Budget Propeller Robot: Merlin prototype on order!
Rayman
Posts: 14,851
Just came across the ArdBot last night...
http://www.budgetrobotics.com/item/ArdBot-Chassis-Kit-340
I've decided to make a low budget robot using that kit plus their wheels (I ordered the blue).
That kit is about $30 delivered.
I've got some $6 servos (digital, all metal) that I can modify for full rotation pretty easily.
If I drive it from a $25 Quickstart board (I think you can get this at RadioShack now, right?), that's $67 for just about everything...
(I'd like to make a custom board, but it's hard to imagine beating the Quickstart...)
Maybe it's slightly cheaper to hack into some mass produced product, but I think this approach has merits.
(BTW: I'm thinking about a $10 adapter board for the Quickstart to make it easier to connect servos...)
Anybody see a cheaper approach?
http://www.budgetrobotics.com/item/ArdBot-Chassis-Kit-340
I've decided to make a low budget robot using that kit plus their wheels (I ordered the blue).
That kit is about $30 delivered.
I've got some $6 servos (digital, all metal) that I can modify for full rotation pretty easily.
If I drive it from a $25 Quickstart board (I think you can get this at RadioShack now, right?), that's $67 for just about everything...
(I'd like to make a custom board, but it's hard to imagine beating the Quickstart...)
Maybe it's slightly cheaper to hack into some mass produced product, but I think this approach has merits.
(BTW: I'm thinking about a $10 adapter board for the Quickstart to make it easier to connect servos...)
Anybody see a cheaper approach?
Comments
I like the servo/sensor board. It would make it very easy to plug servos and 3-pin sensors onto the QuickStart.
I've read digital servos aren't the best to modify to continuous rotation. I think it an issue of speed control. Digital provide full speed if the target position isn't met while analog servos vary the speed proportionally to the distance from the target position.
I have some of each. I'll try them for myself.
A buget robot is an important subject to me. I'm now a Cubmaster. I'd love to be able to find a robot cheap enough that each Cub Scout could afford to build one.
Duane
I got lazy looking for the Velcro....but have all the other parts!
I've got some h-bridge chips for servos that I seem to recall only being ~$2. Can always just build a bridge with mosfets too...
My only concern is that I wanted to add a rotatable platform with a camera on top, not sure this robot is quite big enough.
I think I'll buy this one and then decide between the two when I have them in my hands...
But, it needs 4 Prop pins instead of 1 for a modified servo.
So, I guess that'd be 8 Prop pins for the drive train.
And actually, instead of doing the usual servo "modification" for continuous rotation servos, I think I'll just rip out the control board and run the wires directly to the motor.
That way, this little add-on board I'm thinking about for Quickstart could work for both robots...
Using the 4 Prop pins also gives you the option to use "engine braking" or "coasting", which you don't have with continous rotation servos....
BTW: I got my ArdBot parts in the mail and it looks quite nice actually...
http://www.hobbyking.com/hobbyking/store/__11442__FPV_Fiberglass_Pan_Tilt_Camera_Mount_.html
and
http://www.hobbyking.com/hobbyking/store/__12875__FPV_Fiberglass_Pan_Tilt_Camera_Mount_L_Size_.html
I wonder how HobbyKing's little color camera compares with the DealExtreme camera? At $17, it's more expensive.
Duane
I'm going to take a look at how cheap I can make a Prop-Platform compatible board with the motor drivers on it.
I know I can't beat $25. But, the motor driver add-in board for Quickstart would likely cost $15.
Perhaps I could make a board to sell for $40. I think I can save money by starting from my PTP1 board design and replacing touchscreen section with motor drivers.
This would also have the advantage of allowing the addition of Prop-Platform shields.
My other idea is to try to drive the motors from an I2C i/o expander instead of Prop Pins. Not completely sure that would be fast enough, but it might be... That would save 8 pins that could be used for the shields...
I guess that's OK, considering the price. Took a look at importing the thing myself, but the Sparkfun markup looks very reasonable.
I'm back to thinking of Quickstart as the base board again... Just need a plug-in module to drive the motors...
I'll probably skip the I2C motor control and just give up 8 Prop pins. It's probably better to keep it simple.
I'm assuming that Sparkfun will get this back in stock. They're page says "Backorder Allowed", so I think they will...
Here's what the Magician looks like with the Quickstart on top:
I looked into this idea last year in an attempt to build the very lowest cost robot I could think of to let kids make some form of robot for the absolute minimum cost.
While the cheap motors, servos, etc may be the best idea, I kept thinking about some of the designs I had done that used parts of the circuit board for 3D construction. (Placing an array of light sensors above the board for position detection). On a small scale, the fiberglass of a circuit board, and solder joints locking things together can make a pretty strong framework. If the object of the game is education... then making part of the circuit board into a motor is a great idea for teaching how such things work.
