Trossen Robotics Community Blogs

build my own basic atom pro 28 carrier board

soldier — Thu, 11 Mar 2010 16:28:09 GMT

anyone know how to start to build our own basic atom pro 28 carrier board ?? without using Bot-board ll
i owned a bot-board ll
but i wanna learn more... anyone can help me??? please... how to build my own carrier board for basic atom pro 28... i need help... thanks...

From a New Member of Trossen Robotics Blogs-Introduction

jmhenry — Wed, 03 Mar 2010 13:51:32 GMT

I wanted to let everyone know that I am proud to be a part of the Trossen Robotics Blogs Family. Almost two years ago, I began to write posts on RobotNext. Now, I am attempting to broaden my experience by participating in other sites. Hopefully, you will find my posts of interest. In addition, I hope to create some value for the Trossen readers.

As a retired teacher, I should have a lot of time to spare, but between mentoring FIRST robotics, writing blog articles, and trying to get a business started - there is not much time left in the day. I have been working with robotics teams for almost 11 years both as a teacher and as a mentor.

Also, I try to read widely about robotics and what might be coming next in the field. This is one of the areas of robotics that I will try to share with the readers here.

One step at a time

badcommandorfilename — Wed, 03 Mar 2010 00:10:16 GMT

Accelerometers and gyroscopes are the usual solution for maintaining balance. They simply measure if the robot is tilting or falling over and can then take appropriate actions to maintain balance.

Unfortunately, Robobuilder does not come with either an accelerometer or a gyroscope. Instead, the wCK servos can be queried for their position and velocity which allows for another method of balance - known as kinematic feedback.

Observe:

"> " type="application/x-shockwave-flash" width="425" height="350">

Full details and more videos can be found here.

More up-to date posts can be found at http://buildtherobot.blogspot.com/

Introducing My Hex: Ally

parallax — Mon, 22 Feb 2010 14:54:26 GMT

Hi All!

I have been a lurker on this forum for quite some time (BTW, you guys are incredible) and I have decided to track my progress on my latest project here. I am building a pheonix/micromagic inspired hexapod that will primarily be a testbed for different computer vision and AI programs I am working on.My goal is to make him as autonomous and intelligent as possible. My senior project was a camera on a lynxmotion pan-tilt rig controlled by a phidget advanced servocontroller which tracked peoples faces using OpenCV and uploaded data to a webserver about what it saw. So my first step with Ally is to get that project mobile! My parts list as of now:

6 HS 225MG
8 HS 485HB (will eventually be replaced)
6 HS 645MG
9 Multipurpose Brackets (Servo Erector Set)
1 "C" Bracket
1 1.2G Wireless Surveillance Camera
SSC-32 Servocontroller
BlueSmirf Bluetooth Modem
2.4 GHz Duck Antenna
6.0 Volt 1600mA NiMH Battery
2 9 Volt Recharge-ables
Battery Harness
Lots of 1/8" (3.175mm) Lexan

I'm concerned about the flexibility of the Lexan. I've seen a YouTube video of someone who built a hex out of plexiglass and it was a wobbler. Makes me a little scared for my Ally. I can't have that for my bot! ;) Although Lexan is less flexible that plexiglass, it's not THAT much less flexible. My current plan is to shore his legs in places with thin strips of metal. The alternative materials I was thinking about were HDPE (too heavy) and aluminum, which is not very feasible for me. Any input would be appreciated. My criteria are it must be relatively easy to work with (do to lack of machining equipment), and light.

I have to run, but I will post some CAD drawings tonight.

Later,

parallax

Second Robotics project under way.

Meccano Jay — Fri, 19 Feb 2010 02:17:39 GMT

I really enjoyed building the simple Meccano walker and sharing it with everyone. Building the bot gave me a good basic intro into the world of robotics. I was able to familiarize myself with the Phidgets servo control and get a feel for programming for it. While all this is good, it is still very basic and with only 2 DOF, limiting as to what it can teach me. Not to mention leaving my phidgets controller almost empty.

The Meccano walker has been disassembled to make room for a whole new project! This bot will utilize 6 DOF giving it more control and freedom, not to mention being a much more worthy programming challenge! This bot will consist of a bit more than just your basic erector set pieces. I have purchased a bender and a few other small tools and have begun fabricating some of the pieces myself. I plan to expand this as my knowledge and experience grow.

I am very excited about this project, and though it isnt the greatest bot around I hope you will follow along on my journey of discovery in the world of robotics. I have been waiting to do this since my days at Amarillo College back when robotics was still just scifi.

I wish I still had my pieces or at least the pictures of my early work. I had constructed a torso to hold two servos as hip joints and had constructed a leg that had a knee and ankle. it was very crude and probably would have done a sloppy job of moving since the only means of control then was a multichannel R/C remote. I never finished the bot due to a lack of servos, my test servos had to go back into my race cars for the weekends. I was after all a college student on a budget.

