Sunny Custom Keyboard Illuminates The Past

Ever wonder why keyboard number pads and telephone dials have reversed layouts? Theories abound, but the most plausible one is that, shrug, it just happened that way. And now we’re stuck with it.

Well, that answer’s not good enough for [Jesse], so he punched up his own keyboard design that combines the golden years of function-rich Sun and IBM keyboards with Ma Bell’s DTMF number arrangement. That’s right, Sundial has 24 function keys total, and the number pad matches Ma Bell’s all the way down to the asterisk/zero/octothorpe pattern on the bottom row. How do we know what the unlabeled ones are, you ask? It’s all mapped out in this layout editor. We love that it has all the key lock indicator lights, because that practice should’ve never faded out in the first place.

Though inspired by this beautiful unicorn of an Arduino keyboard we covered a few months ago, the Sundial uses a Teensy 2.0 to translate [Jesse]’s Cherry MX clone-driven wishes into software commands. It’s also painstakingly hand-wired, so here’s the build log for you to drool over. Just cover up your keyboard first.

35 thoughts on “Sunny Custom Keyboard Illuminates The Past

  1. The answer I’ve heard dates back to the rotary dial phone days, where 0 was, technically, 10, and you couldn’t send 0 pulses to the switchboard. Hence, why phone dial pads had 9 near-ish 0.

    As for computers and keyboard numpads? Well, you _could_ literally store zero, and 0<1.

    1. You used to be able to tap out phone numbers (a bit like Morse code really) into old rotary phones by taping with one finger on a hangup button. One tap was dialing the number 1, wait a short duration, and tapping two times and that was dialing the number number 2, …, tapping times was dialing the number 9 and finally taping the hangup button 10 times was dialing the number 0.

      You might ask why this was useful, but on some private payphones they actually locked the dial in place until a coin was inserted to unlock the dial. So ….

      1. This is the exact reason why emergency services aren’t reached by dialling 111… too many kids “dialling” it by repeatedly picking up and dropping the handset.

        You see it in some very old movies, they pick up the handset and prod the hook switch a few times.

  2. I would have thought the most obvious reason is computer scientists count from zero. So it makes logical sense for the numeric keypad to start at zero and go through to 9. Your average [American] person however starts counting from 1 and sees zero as a way of making 10. If you add to that what AmEv says about pulse dialling phones sending 10 pulses for a zero that made it a requirement for mechanical phones to have zero at the end of the dial. When making a keypad the numbers on the dial were simply transposed onto the phone keypad, ie in order of 1 to 0.

      1. so even a quick glance at the paper has me wondering how anyone can claim that their study was scientific in any way or that it was carefully designed. quoting the original paper:

        “Sixteen different arrangements were selected for the first part of this study. These arrangements were compared by separating them into groups of three and having a different sample of six employees try all three arrangements during each of five sessions. The arrangements and how they were grouped for the study are shown in Fig. 3. ”
        Fig 3:

        Two things jump out at me right away, first of all the group size is incredibly small to make any sort of extrapolation that the group represents the standard deviation of the overall population. The second thing that stands out to me is that there are two layouts that are repeated, meaning that there are in essence 14 options. To be fair, the rest of the study seems well done and no glaring errors pop out to me but you cant argue that the layout was carefully designed when there are such glaring errors in their testing procedure regarding potential layouts. It looks to me as if someone had their preferred layout and massaged the experiment to suit their conclusion.

        1. Agreed. I didn’t say their methodology was right, and many of the proposed arrangements seem…improbable.
          I would posit that the calculator, reverse calculator, two row and rotary dial arrangements make sense to evaluate, and the others are “filler”. Some of them, cross and double diagonal, for example, are downright wacky.

          Still, at least they didn’t just pick something and say, “looks like this should work”. At least, they made an effort…and documented it.

    1. ..and, *why* was shorter keying time the deciding factor?

      Back in the day, before digital electronics were cheap and plentiful, the DTMF receiver at the central office was an expensive piece of equipment, and had to be shared among incoming lines. (modern exchanges use digital DTMF decoder chips, one per line card) So, when you picked up your phone to dial, you grabbed the DTMF decoder, and held it until you had finished dialing your number. No one else in your line group could use it until you finished dialing.

      Shorter keying time => less wait time for the DTMF decoder

  3. I thought it was because the keypad wasn’t based on telephone dial pads at all, but on the pads of simpler calculators, hence the larger 0 key and presence of all those other calculation-related buttons. A computer keyboard with built-in DTMF pad would instead have likely retained the keys A through D even after they were dropped from regular phones.

    1. The DTMF spec originally supported 4 columns and 5 rows to allow the A-D keys down the right side.
      Somewhere I have a text book from school with a full dive into how it worked.

      That being said… I am not sure I ever saw so much as a photograph of a consumer telephone dialpad with the 4th column. The only ones I ever saw/used were mostly in strange half-proprietary sets used to program the phone systems or computer controlled systems attached to the phone network, and had much more than 12 keys, usually with the extensions connected through internal modems or teletype-like interfaces.

  4. I was about to say, “octothorpe, what the hell?”
    Then I reminded myself, “you know what? At least she didn’t say ‘hashtag’ like all thosr darn kids” — and I smiled. I learned a new word. Now get off my lawn.

  5. “Ever wonder why keyboard number pads and telephone dials have reversed layouts?”

    Ever wonder why telephone dials don’t have keys or pads and a partial circle isn’t in any way the reverse of a rectangle?

  6. The story I heard decades ago was that accountants could run rings around anyone on a ten key, so they handicapped them with the upside down keys. They would be some of the first users along with other professionals in office situations, not the public yet. That DTMF converter had to have 50mSeconds to acquire the correct tone so zippy had to get his fingers tied.

    I have taken many a phone and reversed the 2 lines and take out the keycaps and correct them, even Ma Bell (GTE) ones in the day. It irks me that I can’t have one correct keypad on my phone, it’s all software. It should be an ADA required choice.

    The warped number-line of 123…90 shapes the math mind of learning challenged kids before they know it. I wonder if math scores are better in countries like Sweden and down under than the US, duh. I have seen a Walmart yardstick with zero at the beginning end and 1 in the second inch. All other rulers and tapes have omitted the zero and have a 1 to the left of the line for second inch. Imagine a voltmeter scale without zero at the left end.

  7. I like the keyoard layout, I really miss the old Sun Type 6 keyboards with the functions keys to the left, where they make a *ton* of sense to me. Sorta like my old Northgate OmniKey Keyboards as well. I’d love to see this made into a 10 keyless module as well.

    The next issue I have is no backlight. Right now I’m using the CODE keyboards from (did I get that right?) and I really like the feel, but I dislike the lack of durability of the keycaps. Anyone know of a good source of black keycaps, with white letters that are doubleshot? And compatible with Cherry Browns?

    But in general, this is a sweet looking build, I’m really impressed with the photos of how he did all the layout and test fittings.

  8. The thing I remember about the 24 function keys on IBM keyboasrds is that they doubled as macro keys. One of the other function keys down the bottom left was the on/off record button, and another was the ‘replay’ button. Press “record on”, type your sequence including control characters, press the numbered function key, press “record off”. Press ‘replay’, and a numbered function key and your recorded sequence came spilling out. Of course, many folk where I worked recorded their logon credentials, but it was an incredibly useful thing to have for coding and system admin.

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.