ShiftBrite is a simple device I am designing and producing. It allows easy control of a bright RGB LED. The interface is a straightforward clocked serial data line and a latch input. All signals are buffered and passed through for good performance over long cables and daisy chaining many devices. Many ShiftBrite devices can be controlled from any type of controller that supports clocked serial data output, which is practically all microcontrollers and even PC parallel port or FTDI bitbang adapters.
Whoa, this wasn't quite ready for primetime yet. Anyway, I'll have a bunch of these running at the Maker Faire Bay Area 2008 this May 3rd and 4th. Well, assuming my Chinese assembler is able to meet the deadline.
AF: BlinkM and ShiftBrite may look similar, but they work differently and are intended for different markets, though some overlap exists. The ShiftBrite was actually in the design stage before the BlinkM was released. If you're writing code from scratch, ShiftBrite is actually easier to use because you simply pulse one line and shift the next bit of data on the other. When you've sent the whole command, pulse the latch line. This means there is no need to implement the full I2C protocol. ShiftBrites are also daisy chained, the identity of an LED is defined by its location on the string, rather than having to pre-program every LED like the BlinkM. However, this means you must send out the status of every LED even if you only want to change one; with the BlinkM you can speak to one LED on the I2C bus and leave the other alone.
The ShiftBrite is intended for users who want to control several RGB LEDs in 30-bit color, and don't need the scripting and auto-sequencing of the BlinkM standalone modes. It uses a chip that needs no programming and does current regulation, dot correction, thermal auto-shutoff and restore, and buffered outputs for longer cable runs between LEDs. Since the ShiftBrite uses inexpensive parts, based on current quoting I should be able to start these near $5 in single quantities with discounts possible for multiples.
Anyway the launch is supposed to happen around Maker Faire, so see you then....
Posted by: Tod E. Kurt on April 2, 2008 at 4:26 PM
Yup, ShiftBrite is great for when you want to control a large number of RGB LEDs and want to do things like video with them. The chip in ShiftBrite is really cool and I'm so glad you're making a board for it, since those QFN packages are hard to work with.
BlinkM is more for when you have limited processing on the controller side and want to push off some of the RGB color changing to the LEDs themselves, since BlinkMs can run little "programs" of color changes. Due to the limitations of I2C (127 address range), it's not really made for large arrays.
BlinkM & ThingM will be at Maker Faire too. Garrett, I hope we can meet up and say "hi"!
Void your warranty, violate a user agreement, fry a circuit, blow a fuse, poke an eye out. Make: The risk-takers, the doers, the makers of things... Welcome to Make: Online!
More contributors: Mark Frauenfelder (Editor-in-Chief, MAKE magazine), Kipp Bradford (Technical Consultant/Writer), Chris Connors (Education), Diana Eng (Guest Author), Peter Horvath (Intern), Brian Jepson (O'Reilly Media), Robert Bruce Thompson (Science Room)
Weekend Project: Making Char Cloth
Learn how to make a cheap and effective fire starter made from an old t-shirt. To download The Char Cloth video click here and subscribe in iTunes. See Char Cloth in action with the Fire Piston from William Gurstelle....
More...
RSS
Isnt the blinkm easier?
Reply to this comment
Whoa, this wasn't quite ready for primetime yet. Anyway, I'll have a bunch of these running at the Maker Faire Bay Area 2008 this May 3rd and 4th. Well, assuming my Chinese assembler is able to meet the deadline.
AF: BlinkM and ShiftBrite may look similar, but they work differently and are intended for different markets, though some overlap exists. The ShiftBrite was actually in the design stage before the BlinkM was released. If you're writing code from scratch, ShiftBrite is actually easier to use because you simply pulse one line and shift the next bit of data on the other. When you've sent the whole command, pulse the latch line. This means there is no need to implement the full I2C protocol. ShiftBrites are also daisy chained, the identity of an LED is defined by its location on the string, rather than having to pre-program every LED like the BlinkM. However, this means you must send out the status of every LED even if you only want to change one; with the BlinkM you can speak to one LED on the I2C bus and leave the other alone.
The ShiftBrite is intended for users who want to control several RGB LEDs in 30-bit color, and don't need the scripting and auto-sequencing of the BlinkM standalone modes. It uses a chip that needs no programming and does current regulation, dot correction, thermal auto-shutoff and restore, and buffered outputs for longer cable runs between LEDs. Since the ShiftBrite uses inexpensive parts, based on current quoting I should be able to start these near $5 in single quantities with discounts possible for multiples.
Anyway the launch is supposed to happen around Maker Faire, so see you then....
Reply to this comment
Yup, ShiftBrite is great for when you want to control a large number of RGB LEDs and want to do things like video with them. The chip in ShiftBrite is really cool and I'm so glad you're making a board for it, since those QFN packages are hard to work with.
BlinkM is more for when you have limited processing on the controller side and want to push off some of the RGB color changing to the LEDs themselves, since BlinkMs can run little "programs" of color changes. Due to the limitations of I2C (127 address range), it's not really made for large arrays.
BlinkM & ThingM will be at Maker Faire too. Garrett, I hope we can meet up and say "hi"!
Reply to this comment