Go Down

Topic: Centipede Shield - also a contest! (Read 24253 times) previous topic - next topic


Nov 09, 2009, 12:53 am Last Edit: Nov 09, 2009, 12:57 am by kyndal Reason: 1
i'm working on a multifunction Alarm system
that needs a loot of  I/O

2x   keypad.LCD,Speaker,key,RFID type "door units"
      Separate micro  "Serial protocol"

1x   www, email, IM, text message  
     alarm indication and remote control interface "Serial protocol"
     Embedded linux,  "dd-wrt, NSLU2"  

16x window sensors
4x   glass break sensors (analog piezos)
4x   PIR motion sensors (on/off)
4x   Smoke/Fire alarm inputs (on/off)
1x   powerfailure sense  (Tamper, Fridge, freezer )
      (on/off maby analog Temp)
2x   "Guest" "door bell"  zones (mailbox and driveway visitor sensors)

2x   outdoor siren, Beacon light (on/off)
2x   indoor alarm siren (on/off)
4x   indoor lights interface (on/off)
2x   outdoor floodlight  (on/off)
1x  SD-Card  for Storage of "event log"  

All of the units need a "TAMPER" sensor, but this can be as simple as just one wire through them all in serial (if broken)

interface to camera server (unknown.  Rs232?)

so yea.. looks like i need Some additional I/O's for the Arduino.
or maby just something bigger ;)


im working on a project for a friend. what i call a "magic" hallway. using a sonar sensor, many pir sensors and a ton of led lights.  as one walks through the hallway, the light follows them. get brighter as they get closer and the gradually dimishes as the move away from the light. funny, i was looking into i2c protocol for this project, and then you come out with a shield that would be perfect


I'm working on building a custom control interface for a model train set my dad is building.  He's gotten pretty ambitious with the track layout, and I know I'm going to need more than 20 i/o pins on the arduino for all the electrical elements I'm going to have to control.

Essentially, we're going to be using multiple arduinos (or maybe just one with a centipede shield) communicating with a touchscreen UI running on some sort of embedded device to control all of the track elements (turnouts, forward/reverse, engine speed, turntables, sound effects), lighting effects (buildings and such), and some safety systems (the layout includes a couple turnaround loops and it's conceivable we'd have two trains on the same loop at the same time - the arduino could shut down an entire loop automatically if it detects a problem).

I originally submitted this for the Nokia n900 push project.  Sadly, it was rejected.  We're moving forward with the build, though.  I'm probably going to re-purpose a TI Beagleboard for the user interface, since it's silent and the serial USB connection with the arduino should be trivial to set up.  Either that or a Nokia N810 or a Pandora handheld -- haven't decided which one I think I can spare yet.  :D

Some of the technical considerations for this project can be reviewed at rootaction.net/~boltonc/train.pdf  Ignore the prototype track; we're skipping that step and working directly with the final track layout.

We're also looking into what kind of stepper motor we would need to build our own custom turn-table, and whether it would be more cost-effective than purchasing one at retail.

The project is long-term, but is already under way.  We finished building the table top this afternoon.  It's sectional in 6 pieces since the overall layout is 10x12 feet.  We'll be getting together in another week or two to finish the structural build (assemble legs and casters on each section and bolt them together), after which we'll be working independently; he'll lay the track while I start prototyping circuits.  It'll take us a month or two before we're ready to put it all together.  I'm not sure if that meets your requirement for 'something plan to build immediately' or not, but I do know the centipede shield will save me some coding and cash.

Even if you don't decide to throw a freebie my way, I'm interested in purchasing one when you're ready to start selling them.  Do you know yet when that might be and/or where you'll make them available?


I would like to interface with a relay board - probably dead-bug style - to switch line current to many several channels of animated christmas lights.

I foresee needing near 50 channels to have a truly explosive experience.  So far my plan involves using shift registers - which I'm quite comfortable with working on, but this board would simplify things greatly...That, or I pick up an arduino mega.

Then it's just a matter of writing sync software.


ok macegr,
heres my idea.  I want to make a cheap and fast motor control system for robotics.  something like the sony aibo used to use...1 brain with tons of sensor in and coordinated servo control with balance...something that stabilizes my robotics platforms as a standard.  then others can implement my code base and control system for their robots.  If every one of us makin robots has to teach them all how to balance and self correct, this will take forever.  If i make a fast atmega328 program and a custom app for motor control,  i can create a standard kit for robotics platforms that is cheap and EASY....so hook me up bro... i need that shield for some serious I/O.  awaiting that PM.  Legion.


Nov 09, 2009, 02:31 pm Last Edit: Nov 09, 2009, 04:38 pm by efluon Reason: 1
great! this will make it easy for me to build a (hardly original nowadays) stepsequencer for drums. sort of like monome but 4 rows with 16 steps (using something like the sparkfun 4x4 button pads (not allowed to post that here)). there will be 16 rotary encoders, how to use them is something i'll still have to figure out for myself, but i guess it should even be possible using two of these shields!


