Society of Robots - Robot Forum
General Misc => Misc => Topic started by: airman00 on October 19, 2008, 04:19:51 PM
-
Hello all ,
A month ago I finished a robotics product of mine ( its a secret for now) . I've been going through the manufacturing phase - you know the planning , getting parts , making calls, etc. The robotics product will come out in stores in 2-5 months time .
Anyways the reason I was making this post was because I could start manufacturing 50 dollar robot boards similar to this one : http://www.societyofrobots.com/robotforum/index.php?topic=5268.msg40956#msg40956 (http://www.societyofrobots.com/robotforum/index.php?topic=5268.msg40956#msg40956). The layout would be very different , but it would be more or less the same features. This extra PCB would not add any significant cost to the manufacturing run of my product, in fact it might even lower costs since the robot board uses almost the same components as my product (save money in bulk orders).
It would be layed out similar to the Axon - like a miniature Axon ;)- based off the ATmega168 ,USB uart( bootloader also?) and sell for approximately the same cost as an Arduino ($35).
What do you guys think of this? Would you buy one? Any suggestions?Any comments?
P.S. The robotics product is really cool , but I don't want to release details just yet. Hint: It can be used to make giant robots . But no more hints than that.
-
Make it look as close as you can to the Axon, and dont make a 100% profit from it ;). And I will buy it ;D
-
Would you want to buy it presoldered or just a solder-it-yourself kit?
-
homemade PCB or professionally manufactured?
who solders the ICs?
-
homemade PCB or professionally manufactured?
who solders the ICs?
all professionally done, soldered by a company .
Like I said, I'll be forced to charge at least 35 dollars for it (because of all the costs for manufacturing and soldering) - is there enough demand for this?
I was 1/4 way through my schematic and then I realize - hey I'm basically making the Freeduino! So I took the Freeduino files and I'm just modifying them a bit and changing the layout.
-
hmmm I'm starting to think its not worth it:
Heres my reasons :
Arduino and Freeduino do basically the same things my board does
It will cost me more to make these boards then a company who makes Freeduinos
I won't make so much profit since I have to compete with the prices of the Arduino and stay below $40.
-
Now see, that is why I gave up doing it :P.
There is no perfect board for everyone. Because aplications are different. However, if you can make the boards with 3 pin connectors, with jumpers to select the power to the digital 3 pin headers, mini USB connector with the chip soldered, that would be perfect for small robots. Users can use the Arduino software (or just plain C) and can practice soldering their own headers, sockets, LEDs, resistors, etc. I may be wrong, but none of the Arduino compatibles has 3 pin headers for all the pins, so you may have a piece of the market there.
You may consider to use DIP version of the microcontroller so the user can plug in ATmega8, ATmega168 or the new ATmega328P.
-
yea the only difference would be the 3 pin headers. I'd be selling 3 pin header Arduinos .
-
And that is a bad thing?
-
And that is a bad thing?
no but then again , is it all worth it in the end - I have to factor in labor.
I expect maybe 100-300 sales a year at around 10-20 dollars profit . Not that much cash... 1000- 6000 dollars a year from that . Its significant I guess, but I have to think about all the labor and stuff.
-
The market really needs something as cheap as the Arduino, but pinouts like the Axon. Seriously, one pin isn't enough to connect a servo!
But the Arduino is mass manufactured, like a year or two ago they claimed 10k sales . . . meaning they can get ultra cheap manufacturing prices.
Manufacturers charge way too much to solder on headers, so you'll either have to do it yourself, hire a highschool student, or have the customer do it.
With the extra headers, I doubt you can ever match the Arduino price and have it still worth your time.
I recommend selling two different types:
pre-assembled
not assembled
And of course price the pre-assembled one higher.
Good luck!
-
Admin is perfectly right. I just want to stress out something. Make sure the microcontroller is DIP version, not SMD. Because noobs may brick it, so they will just change the micro, not the whole board. Back in 2005 someone told me about a manufacturer in Chicago who had good prices and who could make the boards different colors soldermask and silk screen. Some of the colors were cheaper... Let me know if you want me to look him up for you.
As for the board layout, other than jumpers for power selection for the headers, I would like it to have separate headers for TTL serial, jumpers to conect it to the USB chip, headers for I2C interface with power and groud (4 pin header). If you toy with the ideea of the bootloader, try the AdaBoot bootloader, since they say it is universal and also works with Arduino programming environment. Maybe make it in the form of the modules we talked about building as a comunity project, so we actually can use it directly for that.
-
Since three pin headers seem to be in high demand, what about designing an Arduino/Freeduino 3-pin header shield?
