Step 5: Program the Arduino

This is pretty easy and the fellas who I've linked to worked out most of the details. Good thing, we can get straight to playing around with it.

If you need information about arduino's, just google it. I found all of my info there easily. Here is the main arduino page: http://www.arduino.cc/

First thing, get the servo working. Run the servo example code in the arduino programmer to get your feet wet.

Second, download Stephen Hobley's Wii camera library and example code. Here: http://www.stephenhobley.com/arduino/PVision.zip

Connect the arduino to the same pins that he used. Make sure your i2c circuit and clock are all connected and the power is on and then run the code. You may need to insure that the Pvision.h file is in your arduino library so that is it properly reference when compiling.

Once running, in the serial I/O stream in the arduino programmer you should see data coming in as long as the camera sees some IR light.

I first used a match to check the camera. It worked fine. Later I used an IR led since I needed something a little more steady (and less burny.) You can use a digital camera to see if your IR LED's are working. The camera image sensor picks up the IR light that your eyes cannot see. This is a reassuring way to truly know if the LED's are working.
<p>Hiya, great project!</p><p>It's my first time working with arduino, and I was wondering if I could just use the 3.3v output from the arduino to the camera as I only have a 9v supply</p>
<p>Would be usefull for a sentry gun project. :D</p>
<p>I was thinking the same thing. With a laser designator! </p>
<p>hello</p><p>i have a problem. I have build the circuit but the camera gives only an x and y coordinate of 1023. </p><p>can anyone help me?</p>
<p>Why is the inverter needed?</p>
<p>I realize this is way after you posted this, but just in case anyone is wondering:</p><p>The camera has some onboard processing that requires a clock signal. The inverter, crystal, two capacitors, and resistor are a simple oscillator circuit (called a Pierce Oscillator) that generate that clock signal. There's some decent material online about these circuits, but this application note from TI might be a good start! http://www.ti.com/lit/an/szza043/szza043.pdf</p>
<p>Hi, how can I get this 8 pin pixart camera? I bought one wii remote and the camera in that was 12 pin. So can you please help me to get the exact one?</p>
<p>Hi, I actually followed your post and I was able to make it. There are things I changed though such as the I2C bus voltage converter shield. LTC4301L is not available in our region so I managed to make a voltage converter using mosfets. I tried using a 25Mhz oscillator, run the program but strangely enough it can control the servo motor but does not output the x, y coordinates. I always get zero values. So, I tried using a 16Mhz oscillator with 4800 baud rate, surprisingly though, everything went smoothly. I can control the servo motor and acquire the right coordinates. But the problem I am experiencing right now is that it only works for a limited time, after only 30 seconds, i can no longer control the servo as well as output the right coordinate. I have to reset it again to make it work. Does it involve running out of memory?</p>
<p>I just removed the camera from one of mine WiiMotes to try this.</p><p>Unfortunately I mangled some of the copper leads on the camera. they appear to be kinda 'spring locking' pins that were designed to come out [of the camera package] easily. Are these a standard pin that I might be able to buy replacements for? or would pushing solid core wire into their slots work? any other ideas on how i might salvage this?</p>
Sorry, I don't know. I was just really careful with mine and was lucky none of them broke.
You can get a tri wing screwdriver for those &quot;crazy three pronged screws&quot; at <a href="http://nintendoscrewdriver.com" rel="nofollow">http://nintendoscrewdriver.com</a> They are pretty cheap and they ship for free from the US so it won't take a month to get to you from Hong Kong.
Great work!! you've done a great job explaining details...thanks for that :) <br>i have one question though,,how is the IR camera calibrated?or what is the x,y reference?the origin (0,0) point <br>Thanks again, <br>Aws
<div> I bought Wiimote at the <a href="http://www.miniinthebox.com/" rel="nofollow">MiniInTheBox</a>, disassembled it, but camera inside is <strong>different</strong>:<br> <br> <br> <br> Who know something about this? It has the same interface or it's completely different?</div>
<a href="http://clip2net.com/s/29AGG" rel="nofollow">Photo</a>
