Society of Robots - Robot Forum

General Misc => Misc => Topic started by: SmAsH on April 17, 2009, 09:01:39 AM

Title: SOR community project?
Post by: SmAsH on April 17, 2009, 09:01:39 AM
hey guys this has been bugging me for a while and thought i would just ask, did that thing ever go ahead? this thing http://www.societyofrobots.com/robotforum/index.php?topic=2765.0 (http://www.societyofrobots.com/robotforum/index.php?topic=2765.0) has anyone done anything towards it, and is it still going ahead or has it since died? it just seemed that everyone who was 'involved' in this project was really excited and had so many ideas that i think were excellent! is this idea dead?
Title: Re: SOR summer project?
Post by: superchiku on April 17, 2009, 09:30:49 AM
DEAD
Title: Re: SOR summer project?
Post by: Asellith on April 17, 2009, 12:17:52 PM
If my health ever gets better I have plans for my own modular system based on the idea from that project. However I think that project is dead. Feel free to design a module to the specs. Maybe if we get a few modules tested and working we can revive it. The google docs for all that are still up.
Title: Re: SOR summer project?
Post by: SmAsH on April 17, 2009, 04:32:07 PM
well i will certainly look into doing that as this year im not going to be as busy as next so i will have some free time.
i will definitely look into doing either a motor driver module or a stepper motor driver module.
i just have two questions, what are the board sizes, i read somewhere in the thread 1.5X1.5"?
and what are the screw sizes? for the standoffs? i have a week and a half of nothing before i go back to school so i can start designing a motor driver module now.
Title: Re: SOR summer project?
Post by: Weird Fishes on April 17, 2009, 06:23:09 PM
I read over parts of the thread (IE: p1 and p8 :P) and it looks really cool! I'd love to help if it gets going again. I'm going to go read over the whole thread actually. I love these open source projects. I've been wanting to be a part of one for a while now. Best to do it before I go off to Uni next year I suppose... from what I hear the CE program there (University of Waterloo for any of you Canadians out there) is pretty intense.
Title: Re: SOR summer project?
Post by: SmAsH on April 17, 2009, 07:08:33 PM
im going to start designing it today, if i can get my eagle working. can anyone help me?
my eagle is loading up fine but when i click add component and that window pops up there are no components? this happens in the brd editor and sch editor? im wiping everything off my pc to do with eagle and going to try installing again soon.
edit: ok its fine now, turns out my computer added some files when i updated it that i deleted and now the libraries are showing up.
Title: Re: SOR summer project?
Post by: dellagd on April 17, 2009, 07:44:22 PM
I'm pretty good at C++ computer programming, so I could help with code.
Title: Re: SOR summer project?
Post by: SmAsH on April 17, 2009, 08:09:02 PM
just a question, do motor drivers need 2 inputs per motor?
Title: Re: SOR summer project?
Post by: galannthegreat on April 17, 2009, 08:21:37 PM
yes, one for high and one for low (back and forth alternating)
Title: Re: SOR summer project?
Post by: SmAsH on April 17, 2009, 08:34:24 PM
awesome, thanks.
Title: Re: SOR summer project?
Post by: galannthegreat on April 17, 2009, 08:37:06 PM
no problem, glad to be of help
Title: Re: SOR summer project?
Post by: Asellith on April 17, 2009, 08:39:04 PM
All the defined specs are here.

http://docs.google.com/Doc?id=ddp2r5j8_24hhf75vck&hl=en (http://docs.google.com/Doc?id=ddp2r5j8_24hhf75vck&hl=en)

Also that links to the other pages that where worked up. I think Dunk was the one who defined most of the specs.
Title: Re: SOR summer project?
Post by: SmAsH on April 17, 2009, 09:03:59 PM
thanks asellith, and yes dunk did come up with most of the specs.
edit: i have attached what i have so far. could you guys tell me anything else you think it may need in it? leds, caps, diodes? i know i need to put some caps in, i will do that later on. right now i need some lunch.8)
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 18, 2009, 05:26:17 AM
What is the schematic? The layout does not tell much...
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 05:27:59 AM
updates for you guys to check out, less jumpers, less space and ver1.7 needs +5V from somewhere else as it doesn't have an on board regulator as dunk said the modules can get +5V from the main board.. i have attached versions 1.5, 1.6 and 1.7
@ro-bot-x i have attached a crappy schematic, i will make a better one soon.
Title: Re: SOR community project?
Post by: dellagd on April 18, 2009, 06:03:10 AM
just a question SmAsH,
what Ideas exactly are we taking and using from here
http://www.societyofrobots.com/robotforum/index.php?topic=2765.0 (http://www.societyofrobots.com/robotforum/index.php?topic=2765.0)
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 06:09:58 AM
i would say the idea to have a set of modules that are open source and easy to interface with, the standards that dunk put on his pages: http://docs.google.com/Doc?docid=ddgqp9wc_0gpbcwmgm&hl=en (http://docs.google.com/Doc?docid=ddgqp9wc_0gpbcwmgm&hl=en)
and an open source community project
there are a few others which i cant think of right now.
1.8 attached
Title: Re: SOR community project?
Post by: Webbot on April 18, 2009, 11:22:57 AM
It might be worth looking at my 2Amp DC Motor Controller in the Motor Controller member tutorials. This uses a L298 but has some extra components so that you can use two wires to control each motor with variable speed, fwd and backward, and brake. The design is very similar to the Solarbotics L298 motor controller. You may want to redo my board layout - it could be made much smaller - especially if you use a dbl sided pcb.

My C++ library in the member tutorials has software classes to control it and my 'imminent' C library has support for it as well.

Can'[t give direct links at the mo as the member tutorials section of SoR seems to be down.
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 03:46:47 PM
well, i will take a look at that webbot. thanks for that ;D and the tuts are still down :-\
and a question towards eagle, does anyone know how to get a black background instead of white? i used to have it then i reinstalled eagle and i have a white one?
Title: Re: SOR community project?
Post by: Asellith on April 18, 2009, 06:31:31 PM
Don't forget the heatsinks. Not sure if you need a big enough one to justify mount holes on the board but it something you need to think about in the early stages and you might have already done that but I just wanted to check.
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 06:36:14 PM
lol i forgot about those! ive never really used em before but i will order some when i get the components for the boards. they are only needed if pushing the chip to its max right? like drawing close to 2 amps?
EDIT: heatsink added and new schematic added with the vs and vcc pins not mixed up :P
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 18, 2009, 08:00:14 PM
Smash, it will be helpfull to Export the schematic and board to image (from File menu) since not everyone wants to load them up in Eagle. Just to take a look and see what's going on, a picture is more than enough. When the project is finalised, you can attach the brd and sch files. Thanks!
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 08:04:21 PM
sure thing ro-bot-x, will attach a few images now.
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 18, 2009, 08:29:49 PM
Ok, now I see what you're up to. I have a sugestion: try to have all inputs on one side of the board along with the regulated 5V input (use regular pin headers) and the outputs on the other side of the board along with motor power. Perhaps rotate the L298 chip 90 degrees? See what you can come up with!
Title: Re: SOR community project?
Post by: SmAsH on April 18, 2009, 09:55:18 PM
awesome, thanks ro-bot-x! do you think i should have an on board regulator?
EDIT: i was at it all day today but i couldn't find a way to have all outputs on one side without having heaps of jumper cables everywhere, so i think ill keep with the current design although i did make it smaller horizontally by moving some stuff inwards.
Title: Re: SOR community project?
Post by: dellagd on April 19, 2009, 05:57:47 AM
I think we should have a on-board regulator
Title: Re: SOR community project?
Post by: SmAsH on April 19, 2009, 06:04:18 AM
I think we should have a on-board regulator
yes i do agree and it would be a sinch to add a regulator but i was just confued by this post by dunk on his page.
Quote from: Dunk
Standard power bus should be 5V DC.
Modules operating at less than 5V should have on-board voltage regulator and buffer circuitry to allow 5V I/O.
Modules accepting greater than 5V (eg. 12V) which contain regulators to supply 5V to circuitry on board should also be able to supply other boards with 5V power to a minimum of 200mA.
Modules accepting greater than 5V (eg. 12V) to power high voltage components (eg, DC motor drivers) do not need regulators for modules 5V circuitry. They may draw 5V from another module.
but why not, i may as well chuck in a regulator :-\
Title: Re: SOR community project?
Post by: Webbot on April 19, 2009, 01:10:09 PM
Worth checking out Solarbotics board: http://www.solarbotics.com/products/k_cmd/ (http://www.solarbotics.com/products/k_cmd/)
The Resources tab has schematics etc. Board layout may give you some ideas.
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 19, 2009, 03:15:03 PM
If you want to add a regulator, use the 7805L at 500mA or even less, that uses a regular transistor size capsule, it saves some space. Besides, if you power just the L298 with it, makes no sense to use a full amp regulator.

