There's a useful thread about running multiple instances of the Propeller Tool Software and the PST. http://forums.parallax.com/showthread.php?132744-multiple-PSTs
This function can attach to specific Propeller chips and act as virtual reporting screens, take the place of output LEDs, serving as LCD and TV monitor replacements, and other useful functions in a multi-chip Propeller machine.
Yes, 12Blocks now runs on Linux and Mac!
Here's a picture of a class using 12Blocks on Macs to program TBots:
What do you mean "programming multiple propellers"? As long as you just want to program one at a time you're fine: 12Blocks has a nice "port chooser"- that either automatically connects to a Propeller- or let's you manually specify which port to use. In demos I frequently show 12Blocks connecting to multiple bots/projects.
Hanno
Hanno: what about the mixing of languages? On a piece of Spin code, can it embed 12blocks code, or you will probably say it's just the other way around..
:In the case of the rotated picture - no problem with a Snow Leopard Mac. Just drag it to the desktop and turn it.
The Big Brain has a HYBRID interface under development for some time which includes various circuits: examples are Serial Communications, Daisy Chain, Token Ring Topology, Prop to Prop, and Parallel.
Definition
Reduction Interface RI - the simplification and combining of wire interfaces to achieve one interface that can handle the functions of several
HYBRID Interface - several interfaces mixed together, not necessarily reduced
This post introduces the Reduction Interface RI. The implementation of RI will reduce the number of prop wires by combining several interfaces together and mutualizing various functions.
This week sees the reassembly of the first two Partition arrays for 800 cogs after moving to the larger containment rack. A cog testing program is needed and suggested. Pseudo code below.
Loader ' loads code into 2 Partitions of 100 prop chips in Cog 0
Cog Loader ' loads code into all Cogs
Test Cog 0 through 7
Result to LEDn
Repeat
This makes a new smaller machine of 100 props with 800 cogs. Partitions were built at 50 props each, however since the introduction of the Filter, it is likely the number of props per Partition could double (perhaps a future consideration) and continue to work reliably.
Hi Humanoido,
12Blocks let's you "inline" pure spin with the "code" block. You can also type any valid expression or condition into appropriate parameters. If you want to reuse existing spin objects you can "file>import" them and 12Blocks will create blocks for your public functions. Do you have a usb connection to each of your props? Or are you loading one and telling it to copy to others?
Hanno
Hanno: It's good to know Spin code ports to 12blocks. I'm more interested in knowing if anyone has multiple props connected together with various routines already developed in 12blocks.
Correct me if I'm wrong Hanno, but I thought 12Blocks was designed to make programming the Prop a lot easier, especially for non programmers.
I haven't used 12Blocks but I'm guessing it's kind of like the interface to program Lego MindStorms. You arrange the graphical blocks to indicate what you want your robot to do and in what order.
I doubt it's the best programming environment for programming unusually behavior like multiple Propeller communication protocols.
I'd think programming directly in Spin and PASM would be easier than trying to bend 12Blocks into doing things for which it wasn't intended.
Hi Duane,
You're right, I initially designed 12Blocks to simplify programming the Propeller so my 5 year old daughter could do it. It takes away most of the hard things: syntax, grammar, and vocabulary to let users focus on the logic of what they're trying to do. It adds a bunch of features that I initially developed for ViewPort to simplify debugging- graphing variable values, seeing pin states, and changing program parameters while your program runs. While it started as a simple project- it's now very complete- even letting you create stand-alone "exe" files similar to the PropScope application- with bundled firmware, graphical gui and communication stack.
Think of 12Blocks as a hybrid- easy to use, visual, with nice integrated tools calling into low level drivers written in spin/pasm. As such, I
think a high-level interface to the "Big Brain" is a very valid use case. While I continue to develop low-level drivers in pasm/spin, I prefer to program in 12Blocks- especially when I want to demonstrate how to do something...