You can now buy the camera's pre soldered onto a clock circuit from this place: <br> <br>http://rocketbrandstudios.com/store/pcb-s-and-boards <br> <br>He's even got some video and code examples for using it in a followme setup on a bot. <br> <br>I'm not sure if you would still need to run it through a converter circuit to use it with a 5v board like the arduino but it looks to be running fine interfacing to an arduino nano in the video <br> <br>Either way it makes for a much neater little package than unsoldering and proto boarding your own :-)
I'm wondering if a 3.3v to 5v logic converter is necessary. Some folks have had luck with disabling the internal pull-up resistors used in wire library and installing their own pull-up resistors to 3.3v source. <br><br>http://www.varesano.net/blog/fabio/how-disable-internal-arduino-atmega-pullups-sda-and-scl-i2c-bus<br><br>No converter necessary.
I imagine that the IR camera is a custom made component made only for Nintendo. Does anybody know if there is an equivalent product that can be purchased from an third party company be it Pixart or anyone else?
Hi. You know which is the range of IR that the WiiMote detect?
I think that the IR LED's that I was using were in the 940-950 nm range. I imagine that the camera is sensitive to a range around that.
OK, so i'm new to the use of inverting logic gates and i don't understand the schematic. After reading wikipedia i see that the logic gate has 6 inverters in it. Does it matter which you two you use? And as far as i can tell from the photos, your using using the VDD (pin1) and NC (pin16) but what are they connected to?<br /> <br /> Also, would you be able to make a list of recommended digikey parts for the standard parts? I'm not quite familiar with the logic gates and don't know which to pick. I'm also not sure about the crystal either.<br />
&nbsp;It has been a while since I made this, so I had to scratch my brain. <br /> <br /> I used the 74AC04 inverter chip. Pin 1 is not Vcc. It is weird like that. here is the data sheet:<br /> <a href="http://www.datasheetcatalog.org/datasheet/nationalsemiconductor/DS009913.PDF" rel="nofollow">www.datasheetcatalog.org/datasheet/nationalsemiconductor/DS009913.PDF<br /> </a><br /> Pin 7 is ground and Pin 14 is Vcc. The gates are all the same, use any two.<br /> <br /> As far as digikey is concerned. You'll probably want a 14 pin DIP package. We are running the chip at 3.3V, so we need to check that it'll work with that, but it looks like they all do. Otherwise, you'll want to buy something that has a minimum quantity of 1 and is available. I see 4 options with that, they are all good, buy the cheapest. Get a few (they are cheap) in case you blow one out.<br /> <br /> The crystal is easy too. Filter with 'through hole' and 25 Mhz. After that you'll get a bunch. I forget which one I got, but they are pretty much all the same. We are not too concerned with the stability and tolerance. Sometimes I like to get a few different ones in case. It is cheaper to buy a few than to ship stuff again.<br /> <br /> The caps can be annoying to buy because there are a million of them. Again, get a through hole with the right capacitance. We are not dealing with high voltage, so they will all work. Just find the cheapest then. Maybe get a few different ones if you are not sure. That is what I do.<br /> <br /> Just get the cheapest 1/4 watt 1.0 Mohm resistor.&nbsp;<br /> <br /> <br />
Hi there. Is 74LS04 ok with this circuit? Thanks buddy. :)
I am not sure. A quick look at the 74LS04 datasheet shows that it runs at a minimum of 4.75 volts, but the logic will run at 3.4V so I have no idea. I would wire up the clock part and make sure it is working correctly before connecting it to the Wii IR Camera.
Hi there. That is a very interesting project you have. I am wondering if instead of using that voltage converter shield, can we use this instead: <br>http://www.sparkfun.com/products/8745<br>It has the same purpose though, which is to convert voltages.<br>Forgive my knowledge about this, since I am just a beginner as a hobbyist.<br>Thanks a lot buddy! :)
Yeah, I have looked at that too. I don't know for sure if it will work, but it says that it can do i2c conversion, so it seems like it will do. If it does it would be cheaper and easier to work with. Thanks for the comment!
Hi. Sorry for this newbie question but, is the theory behind the circuit of <br>http://www.sparkfun.com/products/8745<br>same as this?<br>http://i52.tinypic.com/2niprtz.png<br>I am trying to work your on your project and I followed all the circuits,<br>except for the voltage converter shield. It seems that the i2c scanner:<br>http://todbot.com/blog/2009/11/29/i2cscanner-pde-arduino-as-i2c-bus-scanner/<br>cannot detect the i2c device, which is the pixart camera.<br>I would really appreciate if you can help me on this. Thanks again. :)