Happy building
Jay

From the Graveyard - Golemech

Noog — Wed, 17 Feb 2010 04:57:39 GMT

Over time there are robots that are built, demonstrated to many people and talked about until they earn some sort of legendary status. Most of us have robots that never get anywhere near this status and now they end up sitting on a shelf wasting their days away. No one gets to enjoy them anymore. 'From the graveyard' is a little series that I would like to use to bring a couple of my old creations back to life. Maybe someone can enjoy them just a little more, and talk about them just one more time. =)

Overview:

Legs : 2
DOF : 6 active, 2 passive
Main Controller : PIC 18F
Height : ~18 cm (foot to main body plate)
Width : ~23 cm (arm to arm)
Length : 19 cm (nose to rear)
Materials : aluminium sheets and acrylic

Last time I worked on this robot: 2002

A few years ago when I was just a budding hobby roboticist, I figured that I should try to build a mech-like robot. Well, not a full sized mech, but my budget would certainly allow me to build one big enough to run around on a table top, which was perfect for me.

My goal was to design and build a small scale bipedal mech. To help achieve this, I set out a few guidelines:

1) Keep the legs slim and keep the weight up towards the body whenever possible.

I had built a few leg prototypes based on hobby servos before starting this robot. Cheap hobby servos could not provide enough torque to account for my poor engineering, so each of my previous designs failed rather quickly. I knew that I could account for some of my design flaws by using the higher torque servos, but my budget could only afford a couple. With this in mind, I would focus more making the mechanics simple so that I could get it up and walking without dropping a fortune.

2) Minimize the servo count.

Servos introduce slop, consume power with a vengeance and quickly add weight to any design. Don't get me wrong, despite their short comings they are still awesome. They are one of the outward connections between your code and the real world. That being said, I wanted to mitigate any problems they introduced by simply keeping the servo count low. My plan was to look at different mechanical solutions instead of just throwing more actuated degrees of freedom (DOF) at my design.

3) Use materials I already had.

Have I mentioned anything about the budget I gave myself at the time? It wasn't miniscule, but I would rather spend money on cool sensors or something that could increase the `wow` factor instead of buying many identical servos. Besides, being resourceful and re-purposing some old technology is its own reward.

Mechanics: Body

One particular mech design from Battletech was called the Marauder. This mech consisted of a jet like cockpit with long lanky arms and legs. It wasn't my favourite mech, but the shape always intrigued me:

(http://images.google.ca/images?q=marauder+mech)

On paper it looked pretty cool, but I was skeptical that this design would work in real life. I could say this for most sci-fi mechs, but that hasn't deterred me from wanting to build them either. =)

The vision of the Marauder with it's stretched out bulky arms inspired me to try and put some weight (specifically the main batteries) stretched out to the side of the my table top mech. This would help to lower my center of gravity without putting all of my batteries in the feet.

Mechanics: Legs

Some of my previous leg designs showed me what kind of problems I could expect with a reverse knee. Golemech incorporated the look of the reverse knee while avoiding the problems: The knee is not actuated at all, the reverse knee is merely cosmetic. Yes, this was the easy way out, but I still got my reverse knee design.

Each hip has 2 DOF: one to roll the leg to the side, one to pitch the leg forward and back (one ring to rule them all, one ring to find them... wait, that's a different tale). A third DOF was used to pitch the foot in the same plane as the hip pitch. Each hip/leg assembly was connected to the body with a passive DOF so that each could be turned (yaw). I planned to lock this passive DOF at first, and as I worked out the kinks from the rest of the DOFs, I would then upgrade this DOF to be actuated by another servo.

Mechanics: Feet

Golemech went through two iterations of the foot design. The first was a two toed approach, the second design only contained a single toe. With the single toe design, each foot would only have two contact points with the ground: the heel and toe.

When both feet were on the ground, a square would be created by the contact points where the center of mass could then be stabilized within. The contact points each had a momentary switch embedded within to signal when contact had been made. This was a low-tech solution that I was able to manufacture at the time. Ultimately, I planned to have a flex sensor or some strain gauges on each contact point.

Control System

The walking gait I envisioned for Golemech would be based off an inverted pendulum. Once standing still in a resting state, the hips would shift the weight of the body to one side, lifting the opposite foot. Golemech would essentially be an inverted pendulum, something that students had easily been controlling with microcontrollers at the time. Next, the lifted leg would be swung out to the side and forward, while the leg in contact with the ground shifted the weight of the body forward. Once the body weight came forward far enough, the leg in the air would then be in the proper position and would make contact with the ground. Repeat as necessary.

A fusion of accelerometer and gyroscope data by something like a Kalman filter would help determine the motion of the robot and be fed into the control system as feedback for the walking gait.