Hanno
Duane, I see your point. However, the ad says it was programmed by a 6-year old. But you never know. It can draw vector graphics, do color, play WAV files, handle TV, VGA, do sprites, create tones, synthesize speech, record sounds, handle mouse, keyboard and terminal events, graph variables - handle plotting and functions, control 32 servos, broadcast messages, handle a receiver block, access a terminal, use functions and calls, variables, arrays, access to VARS library, toggle pins, send and receive data with RS232, shift data, measure frequencies, count edges, and measure duration of discharge in an RC circuit, includes graphical tools, function parameters, can create State Machines... It sounds like the 6-year old in all of us could enjoy.
I initially designed 12Blocks ... to let users focus on the logic of what they're trying to do ... a high-level interface to the "Big Brain" is a very valid use case. While I continue to develop low-level drivers in pasm/spin, I prefer to program in 12Blocks- especially when I want to demonstrate how to do something...Hanno
I can see the potential of programming it to make a Big Brain front end - making complex Brain things very simple to handle and program. It reminds me somewhat of Odesta software - I'm very fond of the iconix method of programming having spent years with a Double Helix programming environment. http://en.wikipedia.org/wiki/Helix_(database)
Think of 12Blocks as a hybrid- easy to use, visual, with nice integrated tools calling into low level drivers written in spin/pasm. As such, I
think a high-level interface to the "Big Brain" is a very valid use case.
Thanks for clarifying what 12Blocks can do. I keep wondering about holding a "Robot Summer School". I ought to investigate 12Blocks better to see if I could use with the school.
SpotLite Index - Propeller & AMD
Introducing various SpotLite Indices
This introduces SpotLite Index, to save time and accumulate a sub index of information for ongoing study, research and project development about specified subject matter. The topic of this SpotLite is AMD and the range is from pages 1 through 55 inclusive.
The AMD project is to connect 720 GPU's to the Propeller Array and for over a 100 props of over 16,000 theoretical MIPS it adds over a TeraFLOP of computing power and numerous resources.
Definition
SpotLite Index - a small sub index designed to save time and accumulate information for ongoing study, research and project development about specified subject matter. Includes the main topic, links, pages, posts, post titles, and the range of posts.
The study and design of Brain Stubs is progressing. The project will also require, in addition to technical merit, some art and imagination. Unfortunately there were no suggestions for the shape of a Brain Stub so the project is progressing slowly. Apparently the right side of the gray matter is mute.
The Investigation
The investigation begins with weight and mass. Since the stub will be servo driven with props, it must not have too much mass. Although Wonder-Board is extremely light weight, it still has mass under consideration and appreciable weight in large sizes at various extended positions.
Determining Mass and weight
Both are determined by shape, and shape must adhere to the requirements set in the previous Brain Stub post. For mass, the formula F = (M)(A) is utilized. The servo has ramping and can therefore control inertia.
Control
Control will be under the Brain Stem as previously planned. This allows many functions of programming for the servo(s).
DOF
Now under consideration are two degrees of freedom but it is not known where these will locate. Enough parts exist to do 2 DOF and this is in compliance with the Brain Stem. The board can immediately handle 2 servos.
Brain Stub Design
This week the trial period was up on the CAD design program. Today, the program was purchased and the Brain Stub design can resume.
Quantity of Stubs
Do we go with a quantity of redundant Octopus-like stubs or could the Brain could get by with one stub? Two stubs open up some possibility of grasping. One stub would be faster, cheaper, lighter, and easier to construct. It would consume less power, and remain easier to program.
Stub Focus
Focus is now on designing the shape of one Brain Stub.
Hardware
The current idea is to use a PING))) kit for the interface connection to a constructed Wonder-Board Giant Brain Arm run by the large servo.
Dimensions
It's possible the Giant Brain Stub Arm could be 4-feet long and 1 or 2 feet in diameter. This is however subjective until the design is more solidified.