Hi, I am having problems getting this circuit to work. I've tried replacing all the components, different wii IR cameras, and still I get a continuous output of all blobs with a size of 0. I lowered the pull-up resistance on the SDA-in and SCL-in to 1.5k ohms, and this worked for a bit, but then after a few days it stopped working again. I've also tried soldering all the components instead of using a breadboard, but this had no effect. Do you know what the problem might be? This is driving me nuts. My next step is to vary the pull-up resistor values and see if something works. Thanks for your time.
It is hard for me to say without working with your setup. I did burn out a camera at one point by accidentally cranking up the voltage on it. I have had no problems on a breadboard. If you have access to an oscilloscope, check the output pins to see if you are getting any action. This may help you isolate the problem. I am going to be working with this again soon. If I come across anything I will let you know.
Thanks for the reply. Turns out it was a grounding issue. I was only connecting pin 2 of the camera to ground, and when I connected pin 3 also, it worked. I don't quite understand it because the pins are connected internally... Also, one of the cameras I was trying to use was burnt out, so that just added to the frustration.
nice work. I was wondering whether similar circuitry could be used to detect finger movements? I am working on an optical projection keyboard project and was planning to use the Wiimote IR sensor for the purpose of detecting finger movements when a user hits a key. I think what you have made could be applied to some extent to my project also?Is that possible? Thanks
You may find it a lot easier if you were to do something like what is in this video for your project. Doesn't require such lengthy amounts of work, and you can keep the wiimote in one piece <object height="390" width="640"><param name="movie" value="http://www.youtube.com/v/0awjPUkBXOU&amp;hl=en_US&amp;feature=player_embedded&amp;version=3"><param name="allowFullScreen" value="true"><param name="allowScriptAccess" value="always"><embed allowfullscreen="true" allowscriptaccess="always" height="390" src="http://www.youtube.com/v/0awjPUkBXOU&hl=en_US&feature=player_embedded&version=3" type="application/x-shockwave-flash" width="640"></embed></object>
Brilliant instructable.<br><br>Do you know how to change the pins that the camera uses on the arduino? I think it defaults to 6 and 5, but I'd like to change these and can't figure out where they are set...
I don't know offhand, but I would assume that it is defined in the Pvision.h library. I haven't had to change it myself.
Very interesting project and also thanks - I have been curious about the 'speed' of processing. <br>I do animatronics as a hobby and one of the 'holy grails' is real time Lip Syncing where the mask will move it's lips in time with yours. <br> <br>If one thinks of the lips as a flexible circle with points every 90 degree's then it's easy to put reflective points on the lips at that location/ Then it's easy to work out how open, cloded, or wide the mouth is and du0plicate it on the puppet. <br> <br>Using an industrual face mask (for dusting) you can make a fully self contain system complete with camera and IR 'spotlight' that will protect the eyes and provice a dark environment. <br> <br>Take Care <br> <br>Marcwolf
Anyone can help with building a smaller controler so i can attach it laptop screen side and use a pen as a mouse?<br><br>the wiimote its just too big, any sugestion?<br>A way to power it by usb would came in handy too.<br>thanks
Does anyone know where you can buy the camera on it's own?<br><br>Also, thanks for the cool instructible - that servo is incredibly responsive!
Thanks, glad to help. The servo is very responsive. It is fun to just sit and watch it move.<br><br>I do not know where to find the camera on it's own. As far as I know this is the company that makes it,<br><br>http://www.pixart.com.tw/about.asp<br><br>If you find them I'd really like to know where you can get them alone.
&nbsp;Could you mount to the servo and cut movement back to a 1:1 movement ratio?
&nbsp;Absolutely. I've been playing with this method. It works better for tracking. You can also use two servos and track with both the X and Y axes.
Miniature Mech Fighting toys, here we come.
This is great. I&nbsp;wonder if you could use it to track the sun? There are cheaper and simpler ways to do that, I suppose, but the wiimote ir camera mounted in the middle of a group of solar panels, or a reflector dish for solar thermal, with the arduino programmed to return the array to a pre-determined position when an ambient light sensor determines it's dark outside (or the solar panel output goes below a threshold, or an internal clock determines it is close to sunset, or a combination). You could even alter the morning position based on the date, so the panels are in the right spot in the morning. <br /> <br /> Very flexible, I like.<br />