Electronics:

Shortly before I had created Golemech, I had helped out the folks at the Ottawa Robotics Group produce a new version of their control board called IgORE2 (http://www.ottawarobotics.org/igore.shtml#igore2). I had a bunch of these laying around and so it was a perfect choice to use as the main controller for Golemech. I was comfortable programming the PIC 18F series of chips and I believed it contained the processing power needed to at least make this robot hobble around.

Since servos actuated all of Golemech's joints, I needed a way to control them. The pic 18F controller could easily provide direct control, but I anticipated that I would need all the cycles available from that controller just to handle the dynamics of walking and balance. I decided that a dedicated servo controller board would be perfect to incorporate into this robot.

I had an old servo controller board that was called the Picopic from a company called Picobytes. I didn't know it at the time, but this product had been discontinued. The company's website and all documentation for this board had disappeared during my development phase. Booo! (As a result of this, now I always download and store a local copy of any documentation associated with 3rd party products I use. Life lesson learned: Check.)

At the time I had enough of a challenge trying to get the mechanics working that I only had a rough idea of what sensors I would try to incorporate into the Golemech. Aside from the 4 momentary switches in the feet, no other sensors were installed into this robot.

Ahh yes, then there were the batteries. I wanted to use something standard that would be able to handle the load. I had purchased some AA rechargable batteries from Rayovac (along with their special charger) that included technology that promised 2000mAH per cell and a 15 minute charge time. The batteries were wonderful! Their charge time usually came in around 13-14 minutes and they certainly packed a lot of juice. In each 'arm' of the robot I had included a 4 cell battery holder which held the batteries in series. The power from each arm of the robot was tied in parallel and then fed to the rest of the robot. This would give a nominal 4.8V @ 4000mAH. I doubt this battery setup would have held up well, but I was hopeful at the time.

In the end, the only electronics I had mounted was the IgORE2, the Picopic servo controller and the batteries. Control commands were successfully being sent from the main controller to the servo controller to test the servos and mechanics as they were being installed. Momentary switches were installed on the feet, but never wired up to the microcontroller.

Why did development stop:

I don't remember exactly what was happening around this time, but I do remember having some problems with the servo controller. When I went to find the answers to my troubleshooting, I realized that the company and their website with all the documentation for the servo controller was gone. I didn't realize it at the time, but this was the final nail in the coffin for this robot. I had other projects on the go that needed my attention and I had placed this robot onto the shelf with the intent to get back to it at a later date. Not much has happened since then.

The clarity of hindsight:

If I worked on this robot again, how would I do things differently? Hmm, the electronics certainly would be modified. The batteries would be upgraded to something that could deliver more of a punch. The sensor suite to handle the balancing of the robot would be prototyped and simulation testing could be in progress before doing a lot of the mechanical work.

I still like the idea behind the mechanics. Keep the design slim and sleek. If the robot is designed to be balanced but also slightly unstable, then the on board control system will handle the rest.

For now, a table top mech is still on my radar, but I'm not sure that the version of Golemech that exists on my shelf will ever be revived.

Noog

ps - here are a couple more pics...

Bluetooth up and working

Chench — Mon, 15 Feb 2010 00:42:04 GMT

I was having some trouble getting the phone to speak to the Picaxe chip over the Bluetooth, it was giving an error but not a very helpful one. I wrote an application for my laptop which had no problem, it had buttons to click to go forward etc. but for some reason the same approach didn't work on the mobile. Then my laptop bluescreened and the bluetooth part of my app got corrupted. Great. But after I rewrote it, it miraculously worked perfectly. I've got it so I can click buttons to turn, go forward, stop etc. and I can tilt the phone forward/back to drive using the accelerometer. All's well that ends well. However, I'd managed to drop the tank chassis off the table at this point and seemingly pizza base foam chassis' aren't up to that sort of treatment. If it's not one thing, it's another. So I'm thinking a somewhat more refined platform is coming up.
Now that I have this part of the project working, it's just fiddling around with numbers and it'll be able to do some decently cool stuff with the sensors I have working. Compass, accelerometer, light sensor and proximity sensor are up and running, sonar is also working but I still need to read the data into the app from the bluetooth. GPS is low priority as this guy will be mostly cruising around my flat. I've also got a commercial application that streams video from the phones camera over the Wi-Fi network to my laptop, and this reads straight into Roborealm, so simple vision processing is possible too. I'd like to add some solar panels to make it a bit more independent, a bit more Mars-roverish. Also it seems easy enough to tell my app to intercept text messages starting with a certain phrase and read commands from that, and also to report it's status, eg. battery state, sensor readings. Smartphones were made for robotics.
Quick video showing how it's working just now, there's a few things that need to be improved.
">" type="application/x-shockwave-flash" width="425" height="350">

Time to get inspired

Noog — Thu, 11 Feb 2010 05:20:16 GMT

Everyone has something different that inspires them to create. Some things are only powerful enough to inspire us for meerly a moment, while other inspirations stick with us for the rest of our lives. These are just a few of the things that have inspired me along the way.

