-
Notifications
You must be signed in to change notification settings - Fork 2
Microcontrollers
At 3.25" x 3.25", the Beagleboard XM for 150$
- WiFi attachment via USB should be like 12 bucks
- Has RS232 serial
- 2W of power @ 5V needed
- weights ~37 grams
BeagleBoard XM Accessories
BeagleBoard Getting Started Guide
BeagleBoard Reference Manual
Ubuntu 12.04 is the goal, on a 32GB microSD.
point of access: /dev/mmcblk0
registered ethernet mac address: 32:62:2d:b8:e0:8c
static IP address: 140.180.128.151
Username: Password: teammugatu
you will need to boot manually with the kernel on partition /dev/mmcblk0p1 and root=/dev/mmcblk0p2 passed as a kernel argument
On the big side (4.5" by 4"), but a total beast, the PandaBoard ES at 161$
- Has included WiFi, bluetooth
Raspberry Pi seems like it's super backed up from their normal distributors, but it looks like it's selling on Amazon for about double its price at 65$
-
Could get WiFi via USB, but then would be out of USB ports. Could have a USB hub to fix this
-
Arduino Boards
- Cheap, easy to program, comes in a lot of sizes, well documented
- Language used is Processing
Linux chip and use robot specific software?
- ROS -> C++ or Python or ?
- RobotBasic
It looks like the EduBot might be using something similar to this (Intel Atom): http://www.adl-usa.com/products/cpu/datapage.php?pid=ADLD25PC
Motion Profile Parameters for leg control(9): -Cycle time (Period of rotation) -Sweep angle (Slow, ground contact rotation) -Offset (Angle of attack) -Duty Cycle (Fraction of time in contact with ground)
PD Control Vars: -Leg angle -Leg velocity (Angular???)
Processing(17): -PC/104 form factor computer stack (mod version of PC/AT bus) -CPU is 300MHz, Pent 2 with 256 MB RAM, 256 MB hard disk (Compact-Flash) -VGA, Ethernet, RS-232 Connects -QNX Operating system compiled with GCC -802.11b wireless ethernet (Ran in traditional accesspoint and ad-hoc) -Analog/Digital interface is custom (We will not be doing this) -Quadrature decoder card gives 6 channels of 2^16 depth for angle sensing -Visual processing is done on separate stack
Sensors(18): -Relative angle measurements done of motor shaft, 2000 counts per rotation, 0.0054 deg resolution -Hall-sensors can give bit whether leg is in a particular position -Motor temp sensor for shutoff -Gyroscope gives roll, pitch, yaw ????
Software(20) -Custom C++ lib for low-level functions (RHexLib, developed by U of Mich) -SimSect++ is used with a gui and Operator Control Unit(Based on a Linux laptop)
Linux base station ->(Wireless)->CPU(QNX)-> Custom I/O card + Angle Decoders