I see now that you want a single layer board, that will work out better with your present layout.
Title: Re: SOR community project?
Post by: SmAsH on April 19, 2009, 04:10:46 PM
okee dokee, ill add a small regulator somewhere on the board.
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 09:10:19 AM
one thing, i know the board size is supposed to be multiples of 1.5" and the screws 4-40 but was there any specs on how far from the edges? i have read several posts about suggestions but nothing suggested, i heard one post from admin about having it cut "on" the corner of the board to make a C instead of the o. can anyone clear this up for me?
Title: Re: SOR community project?
Post by: dellagd on April 20, 2009, 01:51:41 PM
do you think we should send a PM to admin about reviving this topic?
Title: Re: SOR community project?
Post by: Asellith on April 20, 2009, 02:59:20 PM
Do Admin's idea. With the half holes on the sides you can support 1 board or several without adding a bunch of standoffs to screw them down to. Also opens more board space for more complicated designs.

Just some notes.... Does that chip of yours comply with the I2C protocol that is defined? Or can it be programmed to comply? if not there needs to be a front end MCU or something to make the chip comply. I'd look it up but my brain is fried because of work. (I still don't understand how 4 pieces of equipment can fail at the same time)

Also... Make sure your PCB traces are sized right for the power output. Your power traces need to be big enough to handle the current load. They might be just in the standard your using but you need to at least look at into to make sure. You don't want to end up blowing a trace on the board.


Good work so far. Keep up the enthusiasm and posting. Get one or two modules done and maybe we can get this off the ground again.

Title: Re: SOR community project?
Post by: dellagd on April 20, 2009, 03:15:32 PM
can someone please explain/clarify how this will work?
http://docs.google.com/Doc?docid=ddp2r5j8_21dgt72qgz&hl=en (http://docs.google.com/Doc?docid=ddp2r5j8_21dgt72qgz&hl=en)
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 04:25:09 PM
@asellith: i can see how that idea is great but are those holes not weaker? i have no facts about it but the just look weaker as they only have 3/4 the support as the other holes have, and one more thing that was discussed in the other thread was about if we had a 76-76mm (3"^2) and we wanted to mount four of the smaller boards on top we might run into problems with the screw heads bumping into each other, do you think this would be a problem? and lastly, is it a problem if my module doesn't have i2c? i cant really see a use for a motor controller having i2c? i thought it was just going to be plug'n'play type thing?

@dellagd: do you know what i2c is? i don't really understand it either much ;D but this may help a bit http://www.societyofrobots.com/member_tutorials/node/35 (http://www.societyofrobots.com/member_tutorials/node/35)
Title: Re: SOR community project?
Post by: dellagd on April 20, 2009, 04:48:26 PM
Quote
i cant really see a use for a motor controller having i2c? i thought it was just going to be plug'n'play type thing?

if we are using a axon do we really need i2c ( now that I have a rough understanding of it ( yea for me  :P ) )
we have 50 i/o pins.
how much sensors/servos will we really need?
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 04:51:04 PM
Quote
i cant really see a use for a motor controller having i2c? i thought it was just going to be plug'n'play type thing?

if we are using a axon do we really need i2c ( now that I have a rough understanding of it ( yea for me  :P ) )
we have 50 i/o pins.
how much sensors/servos will we really need?
thats what i thought but some people like i2c for its features...
Title: Re: SOR community project?
Post by: dellagd on April 20, 2009, 04:55:02 PM
smash can you make a sketch in microsoft paint or something quick so I ( and probably some others on this topic ) can get a rough Idea on what this robot will look like?
will it be a humanoid?
wheeled?
2 ft tall?
5 ft tall?
Title: Re: SOR community project?
Post by: Asellith on April 20, 2009, 05:17:22 PM
The point of the module system is to just connect everything to a I2C bus. So each module is a stand alone device. The main controller unit will not have any output but I2C. Now there was discussion of using the Axon. So we could change that because the Axon is so cool. Admin will through more of his weight into the idea if we center it around the Axon.

Now if we want to change to be centered around the axon then we need to possible change the entire design. I don't know the actual dimensions of the axon but maybe we should get Admin behind helping us design a module into the new axon 2 design. Using it as a base and build off it. Unless it has really weird board dimensions. 

Or we could just say the axon is the controller and the modules all link together in a weird tower/grid pattern separate and you just connect the two with jumpers. Originally the axon didn't exist or was just coming out so we didn't base it on having such a cool base. So we wanted the modules to link together easily and just a simple connection between each module I.E. a 6 pin connector keyed to not hurt anything if plugged in backwards. That would transfer 5 volts power and the I2C signal to control everything. The only other external connectors would go to the sensors/motors and would all be board specific. We also specified that each module would be independent. So your main processor is not held down by taking sensor readings all the time. When it wants to know something it asks the sonar module what ranges it has stored in memory from the last time. The main controller would just do high end stuff. So lets say we have a servo controller designed to be a hexapod controller. Then the main navigation unit doesn't have to do anything but tell that module go left or rotate 45 degrees. All the hard stuff is precoded by experienced people (us) and it just works.

I prefer the I2C model that will simplify things on the user end. It makes our jobs harder but the noob who wants to make a little robot doesn't have to know that much and can get cooler robots faster. Thats the entire reason for this site anyway. With I2C also we get the ability to make pic and avr versions of each module. So you can still make them at home but don't have to change from your favorite MCU.


Edit: To cover the point about the mounting holes. These things are not holding up a truck. The only point of the screws is to hold it in place. It doesn't support any weight.
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 05:41:10 PM
well i guess i'll have to learn about i2c and see how to implement it into a motor driver... does anybody have any idea about how i would do this, does it need any additional hardware?
Title: Re: SOR community project?
Post by: Weird Fishes on April 20, 2009, 08:42:32 PM
I think the point of the I2C was that it wouldn't need a specific MCU. Any one with I2C will work just fine for it. My vote goes into keeping I2C.