Fiction:

Transformers had to be my first real introduction to robots, yet I`m not sure I can say they inspired me to create robots. These robots had personalities, friends and foes. They seemed less like robots, and more like regular people (with frickin lasers built in).

Role playing games happened to be my first introduction to mechs. These large bipedal tanks were created as the ultimate future weapons. They seemed like an awesome idea (I won't kid myself, they still do!). I never took the time to figure out how legged tanks surpassed wheeled technology, but I'm sure they had a good reason. =)

Shortly after finding games full of mechs, I watched an anime series called Macross (or Robotech depending on the series) (http://en.wikipedia.org/wiki/Robotech) which quickly caught my interest. The series featured alien technology that humankind had reverse-engineered to create their own jet-based mecha as well as more conventional land based mecha. After watching the first Macross series I was hooked.

Don't even get me started on Hollywood.

Indeed, all of these were cool, but I never saw them as something I could create. It wasn`t until I saw some of the robots being created in real life that I had the desire to even attempt to create one myself.

Non-fiction:

I had always been inspired by the robots that came out of the MIT Leg Lab (http://www.ai.mit.edu/projects/leglab/). They had a people working on legged robots, wheeled robots, simulated robots and many more types. I've spent many hours scouring their site to absorb as much information as I possibly could. My early attempts at legged robots were based off a couple specific designs from the MIT Leg Lab:

Spring Turkey (http://www.ai.mit.edu/projects/legla...ng_Turkey.html) which was interesting in that it used Series Elastic Actuators to introduce a bit of 'give' into their design. Motors were built into the body of the robot and then power was transferred out to the joints.

Spring Flamingo (http://www.ai.mit.edu/projects/legla..._Flamingo.html) Unfortunately their website is a little wonky at the moment, but you can see a picture and a video here (http://yobotics.com/robots/robots.htm). This robot was designed with the majority of the weight in the body. Some pretty smart people had already worked on power transfer mechanisms so by studying their design, I could learn a couple tricks to use when building my own similarly styled robots.

Both of these robots were designed as planar robots, meaning they were only meant to work while attached to a pole for balance. This restricted their movement to only 2 dimensions, but at the time I wanted to have Golemech (an early biped of mine) work in the third dimension as well. Peter Dilworth was working at MIT back in the day and he spent some time making a robot called Troody (

YouTube- Troody the Biped Dinosaur

). The design itself is based off of a Troodon (http://en.wikipedia.org/wiki/Troodon), a dinosaur that is thought to have lived around 65 million years ago. Dilworth later went on to start a company where he intended to rent out robotic dinosaurs to museums and other similar institutes. Since I don't seen any robotic dinosaurs running around at my local museum, I assume his business isn't booming. Although I am glad to see that some people (http://www.dinosaurlive.com/) have had better luck with robotic dinosaurs.

More recently, Robo-One (http://en.wikipedia.org/wiki/Robo_One) has been a great source of inspiration as far as bipeds go. The early robots struggled to complete the events. A quick search on Youtube for any Robo-one videos will show you that the robots entered these days have no problem competing, let alone amazing the crowds. With companies creating humanoid kits, it seems than anyone with the cash can get started in this sport pretty quickly. Sure, these kits look wonderful, but the custom built robots seem to inspire me the most.

Whether I find the solutions to my problems or I find a life-long muse, one thing is for sure: inspiration can be found in everything around me. So... what really inspires you?

Noog

Two Steps Forward...

badcommandorfilename — Wed, 10 Feb 2010 03:50:20 GMT

One step back. Improvements to the robobuilder firmware are coming along, but I was met with an unfortunate setback when I accidentally managed to strip another gear in one of the servos.

This time though, it was the #1 drive gear - the only plastic gear in an otherwise all metal chain. This leads me to believe that it broke simply due to regular wear and tear - not because of any specific negligence on my behalf.

You can read the full post here.

As usual, the most recent entries can be found at http://www.buildtherobot.blogspot.com

Mirror Mirror

badcommandorfilename — Tue, 02 Feb 2010 11:47:23 GMT

The wCK servos are some of the best digital servos I've dealt with. You can certainly find servos with more torque or better speed, but don't neglect the importance of all those extra features which are so easy to gloss over.

The wCKs have a list as long as my arm of cool useful extras, but the one I am exploring today is their ability to go into 'passive' mode. This removes all torque from the servo, turning it into a simple rotational encoder. Combined with the ability to query individual servos for their current position, this enables individual servos to become sensors!

The possibilities of such features are numerous, but this is just a quick demonstration of what can be done:

">

" type="application/x-shockwave-flash" width="425" height="350">

For source code and much more, please see the full post.

More recent posts are available at http://buildtherobot.blogspot.com

General update

Meccano Jay — Tue, 02 Feb 2010 09:20:47 GMT

I will be re releasing the walkers software, there is a slight bug in the current version. It is not allowing servo 1 to engage. simple fix, will work on getting it done tomorrow.

