Improvements to first clock; better RTCs found

A project log for A simple non-bomb clock or two

Gets attention, tells time, promotes knowledge of the electronics hobby & DIY/STEM in general, and advertises my local hackerspace a bit

pointyointmentPointyOintment 09/23/2015 at 17:513 Comments

Yesterday I improved the signage on my clock:

I also added a 15-ohm, 5-watt-looking power resistor as a dummy load so the battery stays on and I don't have to press its button every time I want to check the time. My calculations say it's drawing 1/3 amp and dissipating 1 2/3 watts. It gets pretty warm. I could add a transistor and modify the code a little to pulse the load, but I calculated the battery will last over 12 hours like this, so it's not worth it, especially because:

I was at the electronics store yesterday and I found that they now have DS3231 breakout boards in stock! The DS3231 is a much superior RTC to the DS1307. It's temperature-compensated, making it far more accurate, and it can operate with supply voltage below 3 V. Even better, they fit directly onto a Raspberry Pi's header, they're about 1/6 the size (because they're not part of Seeed's Grove prototyping system, which my DS1307 is), and they're about half the price! I bought three, obviously.

This means that I can build a miniaturized version of the clock immediately, without having to get a softRTC running first! So I'm hoping to have a new clock to wear to school tomorrow. The current one's cable is rubbing my collarbone irritatingly, and people seem to be getting slightly bored with it (those who've seen it already, anyway). The next one will pin to my shirt, I think, and have a laser-cut/etched frame, and maybe a slightly larger display.


PointyOintment wrote 09/24/2015 at 03:20 point

The DS3231 is now working on my current clock. Turns out it's a 100% drop-in replacement for the DS1307, which is really nice. I just needed to set the time on it (using a separate sketch I found for that purpose) and that was it. Now I just need to physically build the miniaturized version.

  Are you sure? yes | no

PointyOintment wrote 09/23/2015 at 21:08 point

I unplugged the resistor after running the clock for about four hours today, intending to turn it off. However, it stayed on, and it has stayed on without the resistor for about three hours since then. Perhaps there's some hysteresis in the battery's load sensing. I expect this to extend the battery life for far longer than I need for today; I don't know if it can be used reliably, though.

  Are you sure? yes | no

PointyOintment wrote 09/24/2015 at 02:41 point

Turns out the battery stays on with only the load of the clock now, even when started without the dummy load. I have no idea why.

  Are you sure? yes | no