A community project for the summer perhaps?

[Admin]

What everyone has been waiting for . . . word from Admin. To be honest, I was just really busy, and 200 replies in a single post kinda scared me away for awhile . . .

I have several thoughts concerning:
1) community project
2) leadership
3) modularity
4) funding
5) project goal

1) So lets start from the beginning. I've been dreaming of a SoR community project for about 6 months now, but felt that you guys aren't really ready yet. For example, most of you have only built one robot . . . So I figured I'd dream for another 6 months until you guys get more experienced then start up a topic similar to this. But I guess you guys have the motivation now so my answer is: OK! Yes!

2) Now, I am a firm believer in what I call a 'functional anarchy'. Meaning you put a whole bunch of well meaning motivated intelligent people together, and you don't need a leader. However, many of you have issues on keeping focus and so I'll make some leadership like decisions. I don't really like being a leader, its hard work, but I'm fully capable of keeping you guys on target . . .

3) So now to my next point. As mentioned earlier, modularity isn't an option - it is required for this project to work. We all need to agree on standards of dimensions, voltages, communication interfaces, etc. A few of you have already started discussing this and I think its on track so I won't say any more than I 100% support modularity and standarization.

4) Money is a great motivator. Long story short, I will reach into my pocket and help bank roll this project. Yeap, I will pay for it! But money isn't free . . . sub-projects I financially support must have a great tutorial written to go with it. I also will create mini-contests for money - for example, the best designed module gets $50 from me or something like that. I am also thinking I'll reimburse anyone who makes a great module that uses my Axon (for the full cost of it). I'd have to think more about this so don't quote me on it yet. Opinions welcome.

5) Now this is the most important point I want to make. What if the goal of this project wasn't to do something like make a one-off robot, but instead create the first open source robotics platform in the world? If we all pull together, we can change the world of robotics like open source software has done to PCs. I think SoR is in a position that no one else is to make this happen. And what would we need to do to make this happen?

a) create a set of robot standards
b) develop modules
c) write tutorials

Obvious eh? We already have made progress on all three points.

So here is the defined mission I am proposing:
A) We start a thread on robot standards (there is a lot to say on this, so lets leave it to a new thread I just created)

B) We develop modules. Design a module, write a tutorial on it, you can even develop it into a product and sell on your own if you want. The only condition is that your product is open source and follows our agreed standards.

C) Everyone help grow the community, and document your robots! We are all better off and can build robots faster if we work together. Try building a robot without going to SoR if you don't believe me


Now commenting on what everyone else has said . . .

we all specialized in certain Fields. dunk with his internet control, airman00 with his butler bot voice recognition. but we are all saying we need one leader and i think that is true try admin for example if he was the leader he would see over the entire project. then we should have group leaders (programming, contruction)

cooldog, I entirely agree. If we all specialize on a particular area that we are good at, we can focus more and get more done.

Also we will need a open source programming area as well. Where functions for each module on a pic and avr are already made and you just download the code to go with each one and tie them together with your own code. It has potential but needs to be managed and a good system developed to manage the files and projects.

I can offer project space, but someone else would have to manage that.

does anyone think we need a 3.3v supply? personally i think 5v is a better choice and people can regulate down to 3.3v if they need on a particular board.

dunk, my Axon has a 3.3V regulator built in that can supply up to ~76mA if I remember correctly . . .

licencing. who owns this? how do we stop someone else from patenting out work? i think we need a template Open Source type licence  that anyone can copy into their project.

dunk, I think whoever develops a module owns rights to it. we should follow the open source rules similar to software . . . thoughts anyone?

everyone alright with Eagle for board schematics?

dunk, yesssssssss!

i personally am with the idea of a subforum for the SOR project. it'll be alot easier to find threads about it, than having to look in every section for the SOR PROJECT heading.

Here is where my leadership comes into play . . . for all project posts, make sure this is in the subject line:
SoR Project: [title here]
If it gets crowded, I'll create a new section.