Shape
The shape is estimated to be grotesquely large, overbearing, ungainly, shocking, and with an appearance never before seen. It could have a very large hand at the end of the arm. But the hand would not be fully articulated. It must however meet the minimal requirements set forth in the previous Brain Stub post.
Servos
Should these servos mount standard or should they be continuous rotation servos? Maybe one of each - the large had could continuously rotate at the wrist (for some special functions) while the globby arm could dangle in varying degrees at the lumpy shoulder.
Possible Conclusions & Suggestions
A single Brain Stub
A blobbed shape never before seen
Made from manufactured WonderBoard
Interfaced with Parallax PING)))
Two degrees of freedom
Servo driven
A massively large hand
Side mounted
Articulation at the shoulder or elbow
Articulation at the hand wrist
Extremely large overall size ~ Four Foot extent
Grotesquely appearing
Light weight to minimize inertia
Programmed/Operated by the Brain Stem
Employing one continuous rotation servo at the hand wrist
This is the AX-12A which replaces their earlier AX12+.
I have a robot arm with some AX-12+ controllers (servos). The controller are really cool because you can position them like a standard servo (but not using the same communication protocol) and they can also rotate continuously. They have a 60 degree "blind spot" were they wont know their location but for 300 degrees they'll tell you what their position is as they rotate.
One of (the many) cool things about Dynamixel controllers is the communication protocol. It uses a serial communication protocol that includes the controllers ID so they can be daisy chained together and you can control a whole bunch (256) of them with one IO pin (and two resistors).
The AX-12As have pretty good torque 16.5kg*cm (@12V). If you want a real heavy lifter, you could move up to a EX-106+ with torque of 106kg*cm (@18.5V). These will break finger bones if you're not careful. They cost $499! You would also need a RS-485 chip in order let a Prop communicate with the EX-106+.
I'm guessing you already have the servos you're going to use but you might want to keep the Dynamixels in mind for a future project.
WOW. Duane, you MUST be rich to afford all those fancy schmancy Dynamixel servos. So you're the second rich guy I know who lives in Idaho. Caleb Chung, of Furby and Pleo fame, also lives there.
Caleb, is it you posing as Duane? Remember Letterman?
WOW. Duane, you MUST be rich to afford all those fancy schmancy Dynamixel servos. So you're the second rich guy I know who lives in Idaho. Caleb Chung, of Furby and Pleo fame, also lives there.
Caleb, is it you posing as Duane? Remember Letterman?
Erco, Shush!
Now you know why I don't have my picture for an avatar (or anything else for that matter(yet)).
I once heard that Idaho has more millionaires per capita than California (It might have even been per capita of all 50 states (we don't have a lot of capita here so it would not take many millionaires). This was about 25 years ago. I don't know if it's still true or not. (I don't know if it was true 25 years ago either, but I heard it from a friend.)
Just because I posted a picture of the EX-106+ doesn't mean I own one. I don't.
I do have a relatively good number of AX-12+ servos. They're really nice to use. I think I have enough to make a humanoid robot if I took all the AX-12+s out of their current projects.
P.S. I missed the last decade of Letterman shows. I have no idea who Caleb is.
Duane: the Dynamixel servo has good specs and is a good suggestion - it appears they would work well for the Brain Stub project. Having a servo that can double as both standard and continuous rotation, plus report position, is a good feature. I have Parallax / Futaba standard and continuous rotation servos on hand. It appears the 43.1 oz./in rating will be sufficient to move each stub section. The board is really light weight - lighter than expected. But tests are needed with all three sections connected together. The inertial result can be minimized by moving it more slowly, plus it only needs 2 DOF.
LED Brain Application Chronicles - Part 1 What can it do?
One of the best devices deep inside the Big Brain is the simple and inexpensive LED. While most people appreciate its significance as a diagnostic tool, others may or may not realize it has many more functions and is both versatile and multipurpose.
