For a long time I wanted to make a robot, but my main problem is finding parts for it.

I want to make robot that is basically a computer on wheels. To add sensors, and for a start just to make it go from place to place, and avoid obstacles. Later on to add an arm and pick up things, etc...

Anyways, for a start I need a chassis; I really like the ones here: http://www.pololu.com/catalog/category/88 (the ones on top with 6 wheels), but they look too small to carry a computer and batteries to power that computer, plus sensors arm, etc...

I tried to search on the net for a bigger chassis, but everything I find is for small simple robots...

So does anyone know where I can get a similar chassis but bigger to able to carry the weight of a computer + batteries?

 

Know anyone with a go-kart that's got a bad engine?

 

Know anyone with a go-kart that's got a bad engine?

Haha... that would be going to the other extreme (i.e. too big), I would need something in the middle in terms of size.

 

Give some dimensions if you can.

 

Probably the same size as a mid tower case, so about 15" by 14"

Also, I plan to add layers/floors. So probably the lower layer will have the motors, and batteries, next layer will have the motherboard + DC-DC power supply (I have one of these, I don't remember if it's that exact model), then next layer probably I'll have the sensors, etc...

My only concern is big enough to fit the motherboard (probably 12" by 10"), and the motors powerful enough to push the whole thing...

 

Try googling combat robotics parts (or battlebot parts). 5 years ago there were some really high quality stores online.

 

I googled, but I still did not find anything suitable for what I want (that is already made).

It is difficult for me to build the base on my own, I don't have the tools for cutting, bending, etc... that's why I wanted to buy an already made base (chassis). I can do the electronics & programming, the mechanical part is the difficult one for me...

 

You could always look for places around you that would build one for you. Just a thought.

 

I googled, but I still did not find anything suitable for what I want (that is already made).

It is difficult for me to build the base on my own, I don't have the tools for cutting, bending, etc... that's why I wanted to buy an already made base (chassis). I can do the electronics & programming, the mechanical part is the difficult one for me...

If you can get ahold of some square tubing, angle brackets, and whatever youre wanting to use for the layers (plywood, metal plates, ect) all you would need to build something is a drill with some good bits, some nuts/bolts/washers, and a saw to build something sturdy as a rock. If youre worried about bolts coming undone or such I'm sure you could get a friend to weld it for you (assuming you have a friend with a welder, country kid here). Depending on what you need it for you might be able to use wood for everything.

As for the motors, you might look into the drive motors off of the power-wheels toys like the little hummers kids drive around, one of them was enough to drive me around with no problems.

[EDIT]
These guys probably have a thing or two that might work
http://www.robotmarketplace.com/products/motors_main.html
[EDIT EDIT]
Or maybe the motor out of a handheld drill?

 

You could always look for places around you that would build one for you. Just a thought.

Yes, I did start to look. I am planning to go this week to visit one of the places I found on-line, hopefully they are have what I need. I just sent them an e-mail and asked for the hours they are open, and what kind on help they can provide.

But in the mean time...

magnet18, I have no idea what motors to get. I googled, and I found this one EMG49 kit. It says up to 30kg which is more than enough, but I need 2 of those (for 4 wheel drive), and also, I want to add a motor to turn the wheels (no idea what to use for that), so that the robot can turn left/right, and I would prefer one for each wheel.

The problem with that drive system (EMG49), is that the controller has a limited instruction set. It would have been nice if they would publish the code for the PIC18F2321 that is driving it, so I can modify it to my needs.

And another problem with it... it is f-ing expensive!!! A Canadian store sells the parts for one drive for $404, and there is 13% TAX on top of that, that would mean $913 for 2 drives !! (shipping not included)

 

What type of terrain are you expecting to encounter? If this is just for indoor use, using casters in place of two of those wheels would allow for a skid-steer setup, just turn one motor faster than the other or in the opposite direction to turn. You can turn on a dime this way, it makes things simpler, and it makes things cheaper. That kit you're looking at looks like its quite expensive, are you planning on controlling this via programs directly on the computer or by interfacing with an existing unit? If you manage to get a couple old power-wheel motors (search Ebay) I can guarantee that they will work with a 12V lead acid battery, and you can control them with the maxxtronic 30A speed controller module.
This uses a potentiometer as the user interface to control the speed, but it could be controlled by a servo attached to the pot, or possibly something more directly controlled by the computer. It only goes one direction, but a relay should be able to switch the polarity to the motor simply enough.
The robotmarketplace website i linked to earlier also sells 2 motor mixer boards, capable of either standard controls (one channel forward/back, one left/right), or the skid-steer setup, (one channel for each motor). These are controlled the same way a servo would be.

 

Have you been here http://www.societyofrobots.com/ there are many vendors who advertise kits on here. I started working on building one last week. I have decided to use stepper motors for the drive. I put the stepper functions in a 22v10 so the micro only has to send dir, mode, and step (mode can be fixed - but I coded it anyway). If you use DC motors on both wheels, you can steer by changing speed between the 2 wheels; more turns on the left wheel and you turn right, more on the right and you turn left. Or you can just use one motor and differential on 2 wheels and steer the other 2 using either servo or stepper. With DC motors, you will have to monitor rotation to sync 2 motors. I don't know how much trouble that is, may be over thinking the whole process, but on that site there are several people willing to provide advice to you.