Also... everyone does realize that the project isn't about 1 robot right. (A set of modules and protocols and standards)
Quote
smash can you make a sketch in microsoft paint or something quick so I ( and probably some others on this topic ) can get a rough Idea on what this robot will look like?
will it be a humanoid?
wheeled?
2 ft tall?
5 ft tall?
because this worries me. Come on people its only like 8 pages :). Everyone participating should at least have a decent grasp of the end goal.
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 08:56:27 PM
so is there any hardware needed to set up i2c?
Title: Re: SOR community project?
Post by: Asellith on April 20, 2009, 10:36:49 PM
Smash essentially all you need is a small AVR. The $50 robot core works fine. You need an MCU to translate the I2C into I/O or PWM what ever you use for the motor controller. I2C works kinda like USB but for chip to chip communications. You send a transmission down the bus and each controller looks for it's unique ID. There are 2 types of devices. Masters and Slaves. Masters either tell slaves to do something (Write) or ask them for data (Read). Slaves just wait for masters to ask them something. There can be several masters on one bus but they all need to play nice with each other. For instance your motor controller would be a slave. It could be a master but its not needed. The master would just say have go this speed and the motor controller says ok. Or the master can ask what speed is the left wheel going and the controller will just say 50% or whatever.

The message protocol is outlined in the I2C document on google docs. The commands sent inside the protocol are yours to figure out. For instance a command could be 11001111 for left wheel (1), forward (1), 00 to finish out the 4 bits and then (1111) for full speed. Or 1000 for half speed. You define those commands in the documentation and give you slave an address that will be unique to itself. So the master can say slave 0xFB01 do command 0xCF. Well I switched to hex by accident but no big deal you get the idea. If you can find a premade I2C PWM generator that fits the frame work Dunk worked out then you can replace the MCU with that IC.

Check out the article on Wikipedia for info on the protocol.

http://en.wikipedia.org/wiki/I2C (http://en.wikipedia.org/wiki/I2C)

Also read this tutorial.

http://www.societyofrobots.com/member_tutorials/node/35 (http://www.societyofrobots.com/member_tutorials/node/35)

Its not all that hard but a little bit harder then what you initially thought. But your hard work will pay off in the end if this gets off the ground.

Lets say you want to build a robot. You choice the module framework. Then you state what you want. I need 2 ultrasonic sensors, a motor controller, and a servo controller. The designs are already done. You piece it together and get to work on making your robot do stuff instead of the small stuff like how do I read data from my sensor. The plan is to produce the PCB designs and code for these modules and if you want you can assemble some and sell them to people or people can build it themselves. Kind of an open hardware/software platform to make the process of making really cool original robots faster and easier. No sense reinventing the wheel every time.

Keep at it your enthusiasm is making me think about getting back into this project. As soon as I finish moving to my new place and get my current health situation fixed I might join you :)
Title: Re: SOR community project?
Post by: SmAsH on April 20, 2009, 10:53:25 PM
wow! in depth much? lol thanks for that asellith, i thought i would've needed a microcontroller to do this as the l298 would. is it possible to have say few things attached to one slave? like have an atmega8 attached to the i2c bus then have it running a few things itself? otherwise to me its wasted i/o ports. i did read somewhere that the master can select to read by saying the slaves name then saying a number for something attached eg. 0xFB01 (atmega) then 0xFB02 (servo controller)?
and i keep hearing about this health condition, i figure its something serious? if i am allowed to ask, what is it? its fine if you dont feel like saying i was just curious/worried about our asellith :-\
get well soon,
smash
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 06:02:27 AM
sry guys  :'(
now I read over the whole topic.
It seems even cooler!
it will be like mindstorms just better.
individual modules, can be put together any way you want!
everyone (if we get a lot of people) just makes 1 module
(BTW is there anyone who would be willing to make a tutorial on how to implement I2C?)
then we send them all to one leader and and he tests/puts them together.
Inexpensive, good user interface, this could be better than mindstorms!
(I know, we first have to design/build the thing  :P )
gotta go to school now...
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 06:18:35 AM
yea that's the idea! but im hoping i can get some more people interested/building modules than just me and asellith. @dellagd for a tutorial on implementing c see i2c 101 and this http://www.instructables.com/id/I2C_Bus_for_ATtiny_and_ATmega/ (http://www.instructables.com/id/I2C_Bus_for_ATtiny_and_ATmega/) i completely understand what i2c is and how to implement it and learnt it all in under a day!
@project: if we had an mcu for every module wouldn't that mean programming every module? would it not make more sense to have like 3 modules plugged into one mcu which is plugged into the i2c bus, or would it be a one off program when we make the boards? i do want this board to be noob friendly ::)
i really do hope more people will pick up an interest tho, maybe if we can get some comps going like admin suggested in the other thread like a "best module design wins $50" or something along those lines...
but im defiantly going to give this project my all and make a few modules as i think this idea is a really good one that has not been done before.
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 06:42:37 AM
yea that's the idea! but im hoping i can get some more people interested/building modules than just me and asellith. @dellagd for a tutorial on implementing c see i2c 101 and this http://www.instructables.com/id/I2C_Bus_for_ATtiny_and_ATmega/ (http://www.instructables.com/id/I2C_Bus_for_ATtiny_and_ATmega/) i completely understand what i2c is and how to implement it and learnt it all in under a day!
@project: if we had an mcu for every module wouldn't that mean programming every module? would it not make more sense to have like 3 modules plugged into one mcu which is plugged into the i2c bus, or would it be a one off program when we make the boards? i do want this board to be noob friendly ::)
i really do hope more people will pick up an interest tho, maybe if we can get some comps going like admin suggested in the other thread like a "best module design wins $50" or something along those lines...
but im defiantly going to give this project my all and make a few modules as i think this idea is a really good one that has not been done before.

If I have the money I am defiantly willing to build a module
I  don't know what is it, maybe a acceleration sensor, but I'll try to make one

also I was hoping I could help with programming since I have good knolage of C++
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 06:50:31 AM
the module its self wouldn't be too expensive to make/test prototypes of but it would be the mass production that kicks you in the nads and only gives you an ice pack an hour later if you get what i mean? crappy metaphor i know! im only 14 and i haven't got a job and im going to attempt this so i cant see why half the other "richies" on this forum cant, no offense guys ;D but yea, if this idea takes off sor will be famous! but that's in theory, right? ::)
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 07:24:56 AM
yeah
hopefully this will make sor famous and make admin, you, me and Asellith rich
imagine how rich the guy who invented mindstorms is?
we will be financed to build whatever kind of robot we want!   
sry, off topic.
just had to get that out of my system  ;D
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 07:26:10 AM
nice, im always up for optimism! but i don't think that mindstorms is all that good as you just put blocks together and "hey presto" i got a robot that i made all on my own. there isn't really much room for development as you are tied down to what the have not what you can build. with this project it is open source meaning people will have plans to it so they can build their own modules at their houses as they please to fit their needs. say if i wanted a specific module and i had already bought a full mindstorms kit ($$$) i wouldn't want to start over with real electronics so i would have to go with something different. so what i am saying is that mindstorms is limited to what they've come up with, compared to our idea which is open source anything is possible! always open to expansion, imagine how rich the creators of arduino are?
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 07:28:46 AM
nice, im always up for optimism! but i don't think that mindstorms is all that good as you just put blocks together and "hey presto" i got a robot that i made all on my own. there isnt really much room for development as you are limited to what the have not what you can build.

