I build the first version of my giant spin art back in may after reading build instructions in Make magazine Volume 25.
It work okay but had a few problems. It was direct drive driven and I was worried about burning out the motor. It was scary fast and got unstable as it speeds up.
Three days before Maker Faire I decide to rebuild it. First thing I did was make a new box out of lumber one that I could mount the motor underneath. Then used two pully and a belt to reduce the speed of the motor. I mounted a sheet of thick foam board on top of the center pully. It was much lighter then the wooden board I was using before making it much more stable and safer. I used some plastic fencing, duck tape and paper as a splash guard so people wouldn’t get covered in paint as it spins. The plastic fence worked but looked ugly.
More paper – I purchased 100 sheets of poster board that I cut in half (good quality). As well as a book of painting craft paper with 160 sheets (No as good). For a total of 360 sheets of paper. I was completely out of paper, and things that I could use as paper by noon of the 2nd day. This was good thing as well because it gave me a chance to wonder around the Faire and see what everyone else was making.
Safety Fence – Three kids fell in to the spin art box! No one was hurt but they did get covered with paint. The plastic splash fence that I used as a splash guard couldn’t hold the weight of a child as they reached in to the center to pore paint on to the board.
Ugly Splash fence – The thing looked ugly from the outside. It could have easily been fixed up by putting spin art paints all round the outside of the splash fence.
An assistant – Very important, It was almost impossible to go to the bathroom with a line of children waitting patiently for their turn. I also was trying to get a photo of every child with their spin art but it was impossible while operating the box at the same time.
A time laps from above – It would have been really interesting to watch the entire thing from a camera mounted above the box doing a time laps with a picture once every 5 secs.
This project has somewhat hit a road block and with 12 days left before Maker Faire I’m starting to panic. This leaves me with ~4 good working days left. Here is my current stats.
Editor and Simulator
At burning man 2010 I saw the Illuminatrix made by Ben Delarre. A RGB LED color grid that was programmable from a JavaScript editor that he made. I was impressed when I saw it at burning man and contacted him early after I started this project to see if I could use his editor for my project. Not only did he let me use the editor but he said that he would change it so it would work specifically with my globe project. A week later and I had my own HBD5000 page on his website with an editor, simulator and a few example animations. Amazing Ben Delarre is awesome.
Example animations:
Click the images to see a simulation.
Software is basically done, I’m still having problems steaming frames at full speed over bluetooth, but half frame rate is works great. On the Arduino Mega I can buffer ~19 full color frames and the Arduino uno I can buffer ~10 full frames. As I am displaying the frames to the users I can stream in new frames over bluetooth. The whole system works pretty well and with so little time left I probably won’t be changing it anytime soon.
The Mechanical
The first wheel I made was a 6 foot monstrosity of bent tube stock. While I was doing my testing it was a little unbalanced and bent under the stress. As soon as it bent even a little the problem got exasperated and by the time I was able to shut it down the whole ring was bent out of shape and unrecoverable. It was an expensive and time consuming mistake. Luckily I was not running it at full speed or it could have sheared off its hinge and hurt someone as it flew though the air.
At this point I decided to shrink my project down from a 6 foot tall ring to the biggest bicycle rim that I could find. The bicycle rim is already sturdy enforced with the spokes, light and balanced. The project is smaller but much safer.
I drilled some holes in the rim for wires to run though and buffered the LEDs with expanding foam. Its not pretty but it works surprisingly well.
Next I secured the Led boards, wires and the controller boards to the spokes of the bicycle rim. These components will be spinning at great speeds so I took extra care to make sure that they won’t be going anywhere.
Next I made a box out of wood with two pillows bolted to a center brace. A pillow is ball baring mount that should let the center shaft spin freely while saying in position.
In this video/pictures you can see the LEDs lit up. I used a C Camp to keep the wheel from falling over. Ignore the voices in the background they are from the scary ally behind Vancouver Hackspace.