 

I want more precision out of the wheels, that's why I wanted to use motors with encoders, and that's why I wanted to turn the robot by turning the wheels instead of making one go faster than the other.

I will use the robot in house (on carpet) at the beginning, but later on I also want to take it outside.

What I want it to do (to start with) is basically navigation and obstacle avoidance. So if I tell it, "Go there", to calculate how to go there, and also to go around obstacles, and all this will be done on the computer.

I am also considering using stepper motors for propulsion and turning, because then I can very precisely control and synchronize all motors at the same time.

The problem with that is that I don't know what strength/torque motors to use, and I don't know where to get the gears, metal brackets, etc. needed to attach everything together. And I prefer if everything was made of metal.

 

For many of the reasons you listed, this is why I am using stepper motors for my drive. Steppers can loose sync and not step also, though it is easier to determine/correct. One of the vendors listed at the SOR site has a good selection of gears and chains. None of it is give away and shaft couplers seem expensive to me, I might be making my own. You should note that even though we will monitor the wheel turns very precisely, the wheel can still slip on the ground (especially true outside where the ground is not flat). So if you want to determine your position, it might need more than just math involving wheel turn. As far as strength for motor torque, physics F=MA (A = dv/dt) tells us the force and moment arms help us calculate the torque back to the motor. Essentially what that means is that, less power is needed to accelerate very slowly. But you will still need to overcome the force required to push your mass A) forward and/or B) up an incline. And you will probably want some method of slowing the descent down an incline as well.

 

I have experience with stepper motors, check this: http://www.youtube.com/watch?v=m2CYttY_xPg

The thing I love about them is that the torque is highest at slow speed, so it's kind of like changing gears in a car, low gear means more power, higher gear means more speed and less power. Same with stepper motors, except you don't have to change gears to have this effect.

About this "the wheel can still slip on the ground", yes, but there is not much you can about it regardless on what kind of motors you are using. That has to be compensated by other means, like camera (software must be really advanced for that), or Ultrasonic Range Finder(s) (not precise, cheap), or even Scanning Lasers (very precise, very expensive).

At the beginning, I will just trust that the steppers don't miss too many steps... later on, I will add more sensors.

I did not make up my mind if I will use stepper motors, or just DC motors with encoders...

 

Well, home depot/lowes might get you started on the metal, I know they have plenty of square tubing and angle brackets, I don't know about thick sheets though. If you made the frame sturdy enough to support itself then bolting a thick (1/8") metal grid or mesh to this should provide enough support to mount things to, depending on what you mount 1/16" might work well also.
And if yore lucky you won't have to drill any holes

If you have experience with stepper motors and will be relying on dead reckoning (at least partially) they will probably be a better option. I was just giving advice based around things I've built before. Of course, everything I've built before has either had me riding on it, or been directly RC.

What I've found with things like this (at lease with directly controlled things like remote control) is that it's almost always cheaper to use things that kinda do what you want and modify them or just use things in ways they were never intended. Plumbing mounts hold motors great and 18" plastic floor tiles make great robot bases, for example.
You get good at re-purposing when you're just a couple high-school kids on a tight budget

 

That CNC ploter is very cool. Not the fastest thing, but very cool. Did you write the software or purchase it. I would really like to build one myself, I have seen some that use a dremel as the main cutter. Are you planning on some type of cutter? I noticed that your boards were home brew, how are you connecting them to the PC? I'm thinking about designing a USB to stepper 3 motor controller for low cost hobby projects. What are the specs of the motors you used? Sorry, don't mean to hijack the thread, but homebrew CNC is very nice. I'll PM you about it.

 

CVMichael:

Here's a thought, why not re-purpose a large scale (say 1/8) RC Crawler? The engineering is already done for you. The lipo batteries will be much smaller and lower weight than lead acids. If you could find one with a good metal chassis, an aluminum sheet platform could be mounted for the electronics.

It sure seems like a microcontroller could handle the robot brain duties, at least till more advanced vision or whatever is incorporated. Even then, that might be self contained in a camera system, or wifi router for example.

A random 1/8 Crawler, can you say Mars Surveyor, link..... pretty cool!
http://www.nitrorcx.com/51c875-maxstone8-green-24ghz-artr.html

 

A random 1/8 Crawler, can you say Mars Surveyor, link..... pretty cool!
http://www.nitrorcx.com/51c875-maxstone8-green-24ghz-artr.html

That thing was built to carry a piece of plastic on it, are you sure it can carry an entire computer plus big batteries to power the computer ?

And YES, it want a computer, not microcontroller, there is a big difference between 48MHz and 3GHz.

 

That thing was built to carry a piece of plastic on it, are you sure it can carry an entire computer plus big batteries to power the computer ?

And YES, it want a computer, not microcontroller, there is a big difference between 48MHz and 3GHz.

4wd, servo steering, ground clearance, running over obstacles instead avoiding them seemed kinda neat, and at a reasonable price. Don't know why the Crawler couldn't handle a little netbook and some lipo's, but since I don't have one, I can't say for positive. Your the project manager, so accept or discard ideas as you please, no worries here.

I've got an Invacare power wheelchair that could handle a giant payload out in the shed. It has two 300W motors and a tray for two 12V group-24 gel cell batteries, is that what you need?