I know we wouldn't like that, but regular people would (non roboticists)
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 07:33:27 AM
you cant classify people as "non-robotocists" everyone is a robotocist in my mind they just haven't been educated yet. and we need to keep this on topic, no more talk about mindstorms or people, unless their me...
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 07:35:47 AM
yet. , no more talk about mindstorms or people, .

@you cant classify people as "non-robotocists" everyone is a robotocist in my mind they just haven't been educated true, very true
@and we need to keep this on topic
sry
@unless their me..
lol.
Title: Re: SOR community project?
Post by: Weird Fishes on April 21, 2009, 07:40:32 AM
I think I'll build a module. I don't know which one... or when... but you can count me in.
Title: Re: SOR community project?
Post by: Asellith on April 21, 2009, 07:44:21 AM
@Smash My health problems are all related to the fact I have Crohn's disease. right now I am having complications and doing tests to figure out what else is wrong. But when sick most of the time Hobbies go out the window and I have to spend most of my energy/time keeping my job :)

MCUs are cheap. So wasted I/O isn't that big a deal. I looked and couldn't find a small 8 pin AVR with I2C. Smallest is 28 pins so package size is still rather large but not a big deal. Lets take your motor controller as an example. The base module has the capacity to run 2 motors. With the MCU dealing with the communication side all on one board. You could include in the board design a header strip that could be an off bus connection. Your original MCU needs to support this in code but the idea is there. So you design the main board and an expansion board. The expansion board holds enough to do lets say 4 more motors. These two boards link together to create 1 modules from the main controllers view point but can be expanded to utilize as much of the MCU as you want. The main issue is that each module needs to have 1 function. So you couldn't add IR sensors to your motor controller because thats outside the scope of that module. You could add optional temp sensors to read the motor drivers temperature and output that on the bus if asked for.

The modules can start of simple but over time as people add to the project they can get more and more complex. So if I need a 2 motor controller I need x amount of parts and the board. but if later I want to add more to it then I just buy the optional parts I need now and solder them to the board. The MCUs code is already there to support it I just need to tell the main controller to use it.

@dellagd This is intended to fill the gab between mindstorms and build it from scratch kinda like we do now. Basically you would start building the $50 robot to get your feet wet and understand the basics then with this system you have the skills to make the modules without needed a ton of coding experience and you can get right into building more complex robots faster and easier. It will also work for more advanced guys to be able to build bigger better cooler stuff to inspire the younger crowd.
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 07:47:31 AM
exactly
we can sell a starter kit and then you can buy more individual modules as you advance in robotics
there can be a option to use the graphic interface or make even cooler things by typing your own code
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 07:52:42 AM
i dibs motor controller and low power stepper!
@asellith: im so sorry about the sickness, it must be horrible :-[
and for the mcu since one atmega8 here is $6 ($16 if i buy local) i may get some attiny2313's as they're $2 online. but i do see your point about needing one mcu per module and it makes complete sense. but i will try it with both, although i see the at tiny fit for the job as it has a few I/O ports which are more than enough for this project. and not to mention smaller.
@dellagd, we could make the $50 robot board as a basic board or something like the arduino as the $50 board is pretty much the same as the arduino except for the usb and more memory but the $50 robot is plug'n'play!!
Title: Re: SOR community project?
Post by: Trumpkin on April 21, 2009, 08:21:08 AM
I think it would be kinda cool to have different cubes each housing different components of the robot (sensors, servos, mcu). Then have little pins (male and female) sticking out of each side of the cube which would be 5v and the i2c bus (you would also need a way to mechanically connect the cubes together). So then all you would have to do would be to stick the cubes together and program.
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 01:34:54 PM
ya there could be metal plates on the sides of the cubes, some labeled F and some labeled M for female and male ports. there would be magnets surrounding them, ensuring that they stick together and that then go together the right way.
I think we and people in the old sor project topic are really on to something.
I will make a quick sketch of my idea when I get home
Title: Re: SOR community project?
Post by: Weird Fishes on April 21, 2009, 02:27:54 PM
I don't like that idea. Too inflexible. This should be an easy way to start building a robot. That's too similar to Lego. If you want something to click together buy vex. I don't think we should go into the mechanical, but if we do it should be kept seperate from the electronics.
Title: Re: SOR community project?
Post by: chelmi on April 21, 2009, 03:57:45 PM
MCUs are cheap. So wasted I/O isn't that big a deal. I looked and couldn't find a small 8 pin AVR with I2C. Smallest is 28 pins so package size is still rather large but not a big deal.

Attiny25/45/85 are PDIP8 and support I2C through there USI unit. The trick is this is not called I2C in the spec but TWI for copyright reasons.
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 04:31:24 PM
i think that the cube idea is a bit far ahead and that we should get the actual module going first.
and chelmi, i will take a look at those and see if i can find them anywhere as my usual supplier doesn't stock em.
Title: Re: SOR community project?
Post by: Weird Fishes on April 21, 2009, 05:31:21 PM
I'm going to make a colour sensor board. I'd add it to the list but I don't have access.
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 05:33:42 PM
thats just a list of suggested modules that we could do. you are allowed to do other things. what type of color sensor is it? cam? something else?
Title: Re: SOR community project?
Post by: Weird Fishes on April 21, 2009, 05:35:33 PM
At first I'm gonna have another attempt at the rgb led and photo-resistor approach. If that doesn't work well enough, I'm going to rethink and switch my methods. I think it will work though...
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 05:41:42 PM
i could get one from raI guess I could make a acceleration sensor
I would get these from radioshack, but has anyone made one yet that I can look at to see how I actually build a module
just one person say "this is how we will build the modules" so I can build one correctly
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 05:48:01 PM
@werid fishes: yea, thats what i was thinking would be the easiest.
@dellagd: pretty much you would make it how you would normallly have it ready to be hooked up to an mcu then hook it up to the mcu on your board and use the mcu's i2c to hook it up to the bus. if you dont understand ill pm with with a better explanation.
Title: Re: SOR community project?
Post by: dellagd on April 21, 2009, 05:50:27 PM
@werid fishes: yea, thats what i was thinking would be the easiest.
@dellagd: pretty much you would make it how you would normallly have it ready to be hooked up to an mcu then hook it up to the mcu on your board and use the mcu's i2c to hook it up to the bus. if you dont understand ill pm with with a better explanation.

pls do so

also arn't we doing individual modules like this

Self-Replicating Repairing Robots (http://www.forofriki.com) (http://www.youtube.com/watch?v=VyzVtTiax80&feature=player_embedded#normal)
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 21, 2009, 07:59:31 PM
Right now I'm on the verge of relocating to another apartment, so hobby is postponed. I just look from time to time on the forum to calm down from stress and head ache.

If you check the previous thread of SoR comunity project, you'll see that I was working on a small motor module using tiny26 and SN754410. Also I had plans for L298 to replace the SN.

Now I have other parts bought and I'm preparing the playground to build 2 motor drivers to be installed inside servos for my MiniEric robot. I will be using tiny45 and FAN8082. They are both 8 pin ICs, small enough and easy to work with. I also will be using 2 Hamamatsu P5587 IR photoreflecting sensors (with integrated trigger Schmidt) for a quadrature encoder to be installed inside the servo instead of the potentiometer. It is not a standard module, but I think it will make a nice alternative. Also the code will be the same as for a dual motor module.