Posted a few pics up of the walker in action, still adjusting to the new cam. Bare with me I will get it set right, and ul smaller photos in the future.

Testing a mod to the foot, the angle girder used as a control arm has been removed and the servo connector mounted to the foot itself. Will further test this and report. Getting the log/how to together, should have that posted up with new images soon!

TTFN
Jay

B-Day Goodies Rock!

Meccano Jay — Mon, 01 Feb 2010 06:06:25 GMT

I got a belated (due to icy roads) B-Day trip to Lubbock today! Yay Tool shopping ehehe, too bad we don't have a decent tech store around. Ah well got some cool new goodies and a Kodak M863. No more Cell phone pix !!

This is exciting to me as I can now get the quality pictures I need to do the how to/project log For my Basic Meccano Walker. Not to mention allowing me to get started on some more groovy projects!

Tools Include:
18" bending brake
(3) 2" C clamps
Small wire cutter/stripper
4" digital caliper
Multifunction DMM
4 speed rotary tool kit
pics coming soon ...

Snoopy's looking brainy!

kamondelious — Sun, 31 Jan 2010 20:02:03 GMT

I mean, the lights are on, but there's nobody home.

Part of me is thrilled, but part of me is sad. I've been thrilled with all the custom make-shift work I've been doing to build Snoopy's head, but now I kind of want to redo the rest of him just for the fun of it. Not to fear, I have more changes in the works and I will continue to re-factor him until he's as functional as I intended.

So here's what went into making lightbright brains with a face only Darth Vader could love!

I learned shortly after purchasing the BeagleBoard that there is no built-in microphone amplification and that the stereo audio in was merely a stereo line-in. I did some looking around online for how to build my own mic amp and after looking over a few schematics and discussing it with a colleague at work decided on one that seemed more than adequate. Shortly there-after, I made a trip to a great electronics store I found to pick up parts. After some shopping and picking out all the various components I needed, I wandered over to their section where they have various little electronic kits, some pre-assembled and some leaving the fun to you. Well, wouldn't you know it, they had what I was looking to build in a smaller form factor. I picked up 2 of the DIY kits, one for each ear, along with blue LEDs to replace the red ones and built them the following weekend. They turned out pretty good and work great!

I do love building things. so the little DIY mic amp kit make me feel like a little kid with a new puzzle. I also like to attribute my soldering abilities to growing up with a loving Mother who happened to be a welder. Genetics?

Here's the only picture I think I have of Snoopy's eye mounting outside of his head. This might need some tweaking to get easier position adjusting.

The top screw hole pretty much just hold it up in the head, while the other 2 allow for some independent adjustment as the base is cut halfway up the middle. Definitely not ideal, but seems effective so far.

Next I had to figure out how to fit/mount the microphones for ear onto the head in a manner that would be easy to work with and on. The microphones I used are nothing special, just a couple of really cheap clip on mics I picked up in one of my orders to dealextreme for < $2 each.

I have no hesitations over chopping up $3 worth of microphones. :D

I drilled holes in the sides of Snoopy's skull, just big enough to snuggly fit the housing for the mics.

This might have held up ok on it's own, but I kept having horrible visions of Snoopy falling over (because we all know that's going to happen) and his ears popping out and how sad that would look, even though they would pop back in.
After much deliberation and a general dislike of the idea of using glue I thought of a way to make use of the clips that came with the mics to mount them in a way that was easy to disassemble (because we all know that's going to happen) yet sturdy/secure. I pushed out the hinge pin from the clips and discovered it was the perfect size for the stock Bioloid bolts. That and the thickness of the plastic (bottoms off a soymilk jug) of the head made for an excellent mounting opportunity provided I didn't screw up the required drilling.

I cut the 2 squarish holes in the top of the head above the ear holes to fit the half of the clip that attaches to the mic. This put the hinge holes above the plastic on the outer edge, but covered on the inside. This is where I'd need to drill to allow the bolt to go in.

As you can see, the spacing in the clip hinge was also the perfect size for the stock Bioloid nut. SCORE!

Quick shot of both in place. Yes that's a string on the top front of the head. It's there to hold down the front half of the cap that cover's Snoopy's "brains". The corresponding squarish hole on the back of the head is where the cap is bolted in place.

Here we get to see the completely mounted microphone from inside the head.

I had to mount the eyes in too as I wanted to run the USB cables for the cams behind the ears (inside the head). You can also see that the top mounting bolt for the eyes is also what I hook the string on to. Handy!

I used part of a paper towel tube covered in a balloon in conjunction with the USB powered speaker I mentioned in a previous post to make Snoopy's "voice box".

This is based off the same principals as making a bass speaker which adds a nice richness to the sound. I've been listening to music off it and as long as I don't crank the volume all the way up it sounds pretty good. Too high and it starts to make farting noises, which is comical and fun, but not the desired effect for the short term. The bolt off the bottom is what mounts to the tilt AX-12 and holds the lower half of the head in place.

