Automated window blinds

HW Startups Ideas
#1

For the past few months I’ve been working on automating my window blinds. There is a street light right outside my bedroom window and I don’t like sleeping with the windows uncovered. But I hate trying to wake up in a completely pitch black room. So I’ve been working on a hardware device that will close my blinds at night, and then open them up again in the morning.

Here is a picture of the blinds:

image.jpeg3264×2448 2.05 MB

And here is the motor with a built in gearbox that I’m using:

photo.JPG2448×3264 1.82 MB

And here is my hardware design for the Wifi connected motor controller:

Any thoughts? Suggestions? I’ll post an update later this month when its all working!

1 Like
#2

Great project!

I have the opposite problem. I don’t have very good light in my bedroom and need a controllable way to adjust the light in the morning.

Here’s my project, a 12v PWM Controller. I was planning on generating the PWM signal using a BeagleBone Black.

https://upverter.com/StephenHamer/73d5e5450f4a34ee/12V-PWM-Controller/

Why did you decide to use the Electric IMP as the WiFi + microcontroller for your project?

#3

I don’t have a great answer on the imp. At the time it was because I knew it better than the spark core. I’m not actually even sure which is better, or preferred.

#4

Per my comments in Cost of the Future, I do feel that it’s pretty unfortunate that we need to include a maximal network stack, multi-threading mega-kernel operating system on a serious processor… just to actuate a blind.

Motor costs aside, any thoughts on how you could trim the (computational, electrical…) budget to exclude the $35 network module and all that hefty internet-y stuff?

(Edit: I understand that Spark is able to ship at around $10/unit for large bulk orders. But is that really the best way to go?)

#5

Small power supply, 555,light sensor, relays, feedback micro switches, a few discreet components, perf board, small box, and wire.

job done

#6

ESP8266 + MQTT. Total cost: $3.

This solution only recently became possible, but I think it’s going to become more and more common. This setup allows you to push the brains of your project into any remote server - actuating your widget is no longer a thing of ceremony, it’s just a single message away.

1 Like
#7

Do you actually need Wifi connectivity? Or is it just for getting the time of day? Seems like a BLE solution (such as with a Cypress PSoC 4 BLE device) would have everything you need for direct control and a simple BLE interface to some other BLE control source (iOS, Android, Windows, etc.)

I saw that Nordic has a lightweight IPv6 stack that runs on their nRF51 series, that could be an option to make it addressable if that’s what you’re looking for.

Or maybe you just had the Wifi module lying around waiting for a project to drop into!

1 Like
#8

For BT, I have had good luck with Punch Through Design modules if you wanted to switch over to BLE.

#9

Wow, that ESP8266 is pretty interesting. That’s definitely the cheapest Wi-Fi module I’ve seen. I may have to grab some to play with.

#10

I bet you’re right - BLE probably would have been a great fit, and probably would have had the added benefit of being cheaper.

That said, I wanted it to connect to the wifi in the house, rather than to my phone directly so I would have the most options for driving it. In case for whatever reason I wanted to build and IFTTT for my blinds? Who knows.