Now let's talk a bit about the functionality of the motor module. To be able to set a speed we need to measure the current speed to be able to adjust it to get to the set value. For this we need encoders. Encoders should be using external interrupts to increment/decrement a counter and if we read it at the same time interval we can calculate the speed. So we need to connect the motor driver and encoders to the MCU on the module.

When we will get to the code, we will find out that we might want different types of motor control. For instance, we might want to go with a set speed, or with a set distance, or with a mixture. We also will also want to be able to turn on the spot some degrees or curve nicely. Then we might want to read the traveled distance, the current speed and so on. All these commands have to be hard coded and documented for the module so anyone else can use it easily.

So, there is a lot of thought and work on a motor module. At the time I felt I needed to step back for a while so I learn more to make sure I get it right where I want it. Now I will go ahead and start working on the I2C continuous servos, at least have the hardware ready for programming. But I would like to see you succeed on building the L298 module, so I'll help as much as I can.
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 08:14:24 PM
why thank you ro-bot-x! im glad to see some support from people starting to build up again! i did think about having encoders on there but never really gave it much thought and went on to finishing the pcb. but would encoders mean more wires going down to the motors? and one last question, is it possible to use the l298 to drive one motor@4A?
Title: Re: SOR community project?
Post by: galannthegreat on April 21, 2009, 08:24:33 PM
The L298 can handle that kind of current but beware and take into consideration the stall currents and loads currents.
Title: Re: SOR community project?
Post by: galannthegreat on April 21, 2009, 08:56:30 PM
Hmm... now that I think of it, this is a great method to follow when building robots, and I think I'll use something along these lines for my next 'bot in the pot. Also for beginners out there who think robots are complicated and all, a good way is to think about them is, like this project here, to look at them is in individual modules.
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 09:16:30 PM
well we do want this to be user(noob) friendly, most people who ive tried to get into robotics dropped out when i told them how easy it was then they saw the code.... they think that all robots are these cool bipeds with cannons for arms :-\ sad really...
Title: Re: SOR community project?
Post by: galannthegreat on April 21, 2009, 09:26:20 PM
Yes that's too sad that when people see the code of something they get kind of "scared" and fend away from it, now in all honesty the coding part for me is still like this in a sense, but that's the learning process... and I think I'm stuck in a loop. (no pun intended) ;D
Title: Re: SOR community project?
Post by: SmAsH on April 21, 2009, 09:44:27 PM
hah, yea when i started i was scared s*#tless of the $50 robot and attempted the $40 robot instead.
i think most people who have never been introduced to robotics would see a plastic box with wires coming out of it more easy than the same thing with no box, ie just a pcb with components soldered on.
@dellagd, no we aren't really doing that sort of thing where they link together like that. we are doing something where its just a bunch of pcbs that can be linked via i2c.
@all, ive put together an updated list of modules based on dunks list as he is a busy man and probably has netter things to be doing. if i have missed anything or you want something added please pm me about it.
 modules (http://www.freewebs.com/smash-robotics/motorcontrollermodule.htm/)
 
EDIT: i have added in I2C to my motor driver... this is the order i have my pins in (SDA-SCL-GND-+5V) is this correct?
EDIT2: attached images to help :P
Title: Re: SOR community project?
Post by: dunk on April 22, 2009, 03:00:16 AM
hi guys,
did you see the example module?
http://www.societyofrobots.com/member_tutorials/node/69 (http://www.societyofrobots.com/member_tutorials/node/69)

it's only intended as a test platform but the firmware implements the communication protocol we talked about last time.

in retrospect, the communication protocol between modules might have put a lot of people off the project.
while it is important modules communicate with each other in a standard way it might be simpler limiting this to the i2c bus.
adding the UART to the protocol complicates it significantly.

i have a motorcontroller using the LMD18200 on my robot which implements the communication protocol but the board layout would need some work to make it match the module specifications.
scroll down to the "Motor Controller" section here: http://mrdunk.googlepages.com/electronics (http://mrdunk.googlepages.com/electronics)
let me know if you want the firmware for the motor controller. i think i still have it somewhere.

i really don't have time to work on this just now as i'm getting ready to set off on a year's rock climbing trip. (no robotics for me for a while...)
but good luck with the module project.
having a set of circuit boards designs that fit together and work together will cut out a lot of the development time in future projects.


dunk.
Title: Re: SOR community project?
Post by: SmAsH on April 22, 2009, 03:09:53 AM
darn, i cant believe i missed that in the member tuts. hah, i didn't even know they were going to use uart at all? that probably would've made me only do this in like a years time when i would've had much more experience and grasp over the concept of uart. but i think that I2C will do fine as its simple yet it gets the job done. and we can probably just implement a module for this later cant we?

i think we'll be able to keep this project up while your gone as long, as we can get the first modules up and tested and keep interest growing/staying.
i just have two questions...
1st: in eagle is there a measuring tool. ie draw a line and it tells you how long it is?
2nd: if we decide to put the holes in the boards (not in the C shape) how far in do you think would be good approx?
and good luck with the rock climbing! it really sounds like fun, i love rock climbing!

PS: don't fall, we still need you ;D
Title: Re: SOR community project?
Post by: Ro-Bot-X on April 22, 2009, 04:59:17 AM
About the holes, many people have an Axon or they are going to get one at some point. I think you can go ahead and use the Axon board as a template for the motor controller. That means board size, shape, holes placing and size. I hope Roboduino has at least 2 holes at the same distance as the Axon...
Title: Re: SOR community project?
Post by: SmAsH on April 22, 2009, 05:00:59 AM
is there any way to measure in eagle?
and added on-board regulator :)
Title: Re: SOR community project?
Post by: dellagd on April 22, 2009, 06:09:24 AM
hah, yea when i started i was scared s*#tless of the $50 robot and attempted the $40 robot instead.
i think most people who have never been introduced to robotics would see a plastic box with wires coming out of it more easy than the same thing with no box, ie just a pcb with components soldered on.
@dellagd, no we aren't really doing that sort of thing where they link together like that. we are doing something where its just a bunch of pcbs that can be linked via i2c.
@all, ive put together an updated list of modules based on dunks list as he is a busy man and probably has netter things to be doing. if i have missed anything or you want something added please pm me about it.
 modules (http://www.freewebs.com/smash-robotics/motorcontrollermodule.htm/)
 
EDIT: i have added in I2C to my motor driver... this is the order i have my pins in (SDA-SCL-GND-+5V) is this correct?
EDIT2: attached images to help :P

oh
cool
Title: Re: SOR community project?
Post by: dunk on April 22, 2009, 06:16:20 AM
is there any way to measure in eagle?
look at the numbers at the top left that change as you move the mouse.
they are the cursor's position.
you can set the units to either inches or mm somewhere in Eagle's settings.