I don't seem to have pictures of how I mounted the mic amps on, nor pictures of just the lower half of the head mounted. :(

But when I fully assembled it, it looked like this!

:D Not too bad, but wouldn't you know it, I had to make another trip to my favorite electronics store (again :D) to pick up some heatsinks and caps to go with the 7805's I have to step-down power to my various non-bioloid accessories. While there, right next to the heatsinks, were fan grills and wouldn't you know it, they had one that looked just the right size for Snoopy's mouth. At like $7HK (less than a dollar US or CDN) it was worth a try.

$7 well spent I'd say! Though my craftmanship left a little to be desired. I think I can clean that up a bit and make it prettier later.

There are a few more shots of Snoopy's "brains" in my Bioloid album here :
http://forums.trossenrobotics.com/album.php?albumid=32

Here's a happy little Snoopy sitting all fully intact.

I don't have a lot of room to work here, and have been a bit paranoid of Snoopy falling off my desk. So, I made him a crude sort of Jolly-Jumper.

I think I will refine this a bit and make it more adjustable as well as easier to put in place.

I've had a few people express concern on Snoopy's weight and how he may not be able to walk. This had me concerned for a while as the idea would be dreadful. Then I was reading though a great thread, "Cheap Humanoid"
http://forums.trossenrobotics.com/showthread.php?t=3476
and came across this video.

YouTube- Runbot's 1st prototype

This shows a 6lb 23DOF robot walking with servos that rate at 14kg.cm of torque.
As most people around this forum know, the Bioloid AX-12's rate at 16kg.cm of torque and at Snoopy's last weigh in on the Wii fit, he was a hefty 6lbs. As I expect he's going to be losing more weight than gain over the next while, this should bold well for his walking capabilities. My ability to make him walk however, well, I made him a Jolly-Jumper because I was scared he'd fall off my desk. ;)

This weekend I got to playing with OpenCV a bit and was quite disappointed by it's performance. I wasn't doing anything fancy, just one of the demo video capture examples. The image was all smeary looking. I might poke back at it later, but will probably just stick with v4l (Video 4 Linux) directly. I used one of the example capture programs to snap this little ditty of myself.

The colour palette is obviously off as I'm looking Smurftastic! But I don't mind the image quality otherwise. Snoopy's cams are manual focus, so the bit of blurriness is from that. I wonder if that's why OpenCV was looking smeary? Maybe it's trying to be too smart averaging across frames and with the slight blur, gives a smeary tracer effect. hehe, neat!

Such great fun!

Cheers!

Aaron

Hi I'm Meccano Jay

Meccano Jay — Sat, 30 Jan 2010 07:36:02 GMT

Hi I'm Meccano Jay.

I am a hobbyist, a tinkerer, a DIY er of the digital age. I enjoy Building bots, extreme PC modding, and even designing car audio systems. Not to mention repairing most small gizmoes and what nots that my friends and family bring to me. Whether its a dead PC, a busted PSP, or a Classic Arcade, I have probably had my hands under its hood!

I am very excited about my new robotics hobby! I will be posting my projects along with software and code examples. I currently build using Meccano / erector pieces. Control comes from a Phidgets 1061 Advanced servo 8 motor control. Hitec HS-322HD servos provide the muscle. I hope my projects are both entertaining and informative.

I think what I like most about robotics is that it requires imagination.

Happy Building
Meccano Jay

Saftey First

badcommandorfilename — Thu, 28 Jan 2010 22:32:22 GMT

Rewriting the firmware for a robot with so many servos can be hazardous to it's health. Small bugs can result in servos performing random actions, causing your robot to spaz out like an epileptic kid during a fire drill!

This can result in damage to your servo gears or possibly even more expensive parts. I suggest minimising damage by disassembling as many servos as is possible before testing potentially dangerous changes, as shown here.

For more information, please visit http://buildtherobot.blogspot.com/

All Work and No Play

badcommandorfilename — Wed, 27 Jan 2010 23:48:32 GMT

A lot of code and technical details have been going up on my blog recently, and it's starting to feel a little boring and dry.

I thought I'd take a little time out to introduce some of the most inspiring hobby robots out there. These were the ones which influenced me to start my own robotics project, and I hope they encourage you to do the same.

Full post is here.

As always, more can be found at http://www.buildtherobot.blogspot.com

Some progress!

Chench — Mon, 25 Jan 2010 19:10:16 GMT

Well I had a good read at a book on C#. I didn't really feel I was learning much but I must have soaked it up because I had another look at getting the Samsung SDK for the phone's sensors etc. to work, and it seemed a lot clearer than before! So now my application can access the compass heading, pitch and roll angles, accelerometer x,y and z values and turn the camera flash on and off. The last sensor I'd like to get working for now is the light sensor. I had a try today but got an error so I'll have to work that out. My Bluesmirf wasn't working so I sent it back to Sparkfun for repair, but once it's back I'll work on getting data passed from the phone application to the Picaxe controlling the motors. The first step will be to transmit the speed and direction the motors need to turn in. Then I'll probably write an app to use the pitch and roll reading to steer - like the tilt'n'drive Rovio app for the Iphone :) . Then add the compass, light sensor and sonar for autonomy. My tracked chassis is a bit too small so I'm going to extend it and add a few features.