Ok just to add a little organization to the coming projects I created a list of potential modules. This will be added to and modified over time to include who is working on what module so that people don't do more then one without combining their work.

http://docs.google.com/Doc?docid=ddgqp9wc_0gpbcwmgm&hl=en

asellith, great idea!

About making plastic casings for parts, perhaps that should be a project/tutorial on its own?

And lastly, if we were to use my Axon, I'd have to standarize the software for modularity. I'll make this a priority and get the info out to you guys as soon as I can.

[dunk]

ok, finally got round to packaging this up.
it's not really meant as a finished module (although it does bring all pins on an ATmega8 to header pins).
it's just meant as a working prototype.

http://www.societyofrobots.com/member_tutorials/node/69

it fully implements the master i2c node described in our design doc and conforms to the standards in our physical attributes doc.

it's not pretty though and doesnt do anything more on it's own than blink LEDs yet...
i intend to add servo control to the firmware soon.
it can already act as an interface between my laptop and the i2c bus on my robot though.

i made the worlds dullest youtube vid of me changing LEDs on it if anyone is interested:
http://www.youtube.com/watch?v=S8wv2yDrX7Q

i'll work on documenting my motor controllers next so there is some sample ATtiny i2c slave code for people to work from.


dunk.

[Trumpkin]

good work dunk! Can you reccomend something I can do for the robot that's very simple? I could make a $50 robot microcontroller if we need it  .

[Fredrik Andersson]

I have a suggestion for the board sizes and stacking possibility. If say we have a stack of 76x76 boards and want to stack a 38x38 or even several on top of the 76x76 board the mounting holes wouldn't fit the 38x38. If we as a standard put four holes in the middle of all the 76x76 boards, it would be possible to stack four 38x38 boards on one 76x76. Then we'll do the respectively needed with 76x38 boards.

[hazzer123]

I support this idea, it makes sense to put limitations on the position of the mounting holes more than the actual size of the PCB.
This will make sure that the boards are stackable.

[Ro-Bot-X]

The smaller boards may use only 2 holes in opposite corners.

[dunk]

I have a suggestion for the board sizes and stacking possibility. If say we have a stack of 76x76 boards and want to stack a 38x38 or even several on top of the 76x76 board the mounting holes wouldn't fit the 38x38. If we as a standard put four holes in the middle of all the 76x76 boards, it would be possible to stack four 38x38 boards on one 76x76. Then we'll do the respectively needed with 76x38 boards.

it's a nice idea in theory but very wasteful on board space having screw holes in the middle...
we could make a carrier board that bolts to a 76x76 board that has bolts and bus pins for 4 38x38 boards?
it would waste some height but keep the foot print small.

another thing we have not addressed is what sort of connector to use for boards stacked on top of each other.
with 2 boards we can use a header pin on the top board and a socket on the bottom but if we want to stack 3 or more boards we need some sort of stacking connector type.
i've seen them in use on PC104 boards but have not seen one with only 5 pins....

The smaller boards may use only 2 holes in opposite corners.

yes indeed.  i think we should specify on our template which 2 are fixed and which are optional.


dunk.

[Ro-Bot-X]

Here it is the

"SoR Standard Dual 1 Amp DC Motors Controller"

[Admin]

I should probably comment on the motor driver . . .

#2 for Right-Motor doesn't seem to have anything connecting to it.

Also, the trace width going to the motors is waaaaay too thin. There is lots of room for more red tracing, fill it up!

The trace separation distance for the lower blue plane, on some traces, looks really small . . .

For the programmer connector, put a big white triangle where pin 1 should be. This way a person won't connect it in reverse.

For the big 1000uF cap, it looks a bit small. Make sure the cap is rated for at least 24V.

Lastly, put it into 3D, I'm curious how it will look!

[Admin]

Oh forgot to ask, whats the price of it all?

[Ro-Bot-X]

I should probably comment on the motor driver . . .

#2 for Right-Motor doesn't seem to have anything connecting to it.

Well, it is connected, but can't really see it in the picture atached. I'll try to make a better picture (PNG or something...)