dunk.
Title: Re: SOR community project?
Post by: dellagd on April 22, 2009, 06:18:14 AM
dunk
are you in for the community project like me smash and Asellith?
(plz post under this if you are in)
Title: Re: SOR community project?
Post by: SmAsH on April 22, 2009, 06:24:49 AM
he was one of the main organizers of the project but i don't think he'll be doing it this year as he's busy.
and dellagd, are you making an acceleration module? if so i need to add you to the module list :)
Title: Re: SOR community project?
Post by: Trumpkin on April 22, 2009, 08:09:04 AM
Dohhhh already have to many projects going on to help you guys out with this one.  >:(
Title: Re: SOR community project?
Post by: dellagd on April 22, 2009, 01:57:55 PM
not currently
once I know how to make one for sure I will make a exceleromerter module
Title: Re: SOR community project?
Post by: SmAsH on April 22, 2009, 02:14:34 PM
well i guess me and weird fishes are flying solo for a while (we picked the easy ones :P ) if you need any help in the future you know where to go ;)
Title: Re: SOR community project?
Post by: Asellith on April 22, 2009, 05:04:40 PM
To stop this from turning into another 8 page thread that scares people away from reading it, Why don't we go start splitting things up. If you have new things to talk about concerning the project entitle a new thread labeled SOR Community Project: (Whatever). That way we can keep things seperate. For instance if you start a module and want to ask question concerning just that or discuss a change to the board dimensions. Whatever the case it would be best to seperate them into new posts. People start seeing one thread about a project and they ignore it. If they see new ones popping up all the time it will let people know that the project is alive and then they will investigate it more. Also keeps us organized. This was discussed in the last post but never took off.

Oh and you guys working on module keep posting updates to keep interest alive. Not mundane stuff but every few days post something to let us know your alive and working on it. That will help attract attention.
Title: Re: SOR community project?
Post by: SmAsH on April 22, 2009, 05:37:59 PM
ok asellith, if i find a new question i shall start another thread.
@dellagd (below this). i really doubt it is of that importance to start a new section. maaaybe if this gets off the ground but until then its a definite no.
Title: Re: SOR community project?
Post by: dellagd on April 22, 2009, 06:03:42 PM
good idea
maybe admin will make a new section of the forum this time
Title: Re: SOR community project?
Post by: dellagd on April 24, 2009, 04:21:32 PM
since we plan to make money off of this project do you think it should be COMPLETELY open source?
i mean, a company CEO could be browsing the internet and see this, snatch up the idea, a boom, now lego makes the money instead of us.
we need to like protect these ideas somehow
and my idea is to do this
make a weebly free webpage
like sorproject.weebly.com
and then make a subpage like sorproject.weebly.com/543925894.html
543925894 being the password
Title: Re: SOR community project?
Post by: Razor Concepts on April 24, 2009, 04:31:22 PM
Projects designed to make money will be less sucessful than open-source projects. Just look at Arduino for an example.

These projects should not be designed to make money - just to contribute to the robotics community. Selling stuff is still okay, but everything should be open source. If the project is sucessful enough, people who appreciate it can donate.
Title: Re: SOR community project?
Post by: dellagd on April 24, 2009, 06:22:25 PM
I see what you mean
awww :'( so me and smash (and everyone who helps) wont be come millionaires on this (sarcasm)?
Title: Re: SOR community project?
Post by: Private Reid on April 24, 2009, 08:08:21 PM
Razor Concepts is right.

Open Source all the way !

There has been many topics on a SoR community project, and all (i think) have died... I will follow this thread with interest, i do hope it works out !

I can't really "help out" much cause I ain't that smart  :(  But I'll post a comment if I do know something !
Title: Re: SOR community project?
Post by: Weird Fishes on April 24, 2009, 08:14:54 PM
Yes. Totally open source.

GNU GPL might be a good idea... so no one could do that. I don't plan on making money off of this. I think of it sort of as the linux of robots. I'll sell stuff at cost if there's demand, but mostly things will be free in my book. (Schematics, code, everything you would need to build your own should be free. I'm not going to sell pre-made stuff).
Title: Re: SOR community project?
Post by: SmAsH on April 24, 2009, 08:45:11 PM
i agree on the open source part of things, i never really intended to make money off this project, just to make something good that will benefit others.
Title: Re: SOR community project?
Post by: Private Reid on April 24, 2009, 08:51:18 PM
Man, James ! Throughout the Forum I made about 10 posts this morning.

And then you come along and reply to all of them within 2 minutes !

I was pressing the refresh button on my "Show new replies to your posts" and every time I did there was another reply from "SmAsH !

Your quick !
Title: Re: SOR community project?
Post by: SmAsH on April 24, 2009, 08:58:30 PM
;D im sorry, but sometimes being quick can be a good thing ;) its school holidays for a few more days so i have nothing better to do :P ill calm down a bit as i need to sleep bad.
Title: Re: SOR community project?
Post by: dellagd on April 25, 2009, 07:38:02 AM
ok
making money off of it was just a idea
Title: Re: SOR community project?
Post by: SmAsH on April 25, 2009, 03:49:35 PM
making money off it could come with things but it would depend on how much demand there is.
Title: Re: SOR community project?
Post by: dellagd on May 18, 2009, 06:18:34 AM
is this still on, like smash are you making youre mod?
Title: Re: SOR community project?
Post by: SmAsH on May 18, 2009, 02:23:47 PM
yep, still on.
im going to be placing an order on the parts soon enough.
Title: Re: SOR community project?
Post by: dellagd on May 18, 2009, 02:51:41 PM
ok good I just wanted to make sure because the posting on OSCAR topics kindof slowed down
Title: Re: SOR community project?
Post by: SmAsH on May 18, 2009, 03:13:57 PM
nah, i just had to halt it for a while because of major exams last week ;D
Title: Re: SOR community project?
Post by: hazzer123 on May 19, 2009, 09:56:38 AM
Im also doing exams the whole next month so i don't have any time to work on this for a while. Ill try to get the SSM done by the start of july :)
Title: Re: SOR community project?
Post by: SmAsH on May 31, 2009, 03:43:52 PM
earlier today i was talking to some people in the chat.
we were talking about how some of the modules don't really need I2C and how this is just complicating things.
i can understand how I2C can be useful for some of the most advanced things like gyro modules etc
but is it really that useful for a simple motor driver circuit? i mean... it is a lot easier to program a motor controller
to go forward but it will most likely be a bitch to program into I2C let alone get it working...

i mean, i would rather plug my motor controller into some pins and just turn the pin high than have to set up
a whole i2c program on another microcontroller unit. not to mention the cost that it is putting down on the
modules. i know that the i2c doesn't add that much cost but still... an extra $2-$4 would make a bug difference
for the consumer.

i2c is fine for modules that do more than one thing like hazzers scanning servo module but for a simple little module
is it really worth going through all that effort?
do you guys think we would be able to make some of the modules without I2C or am i going nuts?
Title: Re: SOR community project?
Post by: Webbot on May 31, 2009, 04:22:21 PM
Personally I agree with you. However: I believe one of the goals of the project is to make it so that all modules can be connected and controlled in the same way. So a bit like having a USB hub and you keep adding things to it by just plugging them in - no need to know which pins are the UART or +5v vs Motor supply etc. So easier for the newbie but the downside is that it makes the cost of each module more expensive.
Title: Re: SOR community project?
Post by: dellagd on May 31, 2009, 05:49:50 PM
last time I checked the main point of this was to make a I2C robotics system
If it isn't I2C its just a robot right?
Title: Re: SOR community project?
Post by: Ro-Bot-X on May 31, 2009, 06:02:24 PM
Smash, a motor controller is a smart device. It takes care of the PWM, encoder counting, acceleration, slowing down and so on while the main microcontroller is free to work on the more important things.