While LEDs are easily simulated on monitors and displays, it's preferable in this project to put at least one LED on each Propeller chip. The physical positional relationship is a help during design work and pacing the machine.
The proverbial LED can function as both an input and an output device. This opens up a gamut of possibilities.
LED Functions
Light detection as a photodetector
Debugging Tool
Experimenting with Tri-States
Output, input
Propeller Partition talk
Display
Alerts
CPU variable reporting
Prop Enumeration Indicator
Comparing props
Communicator
Actual Data out
Signaling and Signing
Conveying timing
Cog identification
EOPs and SBs (End Of Programs & Stop Blocks)
Indicate data presence
Work Tasking identifications
Sequential Identifier
Task Exampling
Binary device
Parallel demonstrations
Syncing Illustrations
Other Possible Functions
Touch Pad
Matrix
Talk on a Light Beam
Flood Communicator
Group Functions
Upcoming Part 2
In Part 2 of LED Brain Application Chronicles, we'll cover an idea for adding over a hundred LEDs to the machine and keep the power draw down to an essential minimum.
In response to the possibility of doing this project as one big simulation on a computer...
"Great ideas.... All the fun (to me) is in building the real machine and interfacing it and actually programming it - not simulating it on a PC. It's like simulating a fine dinner on a PC, you just don't get the same satisfaction as the real thing."
In another spinoff in the opposite direction, a simulation project is started. A lot of what's being done with hardware can be accomplished with simulated software on a very powerful platform. But there is a twist..
In this interesting China brain project, PDF linked below, they quote one important criteria.
If the average person can remain intrigued for half an hour by such a robot and the A-Brain that controls it, then we state that the A-Brain has passed the “China Test”.
Here's another brain project that has run for six years! They promised a billion neurons by 2001 but reduced their attempt to 40 million. The project conclusion is unknown.
Table of content
(Download the document to get access to the content)
Preface
Its all about linking and comparison
__Warming Up
__Linking, comparison and decisions
__Neural Networks
Procedural or not ?
__Introduction
__Thanks time and space !
__Scripting is the answer !
____Other advantages of scripting
The main protector
__Introduction
__The amygdala
Emotions and feelings
__Some free thoughts on emotions
__Fear
__Feelings, the derivatives of fear
__Disgust and surprise
__Curiosity and acceptance
__Pain
__Aggression
Believe, the driving force !
__Introduction
__Believe, always believe
__The three states of believe and doubt
The diverse forms of consciousness
__Introduction
__Consciousness
____Some free thoughts
____Another approach towards answers
__Conscious versus subconscious
__Self-awareness (self-conscious)
__Creativity, dreams, imagination and thinking
__Impact on the digital brain
Unexpected processes of the subconscious
__Intuition or following the own instincts
__Imitation and insight
What about hormones ?
__Introduction
__Hormones
__An odd question : is puberty required ?
__The impact on the digital brain
____The hormone part
____The pre-juvenile phase
____The juvenile phase
Deeper thoughts
__'I think, therefore I am ...'
__The buzzword 'perception'
Values and ego
__Introduction
__Relativity
__Standard of values
__The world turns around you !
__Self-interest, always !
__Altruism
A number of important actors
__Motivation
__Appreciation versus punishment
__Instincts
Memory
__Introduction
__What about synapses ?
__Storing information
__The inner voice
__The principles behind belief and fear
__The play-zone
__The long-term memory
__Visualization
__Fragmentation
__Mirror neurons
__What types of memory are needed ?
Attention
__Introduction
__Concentration
Appendix 1 : A sample diagram
__Introduction
__The diagram
__Explanation
____Digital Cerebellum (A) and the Play-zone (B)
____Pain (C)
____Surprise (F)
____Attention (E)
____Curiosity (G)
____Digidala (J)
____Labeling (L)
____Data Temp (M)
____Digi LTP (Long Term Potentiation) (N)
____Main Data (O)
____Acceptance (I)
____Rewarding (K) and Punishment (D)
____Instincts (P)
____Conscious versus subconscious
Epilog
This is a parallel schematic sketch designed for brain testing. Some ideas for use: The parallel circuit can open up eight channels of communication in true parallel. It can also serve out an eight bit byte. It double duties as a serial interface. It can open up packets of send and receive. One or eight can talk while hundred(s) of others listen. Not shown are typical protection resistors across each pin.
