Society of Robots - Robot Forum
General Misc => Misc => Topic started by: Asellith on April 23, 2009, 06:37:57 AM
-
Ok The Axon is 2.542 in square. So we could use this as the building block size for the project. The size is not nice and neat but I'm not sure it matters much. It is slightly larger then the original size proposed here :
http://docs.google.com/Doc?id=ddp2r5j8_24hhf75vck&hl=en (http://docs.google.com/Doc?id=ddp2r5j8_24hhf75vck&hl=en)
however I think if we go in an easy 65mm x 65mm multiples then the units will match the Axon within a 0.5 mm tolerance and that should be fine. We could also define a smaller board that would be 32.5 x 32.5 mm. Also I think rectangular boards should be ok. So a 32.5 x 65 mm board would be fine. With mounting holes spaced 0.1 inch on center from the corners to match the axon. This way if someone designs a module enclosure then the axon will fit into it nicely.
thoughts and idea? is it worth changing?
-
although its an odd shape i think its a good idea to do this for them. but, are we going to be releasing smaller, cheaper brain boards for people who are doing there first robotics project that don't want to spend $130 on one part? like a small mega8/168 board that costs like $20-$30 or maybe even an arduino board.
but.. i suppose i can alter my current design to fit the axon tonight :-\
-
I vote that a eagle library MUST be created
that will feature the AXON pinout on a board..... and schematics...
That must be the first step....
Since I have and AXON, I'll see what I can do...
-
i agree with lefteris on that one, im pretty sure this project should be open source with schematics and board files free for download.
@lefteris, it would be great it you could create an eagle file for the axon, hard though...
-
I'm with smash that we should use a much more inexpensive board, not that I have anything against the axon :D
-
we shouldn't completely rule out the axon but maybe create a board that's easier for beginners to use, the same size as the axon, and has I2C capability. if we were going to put one in a box, why not use something like a tiny23 or other tiny. but if we still needed I/O i would vote for either the $50 robot board (with admins permission) or something similar to the roboduino. but remember, general purpose I/o is still important even if these devices are on an I2C bus, what will happen if someone wanted to use a device we hadn't made a module for? we need to think about some more things before we start designing it, brainstorm anyone?
-
I don't think this modular system should be limited to one of each type of module. There's no reason there can't be the Axon and a smaller one as long as the modules all use the same protocol and conform to the other standards set. Just like I think there should be more than one motor controller they can have different features such as current handling encoder handling multiple motors single motors but they all need firmware one them that allows the main mcu to to control them the same way with the same commands. that way switching from one motor controler to another won't break code on the main board.
-
i cant see that idea being too hard to implement. i mean, if this project really takes off we'll definitely make more modules of one type with more features like madchimp suggested. seeing as im not exactly a whiz with code, can anybody think of a way to differentiate between modules to make the microcontroller say "hey, this isn't the old motor driver, this one has an encoder so i can get feedback" just an idea... but hey, if it can work then why not?
EDIT: ok so ive gotten my motor controller down to below the axons size ;D its now 2.45"X2" but we can always make it bigger to fit.
-
We have strayed and I will attempt to get us back on track here. Ok first off. I want to make the board dimensions conform the the axon as a standard. The axon does not have to be the only control module. Throw a ATMEGA8 on a board layout and pin it out for I2C and power then your done with a controller unit. The idea is to have HUNDREDS of modules. All conforming to a physical and communications standard. This will give us the ability to sit down with a list of possible modules and build whatever. If it requires 2 Ir sensors and a DC motor controller then no the axon as the main board is not the best option. But if you need 45 IR sensors, 12 lasers, 50 servos, and 25 dc motor controllers then the system can still work and you don't need a super computer to crunch the numbers because you have 15 to 20 modules all being controlled by a Axon that simply does high end decision making instead of crunching ADC numbers and calculating motor speeds.
We are creating a standard to allow for tons of modules. Each communicates the same way and fits into the same size or multiples of that size. If you want to make 5 different motor controllers then go for it. I think it would be better to do what a lot of companies do now and design great hardware then work the software in over time. Gives them better time to market and they can continue to attract new customers over time with new features through firmware upgrades.
For instance if you design the motor controller to take inputs from encoders, temp sensors, and any other type of feature you might want but not need then we order a single PCB and upload the version 1 firmware. Then over the years as you and others improve the code it will become better and better and people just need to upload the better firmware to get the new features.