This shield could be ideal for a fancy wall mounted analog clock: 60 outputs which drive 60 leds (for example, high brightness, white leds). One led for each minute (6 degrees), closing the circle. Leds are soldered in order to project each light beam almost parallel to the wall surface (it's preferable to use very narrow beam leds).
The project is very, very simple, since the software must basically drive in a very clever manner 60 different outputs (2 pushbutton inputs for time setting and/or IR input to do the same job).
The hardware part is also quite simple: if we don't want to rely on Arduino clock settings, we only need to add an RTC to the Arduino in order to keep the time as stable as possible. For the code, it's up to our fantasy to find a way to display the time. Some ideas:
1. All leds are powered off except led (or group of 2, 3, 4 or more) to show the hour; slow blinking led (or group) to show the minutes and optionally fast blinking led (leds) to show seconds;
2. All leds are powered on (pay attention to maximum sink/source current from IC) and 2 different blinking rates to show hours and minutes: this way the clock could be used also as  a very original wall mounted lamp/clock!
3. More ideas, please! (consider that you can find other purposes for this gadget, not only a clock, but: countdown timer, analog thermometer, vu-meter, whichever-physic-environmental-variable-meter...)



I find this perfect for application in driving all the indicators and switches on the overhead panel of a home-built flight simulator for a B373-700 airplane.

Usually you need to run a rats nest (hundreds) of cables to the IO board located elsewhere in the electronics bay.

Using this, with a local power supply, a simple two or three wire cable is all that is needed

All the best


I have an old electric-air-blower-powered organ, with several octaves and a chord section, which works by opening/closing holes in a large box (picture a saxophone's key mechanisms).

I'd like to retrofit it with lots of solenoids for pushing the buttons, and interface it to MIDI via an Arduino.

Your Centipede Shield should provide enough outputs to do it right, with a few pins left over for control buttons and an LCD display!


Nov 10, 2009, 03:31 am Last Edit: Nov 10, 2009, 03:31 am by zzz_tek Reason: 1
I am going to be building a custom EEG/EMG monitor. I need multiple channels so I can monitor the electrical activity of various areas of the brain and muscles I'm studying.

This many I/O inputs would allow me to increase my ability to monitor and interpret the electrical impulses from a variety of sources at the same time.


I had an idea for a "distributed production" alphanumeric electronic demonstration sign. the idea is demonstrators get a kit in the mail containing precut led strips and wire, and construct a 2X4 foot 14-segment alphanumeric 1-character sign. the wires all go out to a plug at the bottom. Now - one guy comes out with the same sign, but he also has the control board - arduino+centipede, to control all 5 signs (incl. power supply and switching with uln2003) , and end up with a 5-letter, 10X4 foot electronic sign! I know this does not make use of the input and only uses the centipede as an output expansion, and I know that for the same price I can almost get an arduino mega (which is good for 4 1-letter signs) , but I still want it!
I also discussed this with some guys from graffiti research lab (google it) and they were cautiously interested.

diagram will be up soon. thanks for reading!


Tomas: This won't help you in the contest, but you may want to suggest it to the GRL gang....

Instead of LED signs and lots of wires, put the 14-segment displays on the backs of jackets, using EL wire, powered by batteries, uniquely labeled with a QR code patch on the front (where it's harder to see mid-demo), and control it though an iPhone app. Once enough of these are floating around, you could do impromptu flash-mob signage.


Lots of good ideas so far, glad so many of you think this might work for your projects.

Based on the entries so far, I want to clarify a few things.

PWM may be possible, but probably not too accurate over I2C. I would say that if you're controlling LEDs, you should plan to just have them either on or off. The same goes for LED matrix driving...it could work, but there are probably better and more reliable ways.

Since these are general purpose I/O, I'd like to see at least some mixing of functions. Input-only and output-only are valid uses for this board, but don't exploit any advantages this would have over some shift registers.

I am also paying a lot of attention to how the entries are written, and really appreciate the ones with full details and examples of what you've already done to work on the project. From these posts, it's pretty easy to tell when someone knows what they're doing, and has the ability and motivation to follow through. I would really like to get feedback on these and not have them sit on a shelf until the Arduino is obsolete. If any of you want to expand on your project a little more, that's fine. Feel free to post more details, as long as you get them in by Friday the 13th they'll be included in the overall decision.

Thanks to everyone so far....
Unique RGB LED Modules and Arduino shields: http://www.macetech.com/store


Consider mine withdrawn.

I needed more I/O which the Centipede could give me

But i also need more room for my program so i am upgrading to the Maple (http://leaflabs.com/Maple)

and that will give me the I/O needed also. But if i need more i can add your board on later.


i'm intending to make a hologram generator using the pesistence of vision concept.

its going to consist of two 256x256 RGB matrices (think big, might get at least a small part of it right  ;) ) mounted on a axel (1800rpm+) as shown in a rough top down sketch.....(gimp is not a friendly alternative to MSpaint, but i not complaining  :-/ )

i intend it to show 3D renders at 30fps min when complete.... so i guess the shift register nature of the centipede wil be useful...

[edit] sketch after break...

Go Up