Comments
There's a useful thread about running multiple instances of the Propeller Tool Software and the PST.
http://forums.parallax.com/showthread.php?132744-multiple-PSTs
This function can attach to specific Propeller chips and act as virtual reporting screens, take the place of output LEDs, serving as LCD and TV monitor replacements, and other useful functions in a multi-chip Propeller machine.
12Blocks now has a Mac version. Has anyone tried programming multiple props with it?
http://12blocks.com/download.php
It has a trial version, and after that:
$49 - Lite (no function or arrays)
$79 - Standard (limited time offer- normally $99)
http://12blocks.com/resources.php comprehensive resources
http://forums.parallax.com/showthread.php?132724-12Blocks
http://onerobot.org/education.html 25page+19page downloads+video written by post-doc
http://12blocks.com/tutorial.pdf 35 pages with >20 activities
http://12blocks.com/videos.php several videos
http://12blocks.com/manual.pdf 29 page manual- yes, this is outdated...
Edit: refer to post 1095
Here's a picture of a class using 12Blocks on Macs to program TBots:
What do you mean "programming multiple propellers"? As long as you just want to program one at a time you're fine: 12Blocks has a nice "port chooser"- that either automatically connects to a Propeller- or let's you manually specify which port to use. In demos I frequently show 12Blocks connecting to multiple bots/projects.
Hanno
:In the case of the rotated picture - no problem with a Snow Leopard Mac. Just drag it to the desktop and turn it.
The Big Brain has a HYBRID interface under development for some time which includes various circuits: examples are Serial Communications, Daisy Chain, Token Ring Topology, Prop to Prop, and Parallel.
Definition
Reduction Interface RI - the simplification and combining of wire interfaces to achieve one interface that can handle the functions of several
HYBRID Interface - several interfaces mixed together, not necessarily reduced
This post introduces the Reduction Interface RI. The implementation of RI will reduce the number of prop wires by combining several interfaces together and mutualizing various functions.
This week sees the reassembly of the first two Partition arrays for 800 cogs after moving to the larger containment rack. A cog testing program is needed and suggested. Pseudo code below.
This makes a new smaller machine of 100 props with 800 cogs. Partitions were built at 50 props each, however since the introduction of the Filter, it is likely the number of props per Partition could double (perhaps a future consideration) and continue to work reliably.
POa1 { Loadn [ParaP = Partition1]} + POa2 { Loadn [ParaP = Partition2]} + POa3 { Loadn [ParaP = Partition3]}
PO - Power On
Loadn
ParaP
Partition
a0...an - array 0 through n
POa1 + POa2 + POa3 = ||
Loadn ~ EEPROM ~ Partition1 + Partition2 + Partition3
= ParaP1 + ParaP2 + ParaP3
(continued)
12Blocks let's you "inline" pure spin with the "code" block. You can also type any valid expression or condition into appropriate parameters. If you want to reuse existing spin objects you can "file>import" them and 12Blocks will create blocks for your public functions. Do you have a usb connection to each of your props? Or are you loading one and telling it to copy to others?
Hanno
I haven't used 12Blocks but I'm guessing it's kind of like the interface to program Lego MindStorms. You arrange the graphical blocks to indicate what you want your robot to do and in what order.
I doubt it's the best programming environment for programming unusually behavior like multiple Propeller communication protocols.
I'd think programming directly in Spin and PASM would be easier than trying to bend 12Blocks into doing things for which it wasn't intended.