Yes, that would absolutely work. It would really be powerful in a moving platform where you might need some more dynamic sun tracking. <br /> <br /> You can take this idea further. If I know the time of day and my position on the earth (longitude, latitude and possibly altitude as well) then I know where the sun should be. Then, by using a IR camera (or several of of them), you can measure where the sun is. Then, working backwards, I know (within some error) what the angular orientation of the cameras and therefore the orientation of the robot, airplane, etc as well.<br /> <br /> This gives you a fixed orientation measurement that doesn't require integration as apposed to inertial navigation systems which builds up error over time.<br /> <br /> Of course that would only work during the day. I wonder what IR signal the moon has?<br />
I don't think the moon has much of an IR signal. Not enough for the wiimote camera, probably. It'd be easy to test, though. With cwiid or another wiimote program for pc. Point it at the sky and see what the output looks like. <br /> <br /> For orientation, that method would work great, except at around midday. Especially if your vehicle is on a tilt. That can be compensated for, though. But then again you can get hall effect compass systems for fairly cheap nowdays, which also have no accumulative errors or anything. <br /> <br /> Now if your vehicle was on a body without any magnetic field, like the moon or something, then that would work very well. Especially if you could use the earth as a second reference point for when the sun is directly overhead. <br />
I just tried this out, with wmgui on my laptop. The sun is much too bright for the wiimote to register it! Even with two polerised sunglasses lenses at various angles (crossed at 90 degrees you can't see through them, though the sun still shines through) I got no output - though reflections of the sun on various things produced a myriad of points the wiimote will track.<br /> <br /> I also tried the moon the other night, being a full moon, and I think that it wasn't bright enough either. Though I didn't try putting my sunglasses in front of the lense to test that out. <br /> <br /> <br />
maybe you could try a welding mask? If you have a spare set of oxy acetylene goggles those might work, arc welding mask are actually darker and might block to much light. They might completely block the IR light though<br /> <br /> also does anybody know if the IR cameras from the Nyko or other brand wii remotes will work?<br />
Actually if your system was sensitive enough you could work out longitude and lattitude, too!<br />
Very interesting instructable and well written.<br /> <br /> I'm looking into taking this a bit further and was wondering if you set up an area where you have 1 camera on one side and another facing perpendicular from another side like so...<br /> _____________<br /> |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br /> |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br /> |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; *&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |&lt;<br /> |&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |<br /> |_____________|<br /> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; ^<br /> <br /> Where ^ is a camera direction and * is the light source in this square environment<br /> Would you be able to track 3 dimensions? (1 camera doing xy, the other doing yz). Would arduino be able to handle the two cameras? I also assume that each camera would need it's own interface circuit as well?<br />
Yeah, that is a great idea. You will need to keep in mind that the X-Y on the camera is an angular projection onto the camera surface, thus you will not get a square area (as seen from above, see drawings below) that the cameras are seeing, but instead some kind of quadrilateral. <br /> <br /> Another idea is to use two cameras like eyes. This is essentially the same idea as having them 90 degrees from each other. Either way you will need to do some trig to get the position.<br /> <br /> The arduino should be able to handle the two connections. Only a few pins are used. You may need to use two of the i2c chips, but you might be able to get away with one. I am not sure. It would be more straight forward just to use two.<br /> <br /> The real test would be the update speed. I don't know how much the arduino can handle, your 'refresh rate' may suffer.<br />
It is also important to remember that the camera reports the 4 brightest points. That is how the wii does position detection. The sensor bar is two IR lights positioned at a fixed distance apart. Their distance apart via the IR camera determines their actual distance away (some trig required). You could do other neat tricks with more lights or cameras to get all sorts of information.<br />
&nbsp;Is there anywhere we could get the IR camera separately?

About This Instructable


209 favorites


More by Eric Jacob: Make Your Face Guitar String Hot Wire Cutter Viewing Myself in 3rd Person
Add instructable to: