Putting robotics at your service™

Free shipping on orders over $200

New Quad Project... INPUT VERY WELCOME!

Print view Share :
Previous topicNext topic

Page 1 of 15 [ 218 posts ]

1, 2, 3, 4, 5 ... 15
Veteran ( offline )
Posts: 177
Posted: 2011-03-23 23:13 
Hi Guys,

I've been lurking around these forums for a couple months now and thought I'd finally jump in. I have learned so much from these forums and I'm amazed at all the very talented people here. Your input and ideas would be very welcome for a new project that I plan to undertake :D (<-- one of the things I learned: smilies are big).

I'll start with a little intro: I'm 24 and tinkering and robotics have always been a hobby of mine. I've done RC rovers, autonomous robots and designed and built a couple multi rotor RC aircrafts. I've got a BS in mechanical engineering and also a BS in physics; I couldn't stop at learning the way things worked, I had to know how to apply it too :roll: . I'm a machine designer by profession but focused on controls/mechatronics in college.

Now to describe a bit what I'd like to build.

Attachment:
Stand.jpg
Stand.jpg [ 137.61 KiB | Viewed 10394 times ]




So, obviously it's a quad. But, it also has driven tracks on each tibia(?). My inspiration for the design came from limitations. Walking and rolling both have their advantages and shortcomings so I figured "why not combine them?"(not a new idea, I know).

Another picture in tracked mode.

Attachment:
Roll.jpg
Roll.jpg [ 161.47 KiB | Viewed 10394 times ]


Another showing a rolling turn.

Attachment:
turn.png
turn.png [ 298 KiB | Viewed 10394 times ]


It is not meant to be overly powerful while it is in "tracked" mode, speed was more of the objective here. However, I chose tracks to still handle rough terrain. I figured that if there was a difficult obstacle in its way then walking might be a better option. At the moment, I do not plan to experiment with TA, which will most likely be necessary down the road. First, I just want to develop a stable platform that can walk and roll.

I'll describe where I'm at with this project and hopefully you guys can give me some pointers.

I am in the process of designing a board for my specific purpose using the propeller. It will have:
12 servo outputs
2 motor direction outputs (to control forward, backward, left and right)
4 motor speed outputs
4 encoder inputs (to precisely control motor speeds in turns)
6 radio inputs (all I can afford at the moment until the propeller 2 comes out. It will suffice right now: I currently only plan 4 operational modes)
Also, 2 outputs independent of the micro controller to control the rotation of the top turret and the the pitch of the camera mount.

I chose the propeller because parallel processing should help operation in many ways, especially when I start tinkering with TA. I plan to program this quad from scratch. I've seen that people have used the phoenix code for quads but I would like to optimize it to the propeller and would like to be intimately familiar with its controls. I've already worked out the math for all the inverse kinematics and am currently mulling around with a rule based sequencing engine based on COG, contact patch and direction. Mind you that all my work with this is "pseudo code", I'm a decent programmer and I'm learning SPIN so all my work is still on paper.

There are quite a few parts on here that I want to source from lynxmotion (SES parts, bearings, servos) and others from various places. There are also quite a few custom parts that I have designed and plan to build myself. I have a small CNC mill that is perfectly capable (check out zentoolworks.com , great stuff).

What I hope to hear from you guys are any loose bits of advice that you might have. Are there things I'm overlooking? Things I need to check before proceeding? Are there any adherent problems you see with my mechanical design?

I do not plan on buying any parts for at least a month. I will probably start by buying a prop dip to test with and all the parts for one leg so that I can measure myself and be sure that what I'm going to machine will work.

Thanks for anything guys,

Marc :lol:

P.S. I tried to integrate pictures into the thread. Sorry if its messed up, I'm new.


Last edited by Fish on 2011-03-24 10:02, edited 1 time in total.
You had everything in place properly, but you had BBcode disabled


User avatar
Guru ( offline )
Posts: 9257
Posted: 2011-03-24 10:23 
That is pretty cool! Can't wait to see how it comes out.

_________________
Jim Frye, the Robot Guy
http://www.lynxmotion.com
I've always tried to do my best...


User avatar
Guru ( offline )
Posts: 1130
Posted: 2011-03-24 11:35 
Pretty cool design! I do wonder how heavy the tibia part will be. The tracks and motor will make it pretty heavy I guess. What servos did you had in mind? About the electronics. There is nothing wrong with the propeller. But I guess you will get more support using the BAP since there are more users that use the Phoenix code on the BAP.