Who am I and what am I doing here?

Noog — Sat, 23 Jan 2010 02:33:53 GMT

1) Father.
2) Geek.
3) Maker. (Although builders have had many names through out time, this is currently the popular choice.)

If it is broken I want to fix it.
If it is working I want to take it apart.
I may seek help with this one day. =P

I currently develop software during the day, spending the majority of my time in the C language. I have been known to dabble in any other language when the need arises, but I prefer to stick to my roots.

In the wee hours of the night, after my sanity has completely slipped away but before the morning light has called me to task, I have found a few hours here and there to work on my own projects. Some of them mechanical, some electrical and some software. It is here that my frankenstien creations come to life.

I have been interested in/addicted to robotics for quite a while now. Thankfully, I don`t believe this will end any time soon. My ongoing goal with robotics is simple: to tinker. Sometimes I will strive to make something that no one else has created before, other times I will simply add my own tweak to an existing design. If I can find the time to play with anything in this field, then I will have met my goal. If I can help others along the way, I will have exceeded my goal. If I have had fun doing any of this, then I will know that I am on the right track. =)

Let's build!

Noog

New user-help

Sean K — Mon, 18 Jan 2010 20:13:49 GMT

I need help, I just assembled my first beginner robot and have downloaded the Behavior Control Program to check the assembly, I get the right confirming message after the download but nothing happens when I disconect and press the start button to check the assembly. The robots led lights flash when downloading and I get the message, any suggestions? I also downloaded the demo and nothing happens when I press the mode button to get the start led to flash, then press start

Son of Voltron

badcommandorfilename — Sun, 17 Jan 2010 22:15:27 GMT

Recharging the batteries on the robobuilder RBC is not particularly hard - once you know how to enable it. Like a lot of things in robotics, it's 10% implementation and 90% RTFM.

Source code and more details are available here.

For more cool robot related activities, visit http://buildtherobot.blogspot.com/

A STAMP On My Hype

Nishi — Sat, 09 Jan 2010 22:47:10 GMT

After being so happy with BOE and it's user base, I'm so disappointed with the BASIC Stamp. For one it's limited variable space is fru-stra-ting, 26 bytes
just aint doing it.

... Anyway I doubt that's news to anyone.

Here's a vid of my BOE-Bot With Infra Red And Parallax Ultrasonic Ping Sensors roaming using a Ping))).bs2 Code by Droid Works I picked up from the files section here at Trossen :)

Nice script.

">" type="application/x-shockwave-flash" width="425" height="350">

I want to cover BOEs' back with a 6.V Solar Panel next and and replace the Duracell's with rechargeable batteries.

BOE, IR + Gripper

Nishi — Sat, 09 Jan 2010 02:59:59 GMT

So I bought a BOE-Bot and a few accessories the other day and basically stayed up working through the book over the following 2 days. I have to say it was a great fun and I've learned a lot. But I'm still lacking in know how.

Last night after successfully getting BOE roaming with an ultrasonic ping sensor, I decided to break open and start to construct the 'Gripper' kit, it took me about an hour to fully construct and attach it, by which time I started to look for some example code I could try to modify or use as a jump point, but I couldn't find any. (I must have become spoiled by the abundance of sample code and examples for the ping sensor and panning kit) So far Ive written one test program, and another very sloppy 'fetch and run' program neither are very practical, but the gripperIF may help as a starting point for someone (green in robotics) with a kit and no code like me.

Code:

' BOEGripperIF.bs2   SIMPLE BOE-Bot Gripper Test Using Infra Red Detection by Nishi

' {$STAMP BS2}
' {$PBASIC 2.5}

DEBUG "Program Running!"

irDetectLeft VAR Bit                                            ' Variable Declarations..
irDetectRight VAR Bit
pulseLeft VAR Word
pulseRight VAR Word
pulseGrip   VAR   Word

FREQOUT 4, 2000, 3000                                           ' Signal program start/reset..

DO                                                              ' Main Routine..

    FREQOUT 8, 1, 38500                                         ' Check IR Detectors..
    irDetectLeft = IN9
    FREQOUT 2, 1, 38500
    irDetectRight = IN0
                                                                ' Gripper Action..

          IF (irDetectLeft = 0) AND (irDetectRight = 0) THEN    'Gripper IF..

          pulseGrip = 1000                                           ' Close Gripper


          ELSE

          pulseGrip = 500                                            ' Open Gripper

          ENDIF
          PAUSE 20                                                   '(we PAUSE otherwise the servo hums,
                                                                     'gotta give the program time to run.)