Also, the trace width going to the motors is waaaaay too thin. There is lots of room for more red tracing, fill it up!

OK. I enlarged it to 32 mils, but I have checked the trace width calculator and it says 12 mils for 1 amp at 1 oz/ft^2 copper thickness.

The trace separation distance for the lower blue plane, on some traces, looks really small . . .

Same thing with the picture quality. The spacing is 50 mils. Edit: you were right, changed the Isolate to 16 mils.

For the programmer connector, put a big white triangle where pin 1 should be. This way a person won't connect it in reverse.

Did that, thanks.

For the big 1000uF cap, it looks a bit small. Make sure the cap is rated for at least 24V.

Actually, this capacitor should be layed down on the board. I have created enough space around it for that.

Lastly, put it into 3D, I'm curious how it will look!

I'll do that. Thanks for your coments!

Oh, by the way, I didn't check prices for the parts. But the most expensive is the PCB. If done in quantities, it would worth to build it as a module. I am not a fan of making my own boards. I did that, had good results, but I want my boards to have solder mask and silk screen and most important, metalised holes.

I am also working on a small Dual Modified Servo Controller, similar to this one, and a Dual Stepper Controller. I intend to finish them before I get to Canada.

All these controllers will be controlled by the Main MCU module with the same commands. So swapping motors and controllers won't change a single line of code in the Main MCU. I will need help on writing the firmware for them. When the time comes, I'll open a new thread.

[Admin]

I enlarged it to 32 mils, but I have checked the trace width calculator and it says 12 mils for 1 amp at 1 oz/ft^2 copper thickness.

Larger traces won't hurt. So wait, this only supplies 1A? A bit low . . . I didn't look up the IC's . . . but what if a user wants to stack an extra driver IC to double current? Would be good to have tracing for it.

Oh, by the way, I didn't check prices for the parts. But the most expensive is the PCB. If done in quantities, it would worth to build it as a module. I am not a fan of making my own boards. I did that, had good results, but I want my boards to have solder mask and silk screen and most important, metalised holes.

The reason why I ask is because you can buy drivers with comparable features for about $60, so this would have to be much less to be worthwhile for people.

[Ro-Bot-X]

I have looked on DigiKey for parts prices and I added up parts cost (price per 1 piece). All the parts cost about $20 plus the price of the board, that can be $15 (5 pieces order) or a lot less if ordered 100 pieces (about $4-5 a piece). Plus shipping from 2 places, plus handling, plus a little for design... I gues the price for a one piece controller would be around $50. Ordered in hundreds, price can drop to $35 a piece.

[Rebelgium]

I suggest there should also be a motor driver using the L298, this is a dual H bridge that can supply 2A continuously.

[Ro-Bot-X]

I suggest there should also be a motor driver using the L298, this is a dual H bridge that can supply 2A continuously.

Yes, that would be the Medium Power DC Motor Controller. I'll do that too. The firmware doesn't care the power the H-bridge delivers. Same will work for Stepper Motors.

[airman00]

so when are we building the motor drivers?

I volunteer myself for whatever service needs be done.

[Ro-Bot-X]

Well, I don't know. There is still room for improvement. For instance, I am looking to add a voltage divider to be able to measure the motor battery voltage.

Here is what I'm going to do:
- a Dual Modified Servo Controller with Encoders // I need this ASAP for my project
- a Dual Medium Power DC Motor Controller
- a Dual Low Power Stepper Controller
- a Dual Medium Power Stepper Controller

I'll try to write some code for them, but I need someone to write the code properly in C.

I need to know some things: do we want PID controll in the firmware? If yes, how do we tweak the values? Should I add multiturn pots to the boards or are we going to do it in the firmware?

[Admin]

After Ro-Bot-X posts a 3D image of it plus dimensions, airman00 can you design up a google sketchup of it (since you have started the SoR Sketchup Library)?

