Repurposing IOT Lightbulb Chip For Anything

Home automation products have hit critical mass in the world of consumerism, and now suddenly everyone has a product you can control using some protocol or other. Cree (the maker of LEDs) has a rather cheap IOT-enabled bulb available in Canada and the US for the low price of $15 — not bad considering regular LED bulbs can run you that much, without wireless connectivity!

So if you want to outfit your house in smart lights — great. But what about other things? Well, [Mac Alpine] decided to crack open one of the bulbs to see if he could re-purpose the IOT board. Turns out, you can.

In fact it’s almost too convenient. It’s a remarkably small chip, about half the size of a silver dollar. And it features a small ZigBee radio module. All you need is a 3V power supply, and boom — you have an IOT module that is capable of PWM output. It features an Atmel ATSAMR21E microprocessor which communicates over the radio to a Quirky Wink hub — it can also be triggered using IFTTT.

You can dig deeper into the hack, write your own controls for it — or, you make use of the apps already available for it — but regardless, this could be a very cheap way of adding in some reliable smart controls to your home. Or you could just Arduino everything.

27 thoughts on “Repurposing IOT Lightbulb Chip For Anything

  1. I think Jame, that you need to go on a store-explore. Last month I picked up 6 4000K LED GU10’s for $30 the lot. The store also had packs of 10.

    MiLight and it’s many other brand names are selling RGBW 6W screw-in bulbs globally (80V-240V) for about $10-20, depending on where you look. And Wink’s ‘official’ White, dimmable LED bulbs are the $15 GE. Link (110V) bulbs. At this point, there’s no real reason not to put LED bulbs in every fitting in your house – and Zigbee bulbs are cost effective enough that they are also justifiable.

    Shame Amazon are only shipping used Wink Hubs when you order new ones (I have two unhackable hubs on my desk now). So if you can’t get to a Home Depot, you’re basically stuffed.

      1. I was an early adopter of CFLs. When they were in the $6-$10 range, they used to work reliably for years without exception, had better lumen and color maintenance over time, and started up faster. The power savings over incandescent was dramatic enough that the price was well worth it. Then they got cheaper – in every respect. Wish I could go back in time and buy more of the old ones.

        LEDs are a different matter. Not much power savings over CFL, so they need to run reliably for years to make them worth it. But even now there’s a high enough early failure rate, that it’s questionable whether one can ever realize any cost savings over CFLs. I shudder to think how inadequate the heatsink might be on a $3 LED bulb, it would probably lose half its lumens over the course of a year.

      2. LEDs are only twice as efficient as CFLs. The bigger differences are in the lifetime of the bulbs, and the quality of the light. An LED bulb will last 50,000 hours at 90%-95% of the light output of new, regardless of how frequently you turn them on and off, they are at full brightness instantly, and they work in subzero temperatures. A CFL will last about 25,000 hours, but will only survive a few thousand power cycles, they take a long time to come up to full brightness, they output only about 50% near the end of their lives, and may not come on at all if it’s too cold.

        Neither CFLs nor LEDs will long survive the high temperatures of unventilated, enclosed fixtures.

        A LED is a better deal if it is less than twice the cost of a CFL for equivalent output. And the higher your electric costs, the better deal is the LED.

        1. For me the quality of light and the brightness from LED was the game changer. In my 4″ cans, I went from MR16 low voltage, then changed to GU10 high voltage since they had better sockets (heat was killing old sockets), and finally ditching them for a medium-base socket and some Cree 4″ trims 1100lm. At $12 per can it seems expensive, but benefits include: that I have sealed trims now (no air exchange to attic through old cans), no shadows (better light flooding), brighter and less yellow quality of light (it says war white temp, but it is more of a daylight halogen color), cool fixture (generates no heat to the room).

          In my opinion it was a bargain at $12 per and changed my house completely, something that would have cost a lot to have an electrician come add fixtures to dark areas and circuits for more wattage.

    1. One reason why I still haven’t changed my can lights over is dimmability. Sure you can dim LED bulbs but at its lowest level, they still output quite a bit of light. There’s now way i can match the low dim light coming out of a BR30

      1. Ever since Triac’s, some listeners of nightime AM radio have been on guard. That PWM going to a transistor… They’ll cheat on the necessary filtering just like always.
        Barry Goldwater the proto-conservative and ham tried to outlaw them years ago. I think the mf and hf bands are doomed.

      2. This is the big concern. There’s no reason not to test one, if you’ve got access to a spectrum analyzer, before you go replacing the whole bunch of them.

        It’s probably neither practical nor UL approved nor feasible for a great number of sockets, but I wish someone would come out with a noise suppressing collar that would screw into the socket and then accept the bulb or something to that effect.

        Alternately, the manufacturers should really do their own RF testing to an agreed-upon standard and release a line of premium certified quiet bulbs, but then doing that would be admitting certain things and expose them to liability from the FCC (who, while loathe to do anything about part 15 emmissions from CFLs/LEDs, probably wouldn’t hesitate if someone had already done that much of the work for them).

        Third-party service perhaps? Bueller? Anyone?

  2. Wink needs to stop being dorks and allow people to hack the hubs. Their WINK cloud based system is not the best, but running your own automation server at home is the best solution.

    1. Try sending them a POLITE* email suggesting it, possibly with a suggestion on how it might work, and why it would be a great business plan for them. It’s never going to happen if people don’t ask. They won’t be reading every comment on every IOT post on every blog on the internet, after all.

      * Noting that calling them dorks and insulting their product isn’t very polite.

    2. Has there been any active hacking going on with the WINK hubs? I have one I bought from Home Depot not for home automation but because the presence of all those radios in a single system caught my eye. However, in my (admittedly brief) searches I haven’t seen a lot being done with it beyond the original rooting of the device and spinoffs from that work.

      1. I also got one because of all the radios, especially the CC1110 radios for the Lutron and Caseta protocols (since I have a CurrentCost electricity meter which broadcasts in a band that the CC1110’s can pick up). Mine also has the ICSP and JTAG ports for each microcontroller exposed on the board, suggesting that it should not be too hard to write one’s own code for those micros. and get it loaded. Ideally, one would be able to back up the firmware before blowing it away, but as you say, nobody has really done enough digging to figure out what is possible.

        I also wonder how the Linux OS actually communicates with the various microcontrollers. I assume there are a bunch of serial ports interfacing with the micros, although that could also be something like I2C, or SPI. It would be a lot simpler if they showed up as serial ports, of course :-)

        1. Yep. They show up as serial ports, as I recall. So, if you just want radios – not to use them for home automation, you’re not far off.

          To my knowledge no-one has done the ideal yet, which is to replace the aprontest utility with something opensource and optionally install redsleeve or debian.

          Of course,

          Now that it’s become very difficult to hack them (you have to exploit the updater partition, not the standard app partition), things have slowed down.

      1. That doesn’t count – it still goes out to the cloud.

        It doesn’t matter for a lot of things, but the extra latency is annoying for some Home Automation tasks – there’s already enough latency added with all the various protocols talking to each other…adding some more to go out to a server somewhere on the internet is less than desirable, and adds additional failure modes.

  3. Seems to me like you should be able to find the boards somewhere for less than it costs to buy a full-blown light.

    Of course, if you already have a broken one, that’s one thing.

    1. Which is much easier to get hold of in the US! You can even get a cheap wink hub by buying some.

      I imagine almost (if not) every bulb can be hacked this way. I’d almost be more interested in finding one that couldn’t!

Leave a Reply

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