Duane
You're right, I initially designed 12Blocks to simplify programming the Propeller so my 5 year old daughter could do it. It takes away most of the hard things: syntax, grammar, and vocabulary to let users focus on the logic of what they're trying to do. It adds a bunch of features that I initially developed for ViewPort to simplify debugging- graphing variable values, seeing pin states, and changing program parameters while your program runs. While it started as a simple project- it's now very complete- even letting you create stand-alone "exe" files similar to the PropScope application- with bundled firmware, graphical gui and communication stack.
Think of 12Blocks as a hybrid- easy to use, visual, with nice integrated tools calling into low level drivers written in spin/pasm. As such, I
think a high-level interface to the "Big Brain" is a very valid use case. While I continue to develop low-level drivers in pasm/spin, I prefer to program in 12Blocks- especially when I want to demonstrate how to do something...
Hanno
I can see the potential of programming it to make a Big Brain front end - making complex Brain things very simple to handle and program. It reminds me somewhat of Odesta software - I'm very fond of the iconix method of programming having spent years with a Double Helix programming environment. http://en.wikipedia.org/wiki/Helix_(database)
http://letsmakerobots.com/node/22528
http://www.instructables.com/id/Program-a-microcontroller-With-12Blocks/
http://hannoware.com/
http://12blocks.com/
http://blog.hannoware.com/
http://forums.parallax.com/showthread.php?114059-12Blocks-v1.2-launched!
www.12blocks.com/tutorial.pdf
http://12blocks.com/videos
http://12blocks.com/cheatsheet.pdf
http://12blocks.com/resources.php
http://12blocks.com/download.php
http://forums.hannoware.com/
Thanks for clarifying what 12Blocks can do. I keep wondering about holding a "Robot Summer School". I ought to investigate 12Blocks better to see if I could use with the school.
Duane
Good timing! I just posted a link to a talk I recently gave about Robots and Education- it's here:
http://forums.parallax.com/showthread.php?132806-Hanno-s-latest-Google-Tech-Talk-quot-Robots-and-Education-Bridging-the-Gap-quot
Quite a few schools are now using 12Blocks- I'm even working with a university to develop an official curriculum to bring 12Blocks and TBot into classrooms. Let me know if you need any help...
Here are links to everything related to 12Blocks: http://12blocks.com/resources.php
Hanno
Introducing various SpotLite Indices
This introduces SpotLite Index, to save time and accumulate a sub index of information for ongoing study, research and project development about specified subject matter. The topic of this SpotLite is AMD and the range is from pages 1 through 55 inclusive.
The AMD project is to connect 720 GPU's to the Propeller Array and for over a 100 props of over 16,000 theoretical MIPS it adds over a TeraFLOP of computing power and numerous resources.
The full Index is found here:
http://forums.parallax.com/showthrea...l=1#post977025
Page 33
650 Apple AMD Radeon HD 6750M Graphics Card 480 Stream Processors
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page33
Page 46
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page46
910 Merging Technologies - Big Brain with AMD
912 AMD Sources
Page 47
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page47
923 Developing AMD for Parallax Propellers
924 AMD (to Parallax Propeller) Development Boards Source
925 AMD’s Software Kit Source
929 More about AMD GPUs
Page 51
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page51
1004 Big Brain's AMD Radeon HD
Page 52
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain/page52
1023 Big Brain Expansion with AMD Radeon HD, High Speed Streaming GPUs
Definition
SpotLite Index - a small sub index designed to save time and accumulate information for ongoing study, research and project development about specified subject matter. Includes the main topic, links, pages, posts, post titles, and the range of posts.
Discovery post
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=982979&viewfull=1#post982979
The study and design of Brain Stubs is progressing. The project will also require, in addition to technical merit, some art and imagination. Unfortunately there were no suggestions for the shape of a Brain Stub so the project is progressing slowly. Apparently the right side of the gray matter is mute.