Of course, for a very simple robot this is a bit overkill, but people can use the motor drivers directly on a protoboard, just like the $50 robot tutorial. I did it a couple of times and it works.

But time comes for any roboticist that he/she would want to build a more complex robot, capable of mapping, smooth movement, inverse kinematics and so on. I2C modules specifically take care of bits of the most complex parts of the code, better than multithreading. The only disadvantages are added cost and a bit more complex code. Cost we can't eliminate, but we can make the programming a lot easier with libraries of functions.
Title: Re: SOR community project?
Post by: paulstreats on May 31, 2009, 06:04:20 PM
I got the impression that it was to make a universally inerchangeable MODULAR system. Not necessarily a completely I2C controlled system
Title: Re: SOR community project?
Post by: paulstreats on May 31, 2009, 06:09:06 PM
Quote
Smash, a motor controller is a smart device. It takes care of the PWM, encoder counting, acceleration, slowing down and so on while the main microcontroller is free to work on the more important things.

This is where the problems are happening. The motor controller smash was looking at making was just going to be an L298 on a board. To make it I2C compliant he was going to add an mcu to it. whereas he might aswell just connect it straight to the master without I2C.

What you are describing is to make a motion controller board that is closed loop. i.e. incorporating encoders so it creates a full system which is capable of running its own movement algorithms by command.

A closed loop motor controller would be worthwhile adding I2C to....
Title: Re: SOR community project?
Post by: Ro-Bot-X on May 31, 2009, 06:30:53 PM
I know Paul, to me an L298 on a board is just a motor driver augumented board. There's no control in that board.
Title: Re: SOR community project?
Post by: SmAsH on May 31, 2009, 07:35:48 PM
i can understand if there is an encoder and other more complex stuff but if its just a motor driver i see no need for it...
Title: Re: SOR community project?
Post by: kpmcgurk on May 31, 2009, 07:43:45 PM
I like the Idea of this!! I will help where help is needed!

Do we have a name for this project yet? besides SoR community project?

Also, It would probably be better if there was another way of working on the project, like a forum for just this project?

I know i am turned away by the 4+page thread(not to mention the 8+ page other thread)
Title: Re: SOR community project?
Post by: Ro-Bot-X on May 31, 2009, 07:52:19 PM
The name is OSCAR. All related topics have the name in the topic title, but they are in the proper sections, e.g. Electronics, Software... A search for OSCAR will reveal them all.
Title: Re: SOR community project?
Post by: SmAsH on May 31, 2009, 08:15:07 PM
so do we have any more opinions on the non-I2c option?
Title: Re: SOR community project?
Post by: hazzer123 on June 01, 2009, 01:06:16 AM
This platform will be really flexible. We won't restrict someone to using just our modules for their robot's functionality, if they only want to use 1 of them. If they don't want to use a fancy I2C DC motor controller with nice control algorithms etc, then they don't need to, they can make their own little motor driver circuit and control it directly from the MCU.

I2C takes only 2 pins from the MCU. There will be plenty of pins left over for other things.
Title: Re: SOR community project?
Post by: SmAsH on June 01, 2009, 01:19:56 AM
so are you saying having both I2C functionality and regular pin connection functionality would be alright?
Title: Re: SOR community project?
Post by: chelmi on June 01, 2009, 01:01:28 PM
Ok, here is my opinion.

Before hearing about the community project, I had the idea of building my robot around I2C for on simple reason:
extensibility. A clean an simple interface (I2C) between components allows you to modify your robot without having to
redesign the complete thing each and every time. You can add sensors, change the MCU, ... without interfering
with the rest of the design (think about pin assignment, timer counters, interrupts, ....). With I2C you add very little constraint:
a I2C bus, a minimal API to access the bus, the "standard" connector. That should be it. The MCU, chassis, power, ... is up to you.
If you want to use a peripheral directly (direct pin access), do it ! But it shouldn't be part of the OSCAR system IMO.
Or else we will end up with a bunch of semi compatible boards and beginners will be lost. Of course I2C won't solve this problem
completely but it provides a good way to isolate subsystems.

Smash, about your example of a simple motor driver: a semi-complex robot will probably require more than just a driver. i.e. a complete
motor controller (with feedback loop and PID). In this case, dedicating a microcontroller to this task is a very good idea IMO.

Bottom line: OSCAR must impose I2C communication between boards. But for simple sensors or drivers, we shouldn't impose anything.
Title: Re: SOR community project?
Post by: SmAsH on June 01, 2009, 02:33:46 PM
i see your point chelmi,
i agree that it would be easier to code a change of hardware on the user end with I2C but
i was just saying that it is going to be hard to set up the I2C software.
i will put I2C availability onto my board, but also the option to control it by hooking it up to pins on the microcontroller.
Title: Re: SOR community project?
Post by: chelmi on June 01, 2009, 02:49:27 PM
i was just saying that it is going to be hard to set up the I2C software.

Why? there is already tons of code available for most microcontroller to setup and control I2C. And this is a one time cost: once you have this base layer you never have to touch it again (unless you use a different MCU). After that, it's just read/write to a specific register at a specific address.

In my opinion it's way simpler that having to configure pins, jungle with timers, interrupts,... for each peripheral...
Title: Re: SOR community project?
Post by: SmAsH on June 01, 2009, 02:51:53 PM
i guess your right chelmi :-\
ill take a look at some of those libraries sometime...
Title: Re: SOR community project?
Post by: dellagd on June 01, 2009, 03:16:37 PM
I agree with chemli, but I also like the idea of having the ability to just hook up a servo to a master for easy control
Title: Re: SOR community project?
Post by: chelmi on June 01, 2009, 03:41:07 PM
I agree with chemli, but I also like the idea of having the ability to just hook up a servo to a master for easy control

Yes, but you don't need OSCAR to do that :p
Title: Re: SOR community project?
Post by: SmAsH on June 01, 2009, 09:02:43 PM
lol, that confused me...
Title: Re: SOR community project?
Post by: dellagd on June 02, 2009, 06:07:14 AM
no I liked the idea of it mostly being a I2C system, but we dont take away the  ability to just straight hook up a lightsensor. if you want 1 simple light sensor there is really no point in making a whole module for it! its just impractical
Title: Re: SOR community project?
Post by: SmAsH on June 02, 2009, 06:52:50 AM
but for most of the I2C systems we want them to be more than just a "simple" design that a normal
I/O pin can handle otherwise it defeats the purpose of I2C.
for example... a simple motor driver will take up 2 I/O pins, add an encoder to that, that's another 1/2 pins.
with I2C this "over" usage of pins can be avoided.
but with the axon i don't think many people would run out of I/O but still...
Title: Re: SOR community project?
Post by: hazzer123 on June 02, 2009, 07:45:56 AM
It's not just about IO pins. Having the main controller send out PWM to a H-bridge would use up a timer, or some processing time. Outsourcing this job means the master has more time to ''concentrate' on the higher-level processing,
Title: Re: SOR community project?
Post by: dellagd on June 02, 2009, 01:53:43 PM
the point of this I2C system is like the master can say to the motor-comtroller, get me here. the motor controller Will figure out how to do that not the master. the master is the conductor of a I2C slave orchestra
Title: Re: SOR community project?
Post by: spykid99 on June 04, 2009, 05:17:14 AM
Hi Guys
As far as i can see from reading the first 2 or 3 pages of the original thread we are trying to create a system of "modules" that interconnect with each other like a plug-and-play system on computers.