        PULSOUT 13,pulseLeft                                    ' Apply the pulse..
        PULSOUT 12,pulseRight
        PULSOUT 14,pulseGrip

LOOP

But I'm still looking, so If you know of any practical gripper PBASIC code floating about please link me...

">" type="application/x-shockwave-flash" width="425" height="350">

">" type="application/x-shockwave-flash" width="425" height="350">

">" type="application/x-shockwave-flash" width="425" height="350">

It was only a matter of time...

badcommandorfilename — Thu, 07 Jan 2010 00:54:58 GMT

Nothing lasts forever.

I managed to drop my poor robobuilder off a table, damaging one of the servos. Fortunately, my kit came with a bunch of spare gears!

This post shows how to replace a damaged servo gear for a wCK servo.

More details can be found at http://buildtherobot.blogspot.com/

My First Robot.... And You Know What, It's A BOE-Bot.

Nishi — Wed, 06 Jan 2010 14:57:36 GMT

The other day I expressed my enthusiasm in building a bot, mainly as I've been fascinated with autonomous behaviour for a while and also because I've been learning to program in various languages and I wanted a more practical application for my R&D.

I asked around in the forums for advice on what I should start with and explained my budget was about �400. I had some quit in depth and really helpful responses so thank you to everyone that offered there input.

I had my heart set on getting a brat originally and had downloaded all of the documentation in the days leading up to the 4th of Jan when UPS and the rest of the postal service were set start moving again.

I joined the Lynxmotions forum and asked for some advice on what I was planning to buy, unfortunately I didn't get much help (other than one person who I feel wrong if I don't mention), still I couldn't leave without attracting a lil dose of arrogance and some minor hostility from lurkers hanging about (nothing makes you wanna drop a half a grand on line like be spoken to as if your an idiot.) So anyway that put me of a lil if only cus I live in the UK and the two distributors here hat stock BRATs are both more than a hundred miles away from me.

I personally didn't want to order a bunch of incompatible junk and as I say I'm green as they come.

Anyway lets me not be over dramatic, I was still aiming to buy a BRAT and I still love 'em. But a the same time I was being urged to consider whether I really was capable of getting the most out of a BRAT and if that's what would really bring the best out of me, I was also being reminded that realistically, on top of the hundred or so � I was looking to spend on a few sensors battery, charger, panning etc Id still more than
likely need to pay another 160 or so for some upgraded servos (since no one who had a videos posted seemed to be running the factory servos).

A good few people suggested I look into bioloid, and I do now get the hype. But I didn't have enough for a comprehensive kit, and even if I did it just seemed like to big a initial investment when I still wasn't sure I even wanted one. I looked in to the beginner kit and knew I didn't want that.

I really wanted to build a simple bot and then spend my real time working on the autonomous side of it's behaviour.. I wanted a bunch of sensors and tools that I could through trial ad error grasp a firmer understanding of where I wanted to do go next and how that was possible.

I had been looking at the BOE-Bot, but was put of initially because of the basic stamp and how limited the PBASIC language seemed, but it was all pretty theoretical and in the end spending �120 on my actual bot left me with a few hundred to buy sensors, gripper, voice recognition and a platform with more literature than I could swing a cat at... Well you get what I'm saying.

I ordered my BOE-bot on the 4th at about 10 am. From the 5th about 3 pm till now Ive been through the entire book exercise by exercise, I haven't even been sleep yet, (So please mind my grammar etc)

Its been a real introduction to robotics for me, circuits, microcontrollers and sensors. I've used photo resistors, infrared sensors, ultrasonic ping, all whist gaining a practical understanding of the individual components and setups.

With a wealth of material and vast user base your never short of support or resources. I recommend the BOE-Bot to Anyone looking for a viable gateway into robotics.

Thank You Parallax and thank you Trossen Robotics.

A few Vids....

First Steps:
http://s399.photobucket.com/albums/p...00106-0330.flv

Fine Tuning: (to the best of my ability)
http://s399.photobucket.com/albums/p...00106-0613.flv

Failing to film and guard bot on table (drop off detection: didn't seem to like the bright floors in my house)
http://s399.photobucket.com/albums/p...00106-1047.flv

PING: first scan
http://s399.photobucket.com/albums/p...00106-1306.flv

Sorry for the quality of everything, I was looking for my hd cam, but ended up using my blackberry. :/

Debugging robots for fun and profit

badcommandorfilename — Wed, 06 Jan 2010 07:37:27 GMT

More astute readers who followed the procedure in my last post may have noticed that it didn't actually work.

That's not to say that the instructions were wrong - those are still the steps to build the robobuilder C source files and upload them to the RBC. The problem was that the code which I linked doesn't actually do what it says on the side of the box, so to speak.

I've uploaded a fixed version of the code, so you can get back to hacking away.

More details can be found here.

You can read lots more about my robot project at http://buildtherobot.blogspot.com.