Keep us updated!

Xan

_________________
[b]Share, Use and Improve![/b]
Digging trough: [i]Theory of Applied Robotics: Kinematics, Dynamics, and Control[/i]
Xan's YouTube Channel
http://www.youtube.com/xanore


Veteran ( offline )
Posts: 177
Posted: 2011-03-24 12:39 
The motors for the tracks are actually quite light, only 10g and will still be able to deliver about 10oz of force to the ground each. I aim to at least have a .8 power to weight ratio. Yeah, thats low but speed was more important than power. It should have a top speed of around 4m/s. The material that the body of the tibia will be machined from (kind of an aerated resin 'composite') is also pretty light. I'm hoping to keep the tibia assembly weight under 4oz which, compared to what the tibia usually might way, is still pretty considerable.

As for servos I was thinking of going with the 5685MH's for all of them. The torque is nice but I like that they accept 7.4V. I plan to use a 2 cell Lipo and would like to run my motors and servos off the same header without having to worry about regulating the voltage.


Veteran ( offline )
Posts: 177
Posted: 2011-03-24 12:42 
Oh, and thanks, Fish, for helping me out!


User avatar
Guru ( offline )
Posts: 4125
Posted: 2011-03-24 13:42 
That's pretty wild!

Anxious to see how it comes out!

Alan KM6VV

_________________
Visit:
http://groups.yahoo.com/group/SherlineCNC/
http://tech.groups.yahoo.com/group/HexapodRobotIK/


User avatar
Novice ( offline )
Posts: 59
Posted: 2011-03-24 15:09 
That looks awesome! Always wanted to see a hybrid walker with tracks. Reminds me of this Lego toy I saw a while a go.
Attachment:
File comment: Lego Bionicle Skorpio XV-1
skopio_xv-1.jpg
skopio_xv-1.jpg [ 177.2 KiB | Viewed 10332 times ]


I'm curious, what propeller board are you planning to use? It may be worth looking at the Propeller Servo Controller if you're after something small, although i'm not sure if it has enough IO for what you need. It does support 16 servos out the box though, and can be programmed over USB to use its 6 free COG's (i've not tried this yet).

I look forward to your progress :D.

-1

_________________
BSc Artificial Intelligence & Robotics | PhD in progress
[b]Project Infuser Progress|[/b]
http://www.lynxmotion.net/viewtopic.php?f=20&t=8225
Hardware - [color=#00BF00]95% Complete[/color] |
Software - [color=#FF9F00]40% Complete[/color]


Veteran ( offline )
Posts: 177
Posted: 2011-03-24 17:10 
Thanks, Zodius. Man, I love lego.

I considered using the propeller servo controller but in the end it didn't fit with what I needed. I considered all of the premade propeller boards actually and they either didn't have enough of what I needed or had extra features that got in the way of other things. Designing a board really isn't that bad though, there is a site will custom print double layer PCB's for a good price. It'll be about $60 for three, three being the minimum order. It will be programmable by USB, have just enough IO's and will have the motor drivers built into it.


User avatar
Guru ( offline )
Posts: 1987
Posted: 2011-03-25 09:17 
@Marcham:
Did you take a look at Arduino board ?

This one have 2 motor driver build in:

I have this one:


I realy like your project... realy interesting !

If you want to take a look:
http://www.lynxmotion.net/viewtopic.php?f=8&t=6951

_________________
Eric Nantel - DiaLFonZo
UAV/Drones Specialist

RobotShop inc.
Putting Robotics at your service!™
www.robotshop.com

Lynxmotion
Imagine it. Build it. Control it.™
http://www.Lynxmotion.com


Veteran ( offline )
Posts: 177
Posted: 2011-03-25 11:24 
Thanks, DiaLFonZo.

I've looked into the arduino a little bit. However, I'm very curious to see how the multi processor propeller will handle walking. Having 8 routines running at once can be very useful.

I do like the look of the board you have, 54I/O's!

Your quad is a thread that I've been following, it looks great! I've also followed your DIY remote thread, I'm very impressed. Your solidworks skills are very good too and the renders are always a nice touch.


User avatar
Guru ( offline )
Posts: 1987
Posted: 2011-03-25 11:29 
Marcham,
Do you use SolidWork ?