Also because i live in Australia and i don't think anyone else does i cant help with building but i would be willing to help with high-level(C/C++) programming (I think you guys are talking about assembly, sorry if i'm wrong) i also think we should concentrate on a wide range of MCUs (I have not seen one post on the forum relating to OOPic for example and thats a good MCU (I think)  plus im good at programming that.

What do you think, please tell me if im wrong (Im a noob)
Spykid99
Title: Re: SOR community project?
Post by: SmAsH on June 04, 2009, 05:29:00 AM
Quote
Also because i live in Australia and i don't think anyone else does
er hem... i believe somebody forgot to view smash's profile?
you can still build a module, no matter where you live.
Title: Re: SOR community project?
Post by: Ro-Bot-X on June 04, 2009, 06:07:08 AM
Hi Guys
As far as i can see from reading the first 2 or 3 pages of the original thread we are trying to create a system of "modules" that interconnect with each other like a plug-and-play system on computers.

Also because i live in Australia and i don't think anyone else does i cant help with building but i would be willing to help with high-level(C/C++) programming (I think you guys are talking about assembly, sorry if i'm wrong) i also think we should concentrate on a wide range of MCUs (I have not seen one post on the forum relating to OOPic for example and thats a good MCU (I think)  plus im good at programming that.

What do you think, please tell me if im wrong (Im a noob)
Spykid99

All the embeded programming is done in C. Assembly is used rarely, for tight functions that are time sensitive (I think, I never use it). So, if you like to help with the programming you're wellcome. Many of the electronists are not that good programmers (especially me :D )

About microcontrollers used, we are using AVRs and PICs. OOPic is an augumented PIC that has a firmware installed that interprets the commands stored in EEPROM at the time of execution. This slows things up pretty much, plus you don't have too much space in the EEPROM for user data. I have an OOPic-R board and I coudn't use it to read the Ping))) sensors that require a 5us trigger pulse, because the OOPic takes too much time to do things.

I sugest you take a look at Arduino, since it uses C/C++ programming, uses an AVR ATmega8/168/368 and has a bootloader that allows you to program it just like the OOPic. I am going to use this language to program my modules, at least until someone translates the programs in "real" C.
Title: Re: SOR community project?
Post by: dellagd on June 04, 2009, 02:59:22 PM
I'm a good c++ programmer too. here's a sample.
also, "real" c?

PS actually I have seen a good few Australians here

oh and quick question, I want to add a pic in my signature, what is the code to do that?
Title: Re: SOR community project?
Post by: SmAsH on June 04, 2009, 03:37:38 PM
i think you get the link for the pic then add [img] and [img] around it.
not sure but ill try it
Title: Re: SOR community project?
Post by: kpmcgurk on June 04, 2009, 03:43:24 PM
hrm... I know its off topic, but when your program asks for a product name, And I type it in, It throws me into an Infinite loop... Am I not meant to type in a string? or did you forget to add this at the top

Quote
#include string.h

Thats my Idea... remember that strings are not part of C++, they need to be initialized first before use...

could I see the code for your program?
Title: Re: SOR community project?
Post by: dellagd on June 04, 2009, 07:35:11 PM
where are you getting this from  ???
Title: Re: SOR community project?
Post by: hazzer123 on June 05, 2009, 01:15:30 AM
Please make a new thread for this stuff guys....

Regarding our choice of MCU, this wasn't really a conscious decision (sort of). We are using PICs and AVRs because that's what we and most people in robotics use. If you like OOpic then feel free to port our libraries to work on those, we'd be super-happy if anything increased the flexibility of the platform.

Of course there is no code to port yet... but i think there are a couple of guys who will get the ball rolling soon. And as soon as i finish exams (next friday) ill get to work on my module and see if we can start making real progress with this thing.
Title: Re: SOR community project?
Post by: spykid99 on June 05, 2009, 01:50:40 AM
Please make a new thread for this stuff guys....

Regarding our choice of MCU, this wasn't really a conscious decision (sort of). We are using PICs and AVRs because that's what we and most people in robotics use. If you like OOpic then feel free to port our libraries to work on those, we'd be super-happy if anything increased the flexibility of the platform.

Of course there is no code to port yet... but i think there are a couple of guys who will get the ball rolling soon. And as soon as i finish exams (next friday) ill get to work on my module and see if we can start making real progress with this thing.

I agree that we should make a new thread +ask admin for a new subforum (which would be great) and that its great that we will soon have libraries so people who can understand AVR language (NOT ME) can do that porting and also where is a good website to learn about PIC programming with C

Thanks
Spykid99
Title: Re: SOR community project?
Post by: hazzer123 on June 05, 2009, 02:02:28 AM
Well there are 2 main tools for PIC programming in C. Microchip C18 (for PIC18s) and Hi-Tech C PRO Lite mode (for all PICs).

C18 - Getting started user guide - http://www.kevin.org/frc/C18_3.0_getting_started.pdf (http://www.kevin.org/frc/C18_3.0_getting_started.pdf)
Lots of code examples - http://www.kevin.org/frc/ (http://www.kevin.org/frc/)

C PRO Lite mode - Vendor's site here - http://www.htsoft.com/products/compilers/piccpro-modes.php (http://www.htsoft.com/products/compilers/piccpro-modes.php)
They have a forum and there'll be plenty of manuals and example code in google.
Title: Re: SOR community project?
Post by: spykid99 on June 05, 2009, 02:13:32 AM
Do we have a freewebs thingy or does someone need to make it

Thanks
Spykid99
Title: Re: SOR community project?
Post by: SmAsH on June 05, 2009, 02:26:22 AM
i have a freewebs page that has info on it...
i can make a page specifically devoted to ALL modules if you guys want?
Title: Re: SOR community project?
Post by: spykid99 on June 05, 2009, 02:29:51 AM
I think that if we have a site that everyone could edit and it contained a page for each module +about project etc.

Thanks
Spykid99
Title: Re: SOR community project?
Post by: SmAsH on June 05, 2009, 02:38:07 AM
ok, we need to not post anymore in this thread about this stuff...
we don't want another 8 page thread ::)
if you have any suggestions you can pm me.
Title: Re: SOR community project?
Post by: kpmcgurk on June 05, 2009, 05:23:39 AM
If you guys want, I can make a website...maby with a blog and forum on it? along with just a general home page so onlookers can get a feel for what the project is like before they decide to take part in it?

EDIT: Nevermind, I just saw the other thread...
Title: Re: SOR community project?
Post by: dellagd on June 05, 2009, 05:35:46 AM
why dont we just make a free webs page and pm everyone working on the project the username and password?
like www.sorproject.freewebs.com (http://www.sorproject.freewebs.com)
Title: Re: SOR community project?
Post by: spykid99 on June 05, 2009, 05:46:46 AM
You just missed out on the fun, i did it LOL
Go to the chat

Spykid99
Title: Re: SOR community project?
Post by: Asellith on June 05, 2009, 08:00:16 AM
I am working on a sourceforge page for the project. That was we have forums and news and the ability to mass contribute in an established system. Should be up and running enough to go public next week. I can start working on it again later today or tomorrow still not able to sit in a chair for long after my surgery yesterday.
Title: Re: SOR community project?
Post by: SmAsH on June 05, 2009, 03:25:28 PM
well, asellith.
its great to hear the surgery is over, but don't strain yourself, theres no rush...