[airman00]

After Ro-Bot-X posts a 3D image of it plus dimensions, airman00 can you design up a google sketchup of it (since you have started the SoR Sketchup Library)?

sure I just need exact dimensions of parts

[Ro-Bot-X]

I have completed the 3D rendering of the board. For parts size, please check the zipped Eagle files attached.

[airman00]

OK ill make a sketchup of it when I get the chance..

[Rebelgium]

http://www-micrel.deis.unibo.it/~augusto/doc/Com_i2c.pdf

I found this doc when searching info on I²C, on page 24 I found this text:

The I 2 C-bus specification
Trasmissione Dati-I2C
Wiring pattern of the bus lines
In general, the wiring must be so chosen that crosstalk and interference
to/from the bus lines is minimized. The bus lines are most susceptible to
crosstalk and interference at the HIGH level because of the relatively high
impedance of the pull-up devices.
If the length of the bus lines on a PCB or ribbon cable exceeds 10 cm and
includes the VDD and VSS lines, the wiring pattern must be:
SDA
VDD
VSS
SCL
If only the VSS line is included, the wiring pattern must be:
SDA
VSS
SCL
These wiring patterns also result in identical capacitive loads for the SDA and
SCL lines. The VSS and VDD lines can be omitted if a PCB with a VSS
and/or VDD layer is used.

It might be a good idea to change the order of the wires since now they're
Vdd
VSS
SDA
SCL
NC

Maybe it's not worth it because you think there's not that much chance of having crosstalk.

[Asellith]

I had an idea for a cellular module to control a robot. I am going to develop the board and work on it as a personal project that I might try to sell later on. On that note I can make a special version for the SOR project if there is enough interest. This will be a GSM cellphone that means it will work on AT&T,T-mobile and some others. The module I am using will have Java embedded in it so it can be customized. I will also have several perifirials so that this can be used as a stand alone robot controller that can report back to a computer through the internet. My plans are to make a simple interface to control it and change settings but leave it open for customization by the end user. It will have cool things like RS-232 and USB onboard. However the module does not support I2C. But I think I can find a way to convert the uart to I2C and make the controller act as an I2C master that way it can talk to the other modules on the bus. Right now I have 2 uarts from the module and 1 is used for RS-232 the other can be converted to I2C for this project if you all are interested. If not I'm leaving it as a UART. Unless we want to go ahead and build a UART to I2C module then that would work as well.

Jonathan Bowen

[hazzer123]

I just received an alphanumeric 16x2 display and a 128x64 pixel mono graphic display in the post, so im ready to start making the display module.

How is everyone else doing? The discussion seems to have ground to a halt. Has the progress stopped aswell?

[airman00]

Have we all just given up on this?


C'mon guys!
Lets get some progress done! 

[dunk]

the microcontroller boards on my current robot all run firmware that meets the design specs.
i had already built most of the boards before we settled on the hardware specs though so only one of them meets the board layout design rules.
i really should redo the board layouts but don't know when i'll get round to it.
http://mrdunk.googlepages.com/electronics

anyone else got anywhere with this?


dunk.

[Trumpkin]

Everybody, post what you are doing!

[Asellith]

Ok getting out the paddles to try and get some life into this project....

CLEAR!!!!

ok

Starting work on the Servo Controller module. Going to have to do research into how many servos I can cram onto one board because I know some of you guys like those hexapods and other walkers that require lots of servos. My project only needs like 7 or 8 but why not go the extra distance and design a good board.

Updated potential module list to indicate that I am working on it.

[airman00]

anything finished yet?

[Asellith]

Still in the design phase for the Servo Module. I have the concept and am working on a plan for the code. I want to design the hardware really good with a lot of room for expansion code wise. Then I'll get a base code that works and add a bunch of bells and whistles to the code later.

Right now I'm waiting for work to slow down so I can get some work done. I officially take a new position this Friday and will be on call for the rest of my life. But they tell me the radio station only calls you 4 or 5 times a week and about 3 times a month you actually have to go the the station or the transmitter. So hope to see results soon.

Am I the only one working on a module?

I saw some power designs posted earlier but nothing for months now.