DFRobot have a shield for the Arduino i have. I have it.. ;)
Usefull when trying to plug multiple servo, sensors.. etc


I am on the process to buy a small CNC table... maybe i will be able to cut the "LittleQuadra" myself.

_________________
Eric Nantel - DiaLFonZo
UAV/Drones Specialist

RobotShop inc.
Putting Robotics at your service!™
www.robotshop.com

Lynxmotion
Imagine it. Build it. Control it.™
http://www.Lynxmotion.com


Veteran ( offline )
Posts: 177
Posted: 2011-03-25 12:15 
Yep, I've been using solidworks for about 4 years. I've got 2009 on my home computer and I use 2011 at work. Great Software.

If you are looking into buying a small CNC then check this out.

http://www.zentoolworks.com/product_inf ... cts_id=112

This is the exact machine that I have. It includes everything you need, mechanically and electrically, to get your mill up and running. The best part is it's under $500! They didn't have this package when I bought mine and I ended up spending around $700 get it up and running :x .
If 7x7 isn't big enough for you then you can also buy an extension kit that increases the table size to 7x12. I have to say that I'm really satisfied with it's performance and I love having a desktop CNC machine. Also, there is plenty of free software on the internet to support it, and a great google group for help.

It can cut aluminum, not very fast but it can do it, also plastics (but you need to be careful about melting) and wood. But the material I most recommend is this

http://alumilite.com/ProdDetail.cfm?Category=Casting%20Resins&Name=Alumilite%20Regular

This stuff seriously machines like a dream. Not a bit of melting at any speed, good strength, you can mix in fillers to get different material properties. I also like that I can cast any sized block that I need, just make quick mold calculate volume and you can make plates or block of any shape or size. It is what I plan to use for the majority of the parts on my project.


User avatar
Guru ( offline )
Posts: 1987
Posted: 2011-03-25 12:40 
Quote:
This stuff seriously machines like a dream. Not a bit of melting at any speed, good strength, you can mix in fillers to get different material properties. I also like that I can cast any sized block that I need, just make quick mold calculate volume and you can make plates or block of any shape or size. It is what I plan to use for the majority of the parts on my project.


Yes that's a good idea. I wasn't thinking about that.. But if i buy the CNC would be easy to use.. ;)

_________________
Eric Nantel - DiaLFonZo
UAV/Drones Specialist

RobotShop inc.
Putting Robotics at your service!™
www.robotshop.com

Lynxmotion
Imagine it. Build it. Control it.™
http://www.Lynxmotion.com


Veteran ( offline )
Posts: 177
Posted: 2011-03-25 13:07 
It's not bad. The setup is the hardest part. You'd need to be sure all your corners are square and there is no binding on any of the axes, you'd need to be very careful while assembling. Then, wiring it up wasn't so bad either, I just needed to extend some of the leads. There is tons of online support for getting the thing connected and running.

There was probably about a week of struggling and learning for me (I had zero CNC experience). Then once I got the hang of it it was easy.

Plus, I'll be around here for a while so if you ever run into problems my experience could probably help.

One thing I did forget to mention is that you need a computer with a serial port. It is very difficult to get the stepper driver to run with USB. The good thing is that the CNC software isn't very resource hungry so you could buy an old cheepo computer to dedicate to the mill. I use an old laptop.


Veteran ( offline )
Posts: 177
Posted: 2011-03-27 20:56 
Little bit of an update.

I'm ordering some propeller components to throw on the breadboard. I want to be sure that what I'm laying out on a PCB is going to work. I'm picking up a propeller DIP, a usb port, some eeprom, a 5Mhz crystal, a few buffers, a 3.3 and 5V reg, and some pin headers.

I'm going to wait a few days to order; I'm moving into an apartment this coming sunday and don't want the parts to be delivered there before I move in. If you guys have any advice on other things to order or ideas on things I should try to incorporate it would be very welcome.

In other news: I've changed the design a little. I just made the servos for the coxa and femur parallel instead of perpendicular. I've also started looking into making a wireless link between the top turret and the rest of the body. I thought it would be cool to be able to continuously rotate the body while keeping the turret stationary. I've thought about using infrared for this. I figured I could place few IR blasters on the base and have a receiver in the top with its own power supply and microcontroller to decode the position signal and control the servo.


1, 2, 3, 4, 5 ... 15

All times are UTC - 5 hours [ DST ]. It is currently 2014-11-01 05:30
Feedback Form
Feedback Form