-
I've been toying with the idea for some time of getting into hobby robotics sales but its a hard market to get into at first. I could see after establishing yourself in the market you might be able to produce something like this product after you have the capital to keep your cost low. What you really need is something that will fill a space in the market and be unique. Then you are not competing with others and just have to worry about making a little bit of profit off each unit. Then once your name is known and you have a supply chain setup as well as distributors (ie sparkfun or one of a thousand other hobby style electronic sources) then you can afford to take the chance on something that will strengthen your name but might not make a lot of money. The main problem is finding something that fits a nitch market that has not already been done. might want to try something the the Maxbotics EZ-1 sonar sensor. Making complex sensors easy to use and cheaper is always a good idea.
As far as headers are concerned I would see a problem if someone sold me a board with no headers on it. The problem I always have is them not including the headers in the package so I have to get some of my own.
-
Since three pin headers seem to be in high demand, what about designing an Arduino/Freeduino 3-pin header shield?
people like it when its all in one package - nothing else needed
if people had a choice between a regular arduino and one with 3 pin headers - they'd pick the one with the 3 pin headers
-
people like it when its all in one package - nothing else needed
if people had a choice between a regular arduino and one with 3 pin headers - they'd pick the one with the 3 pin headers
I think you should say robot builders instead of people... Arduino comunity has lots of people that don't build robots. They are using Arduino for lots of different stuff, using shields for interfacing. For some of them, a RoboShield would be a great addition, and some of them would buy a Roboduino board right away. There is a demand waiting to be filled. I would go ahead and build both a shield and a board. Heck, people who allready own a board will buy a shield and the others will buy directly the board, since it will be cheaper than Diecimila + RoboShield.
-
I think I'll offer a board and a shield for robotics(for people who already have an arduino)
what I was originally aiming at was a modified-for-robotics arduino. That if someone was faced with a choice between Roboduino and the regular Arduino they would pick the Roboduino.
-
Anybody want any other features besides the ones below ( features that arent very expensive to add on)
USB bootloader
male headers on all pins
six separate servo pins ( unregulated power)
low dropout regulator
female headers for regulated pwr,gnd,unregulated power
Status LED
reset button
ICSP programming pins
-
USB would be nice as long as it doesnt take up a huge amount of space on the board.
Maybe leave out ICSP pins to make it smaller? The AVRISP needs a usb port to plug into your computer anyways, so might as well stick to the USB bootloader.
-
I think i'll leave ICSP if bootloader gets accidentally deleted or something
-
why not add some of the ports connected to a mosfet or transistor so the ports can turn devices on and off at will without risking damage to the controller. that way sharp rangefinders etc.. can be turned on and off by the microcontroller. (or for a better example if you have a line follower with 5 ir led's, the controller struggles to provide power to all led's on its own but through a mosfet or transistor it can reliably turn them on and off allowing advanced functionality such as background cancelling and modulation).
You could add an i2c dedicated bus etc...
The trouble is knowing where to stop I suppose... Making a simple board with too many features can eventually turn into something difficult to use, always remember that you know how it works because you made it but try and think from the end users point of view, you can easily overwhelm somebody by adding too many features even though you or I think that they would be a good advantage. The only way to make a really well functional board is to allow it to work in a basic way and then switch the extra bits on when the users are ready for it through well structured documentation. I'm trying to create a board for release myself along with a few different types of hardware bases etc... but it really is difficult to know and plan for what the market knows and expects.
(welcome to the commercial developing world, the more people and competition there is the faster the future gets developed!)
ps. Cant wait to see what youre secret is!
-
You may consider to use DIP version of the microcontroller so the user can plug in ATmega8, ATmega168 or the new ATmega328P.
Thought it was a typo?
And yes, please add to the list of features the I2C bus (with power and ground pins, and place for pull-up resistors, just in case one needs them). Use the vertical mount resistor (0.1" spacing) for the pull-ups.
So when will the secret be out on the market? Does it make french fries? Hmm... I should build a robot that makes fries... Just kidding!
-
I think I'm gonna stop by the I2C - i thinks its too much. Maybe in version 2 . ;)
The secret product isn't really so amazing but incredibly useful when making large robots with a lot of DoF
-
quick question
if I do unregulated pins for 6 servos then those pins arent regulated which may cause problems for people who want more sensors.
So how many pins should be dedicated for servos? Minimum of 2 , maximum of 6.