Power
I used a standard computer power supply to power all the LEDs and electronics on the wheel. This works really well and I suggest it for anyone that is powering large amounts of LEDs. The power supply comes with both 5v and 3.3v rails. A common computer power supply also filters and smooths out the current before passing it on to the LEDs.
There is a simple trick for turning on a PC’s power supply without a PC. Basically you need to short the green {PS_ON} pin with any of the black wires {COM} on the power supply.
I am using a 3 connection slip ring to transfer power from the stationary base and power supply to the rotating wheel. There are lots of industrial slip rings available online but they are pretty expensive. You can make your own slip ring but with something so important to this project I decided to buy one.
Safety at Maker Faire
Because of my previous mishap with the larger ring, and because the ring is spinning so fast the good people at Vancouver mini Maker Faire had valid concerns about the safety of this project. The code name of this project probably didn’t help much either. HBD5000 = Hippy Ball of Death 5000.
I decided to encase the spinning parts of my project with Polyethylene terephthalate aka PET-G or better known as the stuff that pop bottles are made from. Its has a few good properties that plexi glass doesn’t. Its easy to mode in to shape, it bends and warps instead of shattering, and most importantly its cheap.
They also connected me with Farrell Segall who has successfully created a few different POV projects that he plans to bring to Vancouver Maker faire this year. We met up at the Vancouver Hackspace and he showed me his impressive POV projects as well as some other very interesting projects that he is working on. He suggested that I make a mount at the top of the POV with a pillow so that the Ring won’t be able to wobble too much. Great idea!
With the top mount and the PET-G even if this project does spin honorably out of control it shouldn’t hurt anyone.
Issues
The bluetooh receiver broke while I was mounting it to the Arduino board. Ordered another one and it should be here 2 days before Maker Faire. If it does not get here in time I will pre program the POV with a single image and use that to demo it.
I have no welding skills and i’m reliant on others to do the welds that I need for this project.
I have burnt out two used motors testing this project. I have purchased a bran new motor for Maker faire but if it burns out on site I will not be able to replace it.
Spinning speed, there is a controversy on what speed I should spin the ring at. Its safer at slower speeds but I will get less frame rate. The POV is not spinning yet so I can not test it at different speeds.
I got these weird errors while attempting to compile a Netburner project with NbEclipse. The project included the HTTP driver as well as some custom generated pages. After bashing my head against the wall for an hour I figure out what caused these errors. Because the HTTP driver has been loaded you are required to have a file in the /http/index.html folder to get NbEclipse to compile.
C:\dev\nburn\lib\NetBurner.a(httpinternal.o): In function `BaseDoHead’:
C:\release\nburn\system/httpinternal.cpp:431: undefined reference to `default_page’
C:\dev\nburn\lib\NetBurner.a(httpinternal.o): In function `BaseDoGet’:
C:\release\nburn\system/httpinternal.cpp:395: undefined reference to `default_page’
C:\dev\nburn\lib\NetBurner.a(htmldecomp.o): In function `SendHtml(int, char*, int, char const*)’:
C:\release\nburn\system/htmldecomp.cpp:176: undefined reference to `html_table’
C:\release\nburn\system/htmldecomp.cpp:180: undefined reference to `html_table’
C:\dev\nburn\lib\NetBurner.a(htmldecomp.o): In function `GetRecordFromName(char*)’:
C:\release\nburn\system/htmldecomp.cpp:88: undefined reference to `file_record’
C:\dev\nburn\lib\NetBurner.a(htmldecomp.o): In function `SendHtml(int, char*, int, char const*)’:
C:\release\nburn\system/htmldecomp.cpp:143: undefined reference to `DoHtmlFunction(int, char const*, unsigned short)’
C:\dev\nburn\lib\NetBurner.a(htmldecomp.o): In function `GetRecordFromName(char*)’:
C:\release\nburn\system/htmldecomp.cpp:86: undefined reference to `n_file_record’
C:\release\nburn\system/htmldecomp.cpp:95: undefined reference to `file_record’
C:\release\nburn\system/htmldecomp.cpp:86: undefined reference to `file_record’
C:\release\nburn\system/htmldecomp.cpp:95: undefined reference to `file_record’
More testing of the RGB LEDs. This set up uses 2pin screw terminals between each of the OctoBrite CYANEA. This set up works much better then the previous test.
