I've been doing stepper motor and serial communication with Arduino for a while, but I now have a project that requires circuitry beyond my knowledge.
I recently acquired some flip dot strips (these guys http://i.imgur.com/ongOn.jpg) and am looking to build a driver for them. The trick is (well, it may be really easy, but I'm not sure how best to do it) is that I need to be able to send a charge either direction to control their orientation. My plan is to wire the display into a matrix, with 30 leads for the columns and 7 for the rows. What I don't know how to wire is the stepping through of each row of each column and being able to send a charge in either direction.
All I could think of was using something like a chain of shift registers to step the charge down each pin, and using some transistors and two pins on the Arduino to control whether a positive, negative or no charge is being sent through the registers. I'm not sure if that would even work, but I feel like there must be a better chip for this.
I am into LED displays, and thought of flipdots as rather antiquated, but I hadn't realised that they are fast enough for video !!
It looks like some of the dots get stuck, which I heard was one of the drawbacks with these displays, but I would love to play with these if I had time / and loads of money.
From the Polish company that makes them
A board 14×28/0.53? is listed at 450 EUR.
I am sure there is a way to drive them with an arduino, but at 3am my brain is not getting round it !
Thanks for the info, that Polish company you mention is actually where I got these, but instead of getting their 450 EUR 14x28 board and driver, I have just the bare strips and need to build that driver myself (multiplexing the dots and flipping the current). That's the part I'm hoping to figure out and am hoping to do for a lot less than 450 EUR.
What comes to mind to save loads of Hbridges is a 16v centre tapped supply ( with big electrolytics to supply the pulses )
with the dots coil connected to the 8v centre tap.
A transistor ( or ULN2003 for 8 dots ) would ground the other side of the coil for driving one way, and a high side version of the ULN ( I cant remember its number ) would pull it up to 16v to drive the other way.
As with H bridges, you would have to ensure both drive logics are not on at the same time...
What do you think of something like a single H-bridge attached to a demultiplexer to iterate through each of the rows/columns? I'll have to start looking at actual chips for these as I'm not sure what power loads they can handle or how fast the demux switching is.
So here's what I have imagined at the moment (hopefully someone can tell me if this is actually feasible).
I'm currently planning to use an SN754410 for my H-bridge ( allowing me to pass through the higher voltage in either direction) and three HEF4067B dual-voltage demultiplexers, which (I think) can handle the voltage I need and work with the current flowing in either direction. (If someone could confirm this for me, I'd really appreciate it.)
I've laid out the wiring here. Ignore the wire that isn't actually connected, I ran out of pins with the Uno and didn't want to rewire it to a Mega just yet. Additionally there probably need to be some resistors along the lines to the flip dots.
The red and black wires carry the normal 5V and the grey and orange lines carry the (flippable) 7.5V, which pass into the demuxes' common input/outputs (one line into the demux controlling the rows, the other into the two controlling the columns). The white and yellow lines control which direction current is flowing through the grey and orange lines, and the green lines control what the demuxes have selected.
The H-bridge has an enable pin (1,2EN) that, when low, places the inputs and outputs in a high-impedance off state. Similarly the demuxes have an ? pin that, when high, places all the outputs in a high-impedance off state.
Currently I have these wired in the diagram to permanently enabled, but I'm not sure if the Atmega will switch the 1A and 2A inputs synchronously enough that there isn't a brief time where the 7.5V input is shorted. Similarly I'm not sure if the ? pin can be safely unused and rely on synchronous pin changing to select the right output.