Thanks
-
So for that group of 6 servo pins add an other 3 pin header where the user can use a jumper (like those on the PC motherboard) to select regulated or unregulated power. I got this ideea from the OOPic-R board, where there are 3 groups of 5 pins, each group with its own jumper to select the kind of power you want. Check out my old board here:
http://www.societyofrobots.com/robotforum/index.php?topic=2228.msg14977#msg14977 (http://www.societyofrobots.com/robotforum/index.php?topic=2228.msg14977#msg14977)
or zoom in and check out the left side of the board where the servo pins are:
-
I think I'm gonna stop by the I2C - i thinks its too much. Maybe in version 2 . ;)
That is too bad, since I2C is good for micro to micro communication (keeps the UART free for other stuff), for newer sensors like compass, Sharp GP2D12 I2C version, etc. The nicest part is that you can stack up many devices on the bus and it works perfectly.
Never mind, I just figured that SCL and SDA are the last 2 analog pins, so if you install the 3 pin headers close to each other, I can have the I2C bus split in two 2 pin female connectors and plug them in an L shape, where 2 pins are power and ground and the other 2 pins are SCL and SDA. The only thing that is missing (pull-up resistors) I can add on one of the slaves. But if you think you can squeeze them on your board, that would be awesome.
-
Hey you could sell photo sensors, no one makes those.
-
Hey you could sell photo sensors, no one makes those.
photoresistors? lol everyone makes those
-
Hey you could sell photo sensors, no one makes those.
photoresistors? lol everyone makes those
I mean the whole sensor, photoresistor + resistor + wires + pin + housing.
-
guess what guys I put on I2C option .
There is holes in place for optional resistors to be pullups for the I2C. So you can use Analog or I2C. Your choice. :)
Cost of the Roboduino will be between $40 and $50. Sounds fair?
-
$40 as a you solder kit and $50 as a all assembled board? It's a little more than an Arduino, but I guess it's worth it.
Thanks for adding I2C pull-ups! :)
-
I was thinking final product to be between 40 and 50 dollars . Solder kit would be less ( not sure by how much)
I'll post up an Eagle3D rendering of it later this week. I'm off to bed , gonna try to sleep a normal amount
-
Updates !!!
Eagle 3D of it :
(http://i273.photobucket.com/albums/jj202/erobot/Misc%20Documentation/roboduino_v1_011.jpg)
and the milled PCB board
(http://i273.photobucket.com/albums/jj202/erobot/Misc%20Documentation/Roboduino.jpg)
-
Looks great! Well done! So, when is going to be available just the board?
O.T. Didn't know you are hungarian... Forkos means Wolf, right?
-
Looks great! Well done! So, when is going to be available just the board?
O.T. Didn't know you are hungarian... Forkos means Wolf, right?
Not exactly sure , when the board will be available - probably in a month or two
Yep I have hungarian ancestry
-
The first Roboduino has been assembled!!!
(http://i273.photobucket.com/albums/jj202/erobot/Roboduino/DSC02309.jpg)
(http://i273.photobucket.com/albums/jj202/erobot/Roboduino/DSC02311.jpg)
(http://i273.photobucket.com/albums/jj202/erobot/Roboduino/DSC02308.jpg)
http://s273.photobucket.com/albums/jj202/erobot/Roboduino/ (http://s273.photobucket.com/albums/jj202/erobot/Roboduino/)
Waiting on a voltage regulator and then the Roboduino will be ready for testing.
-
Nice!!!
Make sure you burn the Adaboot bootloader!
Did you put the cap from the CTS line to the RESET pin? This ensures the autoreset to enter bootloader code when you send the sketch to the ATmega...
-
It looks great!
Do you think it will still be available in a month or two...January maybe?
-
Do you think it will still be available in a month or two...January maybe?
Should be available a few days before the holidays
Did you put the cap from the CTS line to the RESET pin? This ensures the autoreset to enter bootloader code when you send the sketch to the ATmega...
yep I did.
.1uF if I recall correctly......
As for the bootloader I guess there would be an option for that- with bootloader or without. I personally plan to use ISP only but I get it differs from person to person.
-
So, did you manage to add power selection jumpers for the servo pins? If so, where they are placed on the PCB?
-
So, did you manage to add power selection jumpers for the servo pins? If so, where they are placed on the PCB?
no power selection jumpers
All pins marked PWM have unregulated power while all the others are regulated. Shouldn't be a problem if you want regulated power since at the bottom left there are columns of pins dedicated for unregulated, regulated, and GND
-
Well done on developing the Roboduino,
Just reading through this thread. Did you throw away the idea of a robot arduino shield? That I would buy. 50 dollars extra (when I already have an arduino) for convenient headers isn't worth it for me (while I'm sure it would be for others).
With the amount of Arduinos sold already (approx 20,000), I think you would have a nice market for selling shields. I think people would be much more inclined to pay 10 dollars for a robot protoshied, which is only a pcb with some headers. Sell it as a kit and it should be very cheap to make.
Anyway it just a suggestion,
Well done again
David
-
I'll think about the shield.
Its definitely a possibility.
-
HalfShell is doing the Roboshield