_IPA_Export1.jpg and did’nt see the picture. Only after i asked her about why this beer, she saw the picture.

    2. The block symbols were put into the IEEE standard in 1991.. IIRC, the intent was to provide a scheme where new complex operations and blocks could be consistently added, such as latches, latched buffers, etc., to get away from the babel that then existed (see the National Logic selection guide of 1988, the TI guide from the same time, the Fairchild, The NTE, and so on, for little consistency, even within one manufacturers documents). Note that there is an entire language here that allow a standard symbol to be build for a fairly complex MSI-level block, that will be substantially the same no matter who forges the symbol from raw concept.

      The traditional symbols are part of the standard, as well, and were not deprecated. IIRC (no, I’m not going to dig through my files to find the comments, so I may be recalling incorrectly… this may not be the adopted philosophy), the traditional symbols were intended to be preferred when appropriate for clarity, but the square type was explicitly included for consistency.

      I prefer the traditional, especially for complex logic, as there is much less chance to misread and misunderstand. The symbols and spacial relations carry the meaning directly once you learn the language. Think of the common (R-S) flip flop… MUCH more recognizable, even when not drawn in the most common crossing form, using the conventional symbols. The only time I really prefer the blocks (for basic logic) is when the operations are reflected in the mathematical/arithmetic context that they imply.

        1. Yes, they were around, in a number of variations, as well as a number of schema that are now gone (anyone remember the old IBM symbols? Or their symbol for a bipolar transistor?). That is generally how one gets to the standardization process. In this case, there are currently only two competing standards, AFAIK (IEEE/IEC and DIN. IEC did deprecate the traditional symbols at that time, but they still get a lot of use).

          Maybe a new standard should be developed to join them? see https://www.xkcd.com/927/

  1. In the Netherlands, school also teaches us the “new” way and just mentiones the “old” way once. Using this way is universal for most future implementations. The “old” way only works if you already know the function of each shape used, the “new” way allows understanding functions without knowing each shape.

Leave a Reply to kaasbaasCancel 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.