Let's Make Robots!

Pictomatix MkII

Wanders around taking photos of interesting things

I've hacked a cheap camera and mounted on a pan / tilt neck assembly. Pictomatix MkII only has 64Mb of non upgradeable camera memory so unfortunately he can only take 25 hi-res pictures. I've now made him fully autonomous.

The first video is just a quick test to make sure the picaxe could control the camera.

 

 

 

 

 

 

 

The second video is his first time out. The wet concrete made for some nice spinouts. Unfortunately the camera turned itself off when a photo was not taken within 30 sec. Will have to change the code a bit.

 

 

 

 

 

 

 

 

 

 

The chassis is same as before.

Chassis_top__small_.jpg

This is inside of the camera. It is a very cheap and simple camera, this makes it about 1/4 of the weight of the Canon A580 power shot I've been using once it's batteries are removed. If you want to find a cheap camera to hack then I suggest looking for one that takes a memory card. This camera can only take 25 640x480 pictures (uncompressed) and now that the batteries are removed it looses the photos when the power is turned off. This means I have to download the pictures via the USB cable before turning it off.

Inside_1__small_.jpg

It normally runs on three AAA batteries or USB but wont take photos when running from the USB port. I've put a diode between my 5V supply and the camera to provide a 0.6V drop. This runs the camera at 4.4V which should be the same as if it was on alkaline batteries.

High_Voltage__small_.jpg

That large capacitor at the back with 330v written on it is used to store a charge for the flash. Always avoid touching this and it's surrounding circuitry otherwise you'll get a nasty zap and the discharge through you could also damage the low voltage circuitry of the camera. This will discharge very slowly over time. It's probably a good idea to leave the camera without batteries for a couple of days before opening it. Hopefully it will have discharged by then. Test it with a multimeter set for at least 200V before touching it.

Connections__small_.jpg

I've labeled all the important bits. Click on the image for a larger picture. It was fairly easy to test which connections are what. The power connections are directly from the battery holder. I didn't bother with the USB since the camera won't charge the flash from it. There are only two buttons on this camera, MODE and SHOOT. This made hacking it easy. By testing the button connections to ground I found that each button was grounded on one side. This made it an ideal candidate for NPN transistors to operate the buttons. Otherwise I would have used minature relays or analog switches.

Connections_2__small_.jpg

Here you can see my simple 4 wire connection. Brown is ground, Red is +4.4V, Yellow when shorted to ground changes the mode and Orange when shorted to ground takes a photo or changes options depending on the mode.

Re-assembled.jpg

I decided to run the cable out where the mode button use to be, as you can see, I can still press the mode button manually for testing. I was originally going to do a neater job but it's only a $15 AUD camera to experiment with and the button hole was perfect in both size and location.

Polymorph_servo_neck__small_.jpg

The Neck is made with polymorph, the bracket was made as shown in this tip/walkthrough. To join the servos together I just used a small blob that I heated in the microwave so that it was hot enough to bond directly with the servos plastic casing. It will be difficult to seperate the servos without damaging them but they are very cheap (about $4 US) so they can stay that way and I'll just buy more servos if necessary. I can always use this neck joint on another robot.

Here is the Schematic. Click on it for a larger image.

Pictomatic_Bot_MkII_VII_Schematic__small_.jpg

 I've exchanged the IR receiver for a Maxbotix EZ1 sonar. Now the robot can navigate by itself. The LDR is for it to determine if it needs the flash on or off.

 


Update: 14-2-09

 

He's taken his first photos. Unfortunately the grass in the garden caught in his gears so I was forced to put him in the driveway. He's not a very good photographer yet.

testrun1017.jpg

testrun1012.jpg

testrun1018.jpg

I think you get The idea so I won't bore you with shots of my knees or one blurry image that I don't recognise. This will be the end of that meccano chassis as it doesn't work as well as hoped. This was a fun "first camera hack" experiment but I think a better camera, chassis and processor are required. It's not easy to write a good autonomous photographer program with 256 bytes of memory. I had the neck tilting up and down a bit but didn't have enough memory for him to look about in a useful manner.

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
nice job, i have a fujifilm fine pix with a few broken buttons that is suposed to be hacked soon :)

just seen the video now, the little bot is very fast !!!

Just watched the new vid of Mk II - Fab.

Now you need to make a second Pictomatix so that it can film the first! Perhaps they could work together like wall racers.

Mike

@Odd,

Why are there three caps between +5v and Gnd on the left of the schematic? I guess it's something to do with different responses from the different values.

Mike

The 470uf electolytic cap is the main filter. It stores enough to handle reasonably big surges. The 0.1uf monolythic doesn't have a large capacity but can discharge quicker and is good for high frequency noise generated by the processor, servos etc. That's the yes part. All robots should have these for stability.

The no part is simply because the bread board and solid core wires pressed into it are not the best system for electrical noise. I put the 100uF close to the sensor power connections to ensure stability. It may not be required but doesn't hurt.  

I put the 100uF close to the sensor power connections means that you put it close to the voltage regulator?

Thanks

Mike

Because the the bread board has power rails on both sides. Iput it physically close to the power connections of the sensor. The further away from the regulator the sensor is (physically) the more resistance in the power connections, the more capacitance and even inductance. Many sensors will recommend a capacitor in close proximity to it's power connections if it is not near the power source. The L298N dual bridge IC also recommends capacitance on its logic supply as well as the motor supply.

Sometimes these extra capacitances are not required but it doesn't hurt. The one time recently when I forgot to add capacitance to a breadboard power supply, I could not program the picaxe. By adding 100uf to the output of the regulator, the program downloaded without a problem.

lol that thing really flies.
It's not RC, that video shows it using the sonar to navigate. I changed the design since you first looked at it. It's now fully autonomous.
It just missed hitting the wall a couple of times. Its really fast. Does it ever slam into the wall? Do the spin outs ever mess with the sonic sensor?