I been travailing a lot the past few weeks making it hard to work on the mechanics of this project and with only 50 days left till Maker Faire I am running short on time. Travailing doesn’t prevent me from working on the software side of things though but it make it hard to test it. I have created this testing board to use while I finish the mechanics parts.
I created a base class (CFrameBase) for the frame animation then inherited from it for the algorithmic generated animations (test patterns, ect…). I also created a frame buffer class that inherited from CFrameBase that is basically just a big memory block that you can load via the serial port.
Next I added a blue tooth module so I could load the configurations over wireless via serial. Currently the serial can’t keep up with the data rate needed to do even modest frame fate. I will probably end up loading the entire animation in to memory before I start to display it. The draw back of this is that I will only be able to show 19 frames of animation at any given time.
Next I added a Triple Axis Accelerometer (ADXL345) and a Gyro (LPR530AL Dual 300), these will help determining the position of the ring as it spins around its axis. I have tested each of these parts on its own but without the mechanics I can do real world tests. I’m adding in software to manually correct for the motors speed as well just in case.
This source code is still a work in progress but functionally its working.
I have been busy the last few weeks as projects ramp up and we get closer to Maker Faire and the weather gets nicer around Vancouver. Its getting harder to makes these status updates as I would rather be making something or going outside and enjoying the sunshine and now Appa (A black kitty) won’t let me use my laptop without jumping up on my lap.
Giant Spin art box. I built this a few weeks ago and it work great but after playing around with it for a few hours the motor burned out. I guess that’s what I get for buying it from a scrap yard. Currently searching for another motor for this project. I will probably get it from an old washer and dryer or something.
I went as Mario to the Maker Faire per party fundraiser. I had grand plans on making a cardboard EL wire samurai suit but after 4 hours of working on the chest plate I knew I could never finish it in time. I had a flight in the morning so I couldn’t stay very long and missed the tail end of the party.
Picture by Peter Holmes.
Just got back from a trip to Washington DC. While I was there I spent a few days at the Smithsonian Institution and was blown away by the size of it. I spent a good 4 hours at the National Air and Space Museum where I met and talked with the creators of the TJ3Sat project. They built a homemade satellite using a Arduino for its main processor. They have a full write up on their website about their project. Very interesting. While in Washington I also stopped by HacDC the WashingtonDC hacker space as in every hackspace I have visited lots of interesting people doing interesting things.
As the weather gets nicer I been doing more Wheat pasting around the city, mostly in East Vancouver. I learned a lot from my previous attempts and this time its been going a lot smother. Still a lot to learn about speed and where to put them so they don’t get destroyed by weather or shop owners.
RGB POV LED Globe. This project can consumed most of my free time over the last few weeks and I will do a full project update in another post. Here are some pictures.
I wanted to create a very simple project for Maker Faire Vancouver that I could do in a weekend and something that people could take home with them as a memento.
I read about the Giant spin art box in a Make magazine Volume 25. It seemed simple enough and I could spray paint the Maker Faire logo with a stencil.
The build was pretty simple. I found a AC washing machine motor at a local junk yard that had a max speed of 1700 RPM. I created a mount for the motor to sit in so that it wouldn’t shake around too much then attached a flat board to the top of the motor with some screws. The whole process took about 3 hours.
There are a lot of different POV projects out there that have influenced my project.
n00tron 3D volumetric spherical display – interactive exhibit
This is an interesting project. Instead of putting the LEDs on the outside of the ring they put them on the spokes of a bicycle rim. Then they spin the bicycle rim on 2 axes simultaneously creating a 3-dimensional Volumetric display. The display is pretty random but still really interesting to look at. What I like about this project is that they used a bicycle rim as their spinning center ring. I’m having problems sourcing a 6foot ring to use in the center of my project and if I end up scaling down the project I probably will use a bicycle rim like they do.
POV Globe I really like the mechanics of their set up, and they go thou the project set by set explaining all the problems that they had in the build as well.
Illuminatrix
I also wanted a way for people to build animations for this project beforehand. In Burningman 2010 I saw this project Illuminatrix where they had a 16×16 pixel grid of RGB LEDs that people could program animations with their editor. The editor was written in JavaScript and hosted on their project’s website for anyone to submit animations.
After I started this project I contacted Ben from the Illuminatrix project and asked him if I could use his idea for my own project. We sent a few very helpful emails back and forth and he agreed to help me build the animation editor. Woot! I’m so excited about this and I think it will add a lot to my project.
I also like things to be interactive, I want people to be able to play and change them I make while it is in motion. I have been looking at different steampunk and cyber punk interfaces for ideas on the visual aesthetic of a control panel. I also been looking at different DMX lighting controller to get an ideas on different settings that I could make available to the user to play with. Brass plates, big nobes, buttons, switches, sliders, gates, and other tactile feed back devices.
A paranoid inventor goes to any lengths to protect his new time machine. When things go catastrophically wrong, he is forced to use the machine in a desperate bid to put things right. A ridiculous sci-fi conundrum.
I spent most of the last week doing research and ordering parts samples and testing equipment, trying to get everything set up for this project.
Things I have ordered.
$75 USD – 100x Superflux “Piranha” RGB LEDs common-anode ($50 for the LEDs + 25 for shipping). From evilmadscience.com. Definitely not the cheapest place for RGB LEDs but they had the “Piranha” style ones that I wanted in a common-anode and had good reputation online. I selected the Piranha style LEDs after getting a great answer to my question. What makes a good RGB LED.
$0 – Sample chips from Texas instruments. World class sales team, when I called them to talk about the particulars of their chips they offered to send me a few samples. 12hr later the chips where in my hand. Thank you Texas instruments. They also explained to me what a sink driver is and how it is used on these chips use.
$33 USD – 1x, OctoBrite DEFILIPPI ($20 for the board and $13 for shipping). Got one for testing before I buy the rest of the boards. This is one of the most expensive parts of this project and if it doesn’t work right it will be the most costly mistake.
$40 CND – ~1750 RPM AC motor with 1/4 shaft. Found a junkyard and priced by its weight. Tested and it works great comes with a pulley as well. Going to do more extensive testing this weekend.
Things I still need to get
A slip ring, I want to house the power supply in the base below the spinning right. But I need a way to transfer the power up the pole to the electronics and LEDs. You do this with a slip ring. I have contacted a few different distributer in Vancouver asking what they suggest. Waiting on a response. If all else fails I can alway make my own.
Flange Mount with Bearing. I need a way to mount the main pole in to the box. I’m looking for a rather large one as the pole will have quite a lot of stress on it.
A few pulleys. For the belt between the motor and the main pole.
The main pole. I am looking for a hollow 6 foot pole with heavy walls. I plan to run wires up the center of the pole and distribute them across the spinning ring.
A wooden box to house the motor, power supply and other electronics below the spinning ring.
Micro-controller, Still deciding what one to use.
Cables and connectors, I will decide this after I do testing on the OctoBrite boards.
I started searching for a LED to use in this project.
Normally I would go to my local electronics supply story, but this project requires lots (~64) of LEDs and my local electronics store charges a premium for the LEDs, (@1.50 each) and I know I can get them for cheaper then that.
What type of LED
RGB LED Thou hole, I want to use the thou hole LEDs as they are a lot more durable then the SMT versions and last longer.
Wide angle. I would like a ~100 degree viewing angle (its measured from the center of the LED, so it more like a 200 viewing angle) but these types of LEDs are hard to come by so I might have to settle with 60 degree viewing angle.
30 mA or less at full brightness. The TLC5947 shift regulator chip that I am using supports a max of 30 mA and i am hoping to get a LED that uses less then that to prevent over heating.
I trust adafruit and sparkfun and have had good experiences from both of them in the past but I am worried that there is a such a huge range in price difference between each of these sites that I might be missing something critical. Again I asked more questions about the differences between different LEDs and what makes a good RGB LED.
I have ordered a few samples from different sites that supported free shipping, while I wait on a response to my questions.
It has been since early-February where we’ve had Super Happy Hacker House and it’ll be more exciting than ever! Come down to Vancouver Hack Space this upcoming Saturday, March 19th at 7 PM and join in on some good fun! Not a member? We simply ask for a donation small or large! Are you a member? What excuse do you have for not coming?!? Bring a project or perhaps assist someone on one! Big or small, it is probably cool! Just want to come hang out and talk? Sure! Want to bring some beer? We always love beer! If you want to present something, we even allocate some time to do two-minute lightning talks! VHS is located at 45 West Hastings Street in Vancouver, just a few blocks northeast of Stadium-Chinatown Station with the entrance in the rear. If you run into problems getting in, we can be reached at (778) 785-5982. Someone should be there to assist you!
Come on down!
I have decided to create a 64x RGBLEDPOV Globe for Maker faire and Buring man this year. This project is a factor more complex then any other project that I have attempted before, and quite a bit more expensive.
But my project has more LEDs, it will play almost full motion video instead of a static image, and the globe will be a quite a bit larger, approximatively 1.5 – 2 meters in diameter, big enough to put a human in the center of the globe.
Each of the 64x RGB LED will require 3x PWM pins for a total of 192 PWM pins. No Micro-controller has that many output pins, I know I have asked. So you need to either create a matrix of LEDs or use some shift registers (SIPO). I decided to go with the shift register as is much harder to power a matrix of LEDs properly compared the shift registers.
The speed at which I an do a a full refresh was going to be a problem as I needed to update 64x RGB LEDs (192 pins) using 12 BITs PWM 120 times a revolution, 192*12*120 = ~276480 bits per revolution. Basic old school animation frame rate is ~24 fames per sec but I would want it get the refresh rate as high as possible. ~24 fps * 276,480 bits = 6,635,520 bits per sec or 810k per sec. Thats a hell of a lot of raw data!
I found a few boards that used shift registers to expand the IO such as the MondoMatrix LEDMatrix board. But because the board functions on 115.2 kbps RS485 bus, I abandoned it as it would be way too slow to update the LEDs fast enough to do video.
After researching several other boards, I decided to make my own specifically designed for this project . Who knows if I make it right I might be able to sell it as a kit. I have never built a board from scratch before but its a skill I have been meaning to learn for a while.
The most commonly used and basic, shift register is the 74HC595. There are lots of tutorials and project using this chip. It has 8 output pins and can be connected in series. After asking a few questions on the Vancouver HackSpace (VHS) mailing list. I was pointed at the 595 Shift Ease – 74HC595 Shift Register Breakout Board. An ingenious little board that lets you combine shift registers together easily to create a long series of shift registers.
I started searching for other shift register that had more output pins and better power management. I found the TLC5947. It has 24x PWM 12bit outputs, better power management, and it has a neat thermal shutdown function (Automatic shutdown at over temperature by that turns off all output drivers). So it if ever gets too hot it turns it self off to protect it. Sounds like the perfect chip, except that it only comes as a surface mount chip (SMT). I never done a SMT before and i have been told that they can be troublesome. One more new skill to learn.