The Investigation
The investigation begins with weight and mass. Since the stub will be servo driven with props, it must not have too much mass. Although Wonder-Board is extremely light weight, it still has mass under consideration and appreciable weight in large sizes at various extended positions.
Determining Mass and weight
Both are determined by shape, and shape must adhere to the requirements set in the previous Brain Stub post. For mass, the formula F = (M)(A) is utilized. The servo has ramping and can therefore control inertia.
Control
Control will be under the Brain Stem as previously planned. This allows many functions of programming for the servo(s).
DOF
Now under consideration are two degrees of freedom but it is not known where these will locate. Enough parts exist to do 2 DOF and this is in compliance with the Brain Stem. The board can immediately handle 2 servos.
Brain Stub Design
This week the trial period was up on the CAD design program. Today, the program was purchased and the Brain Stub design can resume.
Quantity of Stubs
Do we go with a quantity of redundant Octopus-like stubs or could the Brain could get by with one stub? Two stubs open up some possibility of grasping. One stub would be faster, cheaper, lighter, and easier to construct. It would consume less power, and remain easier to program.
Stub Focus
Focus is now on designing the shape of one Brain Stub.
Hardware
The current idea is to use a PING))) kit for the interface connection to a constructed Wonder-Board Giant Brain Arm run by the large servo.
Dimensions
It's possible the Giant Brain Stub Arm could be 4-feet long and 1 or 2 feet in diameter. This is however subjective until the design is more solidified.
Shape
The shape is estimated to be grotesquely large, overbearing, ungainly, shocking, and with an appearance never before seen. It could have a very large hand at the end of the arm. But the hand would not be fully articulated. It must however meet the minimal requirements set forth in the previous Brain Stub post.
Servos
Should these servos mount standard or should they be continuous rotation servos? Maybe one of each - the large had could continuously rotate at the wrist (for some special functions) while the globby arm could dangle in varying degrees at the lumpy shoulder.
Possible Conclusions & Suggestions
Do you know about the "servos" Dynamixel makes?
This is the AX-12A which replaces their earlier AX12+.
I have a robot arm with some AX-12+ controllers (servos). The controller are really cool because you can position them like a standard servo (but not using the same communication protocol) and they can also rotate continuously. They have a 60 degree "blind spot" were they wont know their location but for 300 degrees they'll tell you what their position is as they rotate.
One of (the many) cool things about Dynamixel controllers is the communication protocol. It uses a serial communication protocol that includes the controllers ID so they can be daisy chained together and you can control a whole bunch (256) of them with one IO pin (and two resistors).
The AX-12As have pretty good torque 16.5kg*cm (@12V). If you want a real heavy lifter, you could move up to a EX-106+ with torque of 106kg*cm (@18.5V). These will break finger bones if you're not careful. They cost $499! You would also need a RS-485 chip in order let a Prop communicate with the EX-106+.
I'm guessing you already have the servos you're going to use but you might want to keep the Dynamixels in mind for a future project.
Duane
Caleb, is it you posing as Duane? Remember Letterman?
Erco, Shush!
Now you know why I don't have my picture for an avatar (or anything else for that matter(yet)).
I once heard that Idaho has more millionaires per capita than California (It might have even been per capita of all 50 states (we don't have a lot of capita here so it would not take many millionaires). This was about 25 years ago. I don't know if it's still true or not. (I don't know if it was true 25 years ago either, but I heard it from a friend.)
Just because I posted a picture of the EX-106+ doesn't mean I own one. I don't.
I do have a relatively good number of AX-12+ servos. They're really nice to use. I think I have enough to make a humanoid robot if I took all the AX-12+s out of their current projects.
P.S. I missed the last decade of Letterman shows. I have no idea who Caleb is.
What can it do?
One of the best devices deep inside the Big Brain is the simple and inexpensive LED. While most people appreciate its significance as a diagnostic tool, others may or may not realize it has many more functions and is both versatile and multipurpose.