@madchimp You have it right we need several versions of every module. I think we need at least two because we should have a Pic and an AVR version of each. however I'll have to put more thought into how to assign addresses to each of the modules. It could be user done simply by setting a global variable in the firmware so when you load the code onto the module it can be any number you want or it could be assigned to each as they get designed. Also you don't want to prevent the more advanced users from changing it. If I want to have two of the same module I need to be able to change the address so I can tell which is which.
I envision two types of modules. The store bought ones that come prebuilt (I.E. smaller SMD versions) and DIY kits or just plans. The end user get the choice as to what they want to pay for and how much experience they have. They can order their own boards or someone can do a big run and sell them to people through a website. The store bought ones would come hard coded with an address and the DIY ones could easily be changed to meet the users needs.
Anyway all that to say this, I think we need to use the Axon dimensions because it doesn't change much right now as there are no working modules in production now and moving forward if we can say yes the Axon with this code fits the standard for the SOR Open Source Robot. Those without the cash for Axons can use other main boards designed by themselves or others and those with Axons get the extra power and ease of use.
-
I agree. they could be sold as pre-builts (more expensive but come with support and warranty), kits (a little less expensive come with support but little to none warranty (but can be purchased separately)) and then the plans (cheap but get no support and warranty)
-
dellagd, i don't think the plans should cost anything at all! i mean, they're just plans! something like the way the roboduino is sold may be good. but we need to get some boards done and tested before we start trying to market these things. I'm getting some more copper clad boards next week so i can start etching some proto's to test.
-
sry. I'm so ill-funded I'm trying to get money anywhere I can get it. sry.
-
dellagd, i don't think the plans should cost anything at all! i mean, they're just plans! something like the way the roboduino is sold may be good. but we need to get some boards done and tested before we start trying to market these things. I'm getting some more copper clad boards next week so i can start etching some proto's to test.
Sorry to interrupt this thread... but are you trying to integrate this in the axon II ? Or are you trying to make this board connecting to the Axon I/II ?
Or like an add on board for the Axon I/II ? Why not combine this board I2C (dual) and H-Bridge (dual) like an universal motor controller. I think 5A max must be enough.
-
yeah, this will pretty much "marry up" with the axon/axon 2's mounting holes so you can stack them.
im working on a dual motor controller that can handle 2A per motor or 4A for one motor.
i don't really what you mean by "combine them"....
but this isn't really combining anything but just making a set of boards being able to stack on top of one another to save horizontal space. and yes these will operate through I2C but can still be operated through normal microcontroller pins if that's what you were asking?
-
yeah, this will pretty much "marry up" with the axon/axon 2's mounting holes so you can stack them.
im working on a dual motor controller that can handle 2A per motor or 4A for one motor.
i don't really what you mean by "combine them"....
but this isn't really combining anything but just making a set of boards being able to stack on top of one another to save horizontal space. and yes these will operate through I2C but can still be operated through normal microcontroller pins if that's what you were asking?
To create a size compatible board and place both a H-bridge and a i2c controller on it, so you can have it both ways at once instead of having 2 different boards needed.
-
yes, that's pretty much it... ok ive got pics up of the board that has the same mounting holes as the axon! plus no more voltage regulator on board... and some optional pads for the extra I/O on the other side as i had a bit of room left.
-
Just a comment about stackable boards - don't forget accommodating for wiring!
Otherwise all the wires will be strewn everywhere . . . perhaps make it so there is a known hole location at the top right of every board strictly for wires to pass through the stack. (instead of going outside the stack and getting rats-nest-like)
Or something like that . . .
-
i suppose it wouldn't hurt to have a small hole for the wires to pass through.
how big do you think would be suitable?
-
i suppose it wouldn't hurt to have a small hole for the wires to pass through.
how big do you think would be suitable?
big enough for a servo hitec connector to pass through . . . it doesn't have to be a hole, it can be a slit on the board's side, too . . .
-
yeah, by "hole" i kinda meant rectangular...
ill take a look and see if i can fit one in my design.
-
hey guys, just a quick question.
is there any hardware difference between a slave and master or is it all coding
-
the hardware can be the same.
yes, it is in the code.
although generally the master has got the better electronics... but meh, still can be the same...