Now that I found a chip that I can use to operate 192 PWM 12bit output pins I need to build a board for each of these chips. But before stared down the ugly PCB Eagle path, I searched to see what other projects where using that chip. After a little googling I found the OctoBrite DEFILIPPI board. Almost exactly what I was about to make with a few exceptions. The LED connections use a header instead of a 4 pin polarized connectors with common cathode. I can easily work around that but I emailed the manufacture of the board to see if they would make a change to their board for me. Waiting on their responses but even if they can’t make the change I think I will still buy this board.
Next is how to power this project. 192 * ~30 mA = ~5.8 Amps at full brightness (white). Thats a hell of a lot of power and that doesn’t even include the micro-controller (+200 mA) and the motor (~1 Amp). It’s not going to be too much of a problem at MakerFaire where I can connect to the grid but when I am out in the desert at burning man its another story. I will need to power this project off batteries and its going to be troublesome to say the lest. 5.8 Amps is only at full brightness and I will probably code in a brightness limiter so it uses as little power as possible to conserve on battery life and temperature of the chip.
A 100 Amp/hr deep cycle car battery costs ~$250 CAN and this project could drain it in ~14hrs. So I am going to need at lest two of them and a way of charging them during the day. A ~1000 W solar power set up that is able to charge these batteries is going cost >$2k. A 1000 W generator new is about ~$950 USD to buy but I think I could rent it for a week for a quarter of that, or beg, borrow, or steal it for much less.
Next problem is the micro-controller has to be able to handle a buffer of 810k per sec. A typical Arduino Mega 2560 (thats the big one) has 8k of SRAM. I could use some very tricky programing methods to read the file only as it is needed but even still it would be extremely difficult if not impossible to get it working right. Ideally I would like to have a multi-threaded chip with at lest 4mb of SRAM to buffer as many of the 32k frames between refresh. Still working on this one.
Still lots to think about but at lest some of the parts are coming together.
Super Happy Crafter House is an open event intended for crafters of all kinds. Those interested in papercraft, origami, technology exploration, sewing, knitting, spinning, weaving, felting, crocheting — and beyond! We’re open to new ideas and craft projects and would love to see you there! Come have fun, create, learn, and meet new people. Bring your own projects and friends, and we will supply sewing machines, felting and knitting needles, soldering irons, mending kits, and wool.
Ray Gun Revival has been launched. We sent out 8k unsolicited email newsletter to anyone that had anything to do with the old site and only ~50 people followed the link and signed up for the newsletter. Showing me that spam just doesn’t work.
Starring:, Rod Bird as Aging Space Ace, Adam Kerby as Space Radio Announcer Direction, Cinematography, Editing, Story: Andy LeBlanc Script:Jordan Ellinger Producers: Steven Smethurst, Jordan Ellinger, Andy LeBlanc
Maker faire Vancouver 2011 was announced. Including the call to makers for projects. I got a few projects planned, mostly for kids. Its just easier to please kids. I am planning on making a giant spin art rig. Its easy, the parts aren’t that expensive and kids will love it. Build instructions appeared in Maker magazine (volume #25)
Still working on my stencils but currently I only have gold spay paint and it limits what I can do. Next week will be 2 layer and three layer stencils.
I moved the data sensor from inside the VHS where they NEVER turn off the lights to just outside the window. Because it captures the sunrise and sunset the data that it produces is a little more interesting.
Tag your life, Tag each day with keywords, over time I want to see if there is correlation between certain actions and results of these actions. For example do i have a trouble waking up in the morning if I eat pizza, or am I more likely eat sushi on Sundays over Mondays. It uses Facebook for user accounts so anyone can create an account easily. later I hope to add the ability to update your Facebook feed with the gets of the week. Feel free to try it out.
I have been keeping up with my Happiness survey. Since I have added the analog happiness selector I find that people report there own happiness more accurately. We currently have 33 uses, feel free to contact me if you would like to be part of the survey.
I went to the Yaletown Illuminate festival to see a fellow VHSer flyingoctopus.netproject. It was an interesting night but there was only 13 projects and I was done walking the street in about an hour. Took some interesting pictures.
A 