While LEDs are easily simulated on monitors and displays, it's preferable in this project to put at least one LED on each Propeller chip. The physical positional relationship is a help during design work and pacing the machine.
The proverbial LED can function as both an input and an output device. This opens up a gamut of possibilities.
LED Functions
Light detection as a photodetector
Debugging Tool
Experimenting with Tri-States
Output, input
Propeller Partition talk
Display
Alerts
CPU variable reporting
Prop Enumeration Indicator
Comparing props
Communicator
Actual Data out
Signaling and Signing
Conveying timing
Cog identification
EOPs and SBs (End Of Programs & Stop Blocks)
Indicate data presence
Work Tasking identifications
Sequential Identifier
Task Exampling
Binary device
Parallel demonstrations
Syncing Illustrations
Other Possible Functions
Touch Pad
Matrix
Talk on a Light Beam
Flood Communicator
Group Functions
Upcoming Part 2
In Part 2 of LED Brain Application Chronicles, we'll cover an idea for adding over a hundred LEDs to the machine and keep the power draw down to an essential minimum.
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=1015760&viewfull=1#post1015760
In response to the possibility of doing this project as one big simulation on a computer...
"Great ideas.... All the fun (to me) is in building the real machine and interfacing it and actually programming it - not simulating it on a PC. It's like simulating a fine dinner on a PC, you just don't get the same satisfaction as the real thing."
http://forums.parallaxinc.com/forums/default.aspx?f=25&m=475282
Humanoido
In another spinoff in the opposite direction, a simulation project is started. A lot of what's being done with hardware can be accomplished with simulated software on a very powerful platform. But there is a twist..
In this interesting China brain project, PDF linked below, they quote one important criteria.
If the average person can remain intrigued for half an hour by such a robot and the A-Brain that controls it, then we state that the A-Brain has passed the “China Test”.
http://www.agiri.org/docs/China-BrainProject.pdf
Hugo de Garis at Wuhan University (Wikipedia) - in 2008 awarded a ~$450,000 grant to build an artificial brain for China.
http://en.wikipedia.org/wiki/Hugo_de_Garis
Book to be published in November 2010: Artificial Brains - An Evolved Neural Net Module Approach (Amazon).
http://www.worldscibooks.com/compsci/7713.html
http://artificialbrains.com/
Here's another brain project that has run for six years! They promised a billion neurons by 2001 but reduced their attempt to 40 million. The project conclusion is unknown.
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.50.7592
Artificial Brain Projects Search
http://search.yahoo.com/search;_ylt=A0oGdW6XEBlO0AUABm6l87UF;_ylc=X1MDUCM5NTgxMDQ2OQRfcgMyBGNzcmNwdmlkA1Q3MzYuRW9HZFRBbTFjdlVUaFhtRkFNZjBESEhWVTRaRUpjQURZTVgEZnIDc2ZwBGZyMgNzZy1nYWMEZ3ByaWQDBGl0A2dwBG5fZ3BzAzEwBG9yaWdpbgNzcnAEcG9zAzQEcHFzdHIDYXJ0aWZpY2lhbCBicmFpbgRxdWVyeQNhcnRpZmljaWFsIGJyYWluIHByb2plY3QEc2FjAzEEc2FvAzEEc2VjA3JlbC1zYQR2dGVzdGlkAw--?p=artificial%20brain%20project&fr=sfp&pqstr=artificial%20brain
The document is still open source but the project has moved to closed source. Contains unique perspectives of artificial brains.
http://theartificialbrain.com/Data/The%20Artificial%20Brain.pdf
http://theartificialbrain.com/ArtBr.htm
This is a parallel schematic sketch designed for brain testing. Some ideas for use: The parallel circuit can open up eight channels of communication in true parallel. It can also serve out an eight bit byte. It double duties as a serial interface. It can open up packets of send and receive. One or eight can talk while hundred(s) of others listen. Not shown are typical protection resistors across each pin.