Coils, wound on a framework provided by the circuit board, combined with driver transistors to build a simple stepper motor, would let you build the motors into the design of the circuit board. (It wouldn't be very efficient, but it SHOULD work.) Magnets, glued in position to 2 DVD's that could come with the kit, would be the drive motors AND wheels. AND a demonstration on multi-purposing. Tip121 transistors sell for about a quarter each.. (2 dollars for the motor drive ) Magnets: 6 - 8 per wheel needed at about 50 cents each, let's say 8 dollars, that's ten dollars for left and right drive wheels, along with the theory AND practice of building an electric motor.
It's hard to beat the price of one of the Prop development boards... but if someone (Parallax for instance) wanted to produce this board in the same way as the other boards... ( or design the robot frame to easily adapt to one of the current boards ) it might be possible to have a working, educational robot for under $50.00.
A kit like this would let budding robot engineers play with other control/motion projects like... clocks... moving lights, etc.
Might be worth looking into.
KB
The Quickstart is $25.
The Magician chassis is $15.
I'm looking at $15 for this plug-in to Magician (tentatively called Merlin).
So, that's $55 total (if I did the math right).
Where people can get creative is adding on low cost sensors...
For added fun you could face the quickstart towards the ground, solder 3 or more LDRs onto some of the button pads, and sequence the 8 leds to help it follow high contrast lines. Or just for a cool scanning effect.
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page73
post 1459
How set are you on using the QuickStart as the controller board.
As I mentioned, I'm also interested in make an inexpensive robot kit. I can't decide if it's better to use a QuickStart or a USB Propeller Protoboard. The USB Prop Proto is also $25 if you buy four at a time. It provides a regulated 5V source as well as the 3.3V source. The USB Prop Proto's regulator are much beefier than the QS's regulator.
I've mounted a breadboard to the backside of a Prop Protoboard. I brought all the pins out to female headers to two sides of the breadboard. A third side of the breadboard has a female header with power and ground connections.
Here's a picture.
Soldering wires from the pins to the headers took more time than I'd want to spend if I had to make a bunch of these. It might be a good first project for someone just learning to solder.
I cut a larger breadboard in half to use with the Prop Protoboard. I have since seen this size of breadboard for sale. I think the $6 is worth it not to half to cut a breadboard.
I think the Propeller Protoboards have a lot going for them as a beginner's robot board.
I also like the QuickStart boards. It's great to be able to power the board from the USB connection while programming. I also like the LEDs on the QS board. The female header makes connecting wires easy.
I think I favor the USB Propeller Protoboard over the QuickStart board to use with a budget robot. What makes the QuickStart your board of choice for these robot?
Are you planning on having others buy the materials or would you purchase the compents and sell them as a kit?
It seems like a lot of money could be saved on shipping if one person purchased all the parts and then sold them as a kit.
Next week I'm going to show a bunch of Cubscouts how to wire up a LED circuit on a breadboard. I'm hoping I can come up with a way they could all make a robot in the near future.
And guess what, the wheels turn! And nothing blew up - not the laptop usb port, not the regulator on the quickstart, not the propeller itself. The prop didn't reset. The quickstart regulator hardly gets warm. However I don't dare try stall current while drawing usb power from my laptop.
Tested here is 8 prop pins in parallel forming each leg of an h-bridge. With 2 motors you could have 3 half bridges, using 24 pins. However I would only drive one motor at a time, because their no load current (incl gearbox) appears to be ~170-200mA at 3 volts. When being driven by the prop they will draw less than that due to the internal fet resistance.
*** don't try this with anything you would miss should it fail! ***
tubular
You're powering the motor directly from the Prop? Ouch.
I thought a read about limiting the current to groups of pins. I wonder if it would be better to spread out the pins used a bit so the current wasn't all coming from the same set of eight pins?
Edit: I just read your post about the Prop's driving fets (good stuff). I still think it might help to spread the load over different sets of pins.
I need to better understand the short circuit condition, because something doesn't quite add up with only being able to get 42mA out. And of course the situation with 8 high side switches on and only 1 low side active would be very bad indeed as all that current would flow through the single low side switch.
But I'm prepared to sacrifice a prop (maybe not a quickstart) investigating how long it runs for. Perhaps we could run a sweepstake.
build this bot 2 years ago with a group of kids.(10 /12 years old)
Parts:
Gadget gangster propeller platform
Tamiya twin Dc motors (like in Tubular note )
SN754410 (H bridge)
220 , 22 and 0,1 microF condos
2 LEDs
plastic sheet
IR receiver + TV remote
It is fun to make competitions through a course with red and green round boxes
Jean Paul
Rather than hijack Rayman's thread any further, how about I investigate the stall current in detail, and start a new thread