SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
This thread is privately moderated by JussiH, who may elect to delete unwanted replies.
Old Dec 25, 2010, 06:10 AM
Friend is a 4 letter word...
JussiH's Avatar
3rd stone from the sun
Joined May 2007
2,670 Posts
Andrej....very interesting project!

Parallax Propellor is such a nice uC @ an unbeatable price......I am interested to know more about your project with the FEZ line of developement boards....do you have a link to your project?

I like the idea of a sub 50 dollar IMU board. For the MultiWii I still think we should start out with Gyro, accel and Baro - but adding pads for magneto and GPS connector for those wanting to evolve MWC in this direction. (Only problem is that most commonly used GPS units are serial - so I need a suggestion for an I2C unit....)

I still think the 50*50 format should be kept! More or less that is standard for flightcontroller boards....

I am thinking we should perhaps wait a bit with going to a new uC! Too bad with the Teensy 2.0++ ports....It would have been a nice and cheap upgrade from the Promini, and offering 3 extra PWMs and freeing up serial for GPS or other stuff!

I agree that 1280USB or 2560USB would be very nice candidates. And I am sure we could squeeze it into a smaller footprint than the Mega and APM bricks. But I higly doubt we could make it cheaper....unless we go for a fullon production volume (1000 pcs)...... But a nice project for the "mad scientists".

I dont want to derail the MWC. So I think its important that we do not stray too far from current hardware and that Arduino and Wii sensor integration is kept! It should still be simple and cheap to build with Wii sensors!

Best regards

Jussi
JussiH is offline Find More Posts by JussiH
RCG Plus Member
Sign up now
to remove ads between posts
Old Dec 25, 2010, 08:06 AM
Registered User
rimshotcopter's Avatar
United States, CA, Los Angeles
Joined Aug 2009
11,692 Posts
Quote:
Originally Posted by kinderkram View Post
Mariano (mgros) already ported the MultiWii code to the Mega1280 and integrated the code for the BMA020. Maybe you should contact him?
Absolutely Freaking AWESOME!!!!
rimshotcopter is online now Find More Posts by rimshotcopter
Reply With Quote
Old Dec 25, 2010, 08:15 AM
Registered User
rimshotcopter's Avatar
United States, CA, Los Angeles
Joined Aug 2009
11,692 Posts
Quote:
Originally Posted by 420RcPilot View Post

I agree that 1280USB or 2560USB would be very nice candidates. And I am sure we could squeeze it into a smaller footprint than the Mega and APM bricks. But I higly doubt we could make it cheaper....unless we go for a fullon production volume (1000 pcs)...... But a nice project for the "mad scientists".

Best regards

Jussi
In my opinion, I believe there should be Two versions. One for the palm copters that is small and light weight, then the mega with all of the options available today such as GPS, Barometer etc.. My train of thought here is, the palm copter don't require nothing more than stabilization where as when building a VTOL, Servos, ESCs and all of the fun gadgets are preferred.
rimshotcopter is online now Find More Posts by rimshotcopter
Reply With Quote
Old Dec 25, 2010, 08:18 AM
Registered User
Joined Jun 2006
1,881 Posts
great idea i like it
now, i hope we will see it rapidly on the market
titeuf007 is online now Find More Posts by titeuf007
Reply With Quote
Old Dec 25, 2010, 01:22 PM
Friend is a 4 letter word...
JussiH's Avatar
3rd stone from the sun
Joined May 2007
2,670 Posts
This one looks promising:

http://invensense.com/mems/gyro/mpu6000.html
JussiH is offline Find More Posts by JussiH
RCG Plus Member
Old Dec 25, 2010, 01:26 PM
Registered User
berkely's Avatar
Belgium, Flemish Region, coast
Joined Nov 2009
219 Posts
Jussi, great idea about this thread.

I personally don't think the project should evolve to fast to a different C. Like Alex said the basic idea was to have a good cheap flight controller board, something he achieved with success.

But let's face it, despite several good blogs, diagrams & pictures, 70% of the post in the main thread are related to problems hooking the WM+/NK up to the board. Now with different ACC's, baro sensors, LLC ... this will be even worse. If a shield could be designed to have a cheap "out of the box solution" flight controller board with the basic sensors I think much more people will (successfully) chime in.

I think we're all kind of jealous of Warthox his copters & flying videos and he's just using a basic WM+, not even an ACC correct?

So in my opinion, a basic shield with a (stackable) option board to connect more sensors via I2C is the way to go.
I love the simplicity & performance of the MultiWii but if Openpilot comes into the game I (& probably others) will jump to that platform for the extras or .....
berkely is offline Find More Posts by berkely
Reply With Quote
Old Dec 25, 2010, 02:28 PM
Registered User
Joined Jun 2006
1,881 Posts
i think a specifique board design for multiwiccopter code is the best idea because lots of difficulty about wmp..for an extension like pressur sensor,and gyro is not a bad idea but i prefer let this for arducopter or other plateform..(i only like pressur sensor for hold altitude)
titeuf007 is online now Find More Posts by titeuf007
Reply With Quote
Old Dec 25, 2010, 02:34 PM
Registered User
USA, OH, Mason
Joined Jun 2010
136 Posts
Quote:
Originally Posted by 420RcPilot View Post
I have an inquriry on Invensense on getting samples and production quanties. As long as the accel is good and the price is right, it's a no-brainer swap for separete gyro and accel. For me, it would free up space for a baro sensor. If there's enough interest I could make a simple breakout (in Sparkfun style).
andrejk is offline Find More Posts by andrejk
Reply With Quote
Old Dec 26, 2010, 09:25 AM
Friend is a 4 letter word...
JussiH's Avatar
3rd stone from the sun
Joined May 2007
2,670 Posts
Thanks for all the inputs and opinions!

I think we should try and branch this out:

First branch is a sensor-shield to stack with the current 50*50 mm shield. Prototype versions should have pads for all the different sensors people are working with ATM:

MPU6000
ITG3200
ADXL345
BMA020
BMA185
BMP085 or SCP1000 Baro
Possibly also a magneto if a suitable and cheap one can be found!

I am also considering a 3.3 regulator with enable. Power comes from 5V and enable hooked to D12, so we dont loose the hardreset option. Also a LLC to pull all logic to 3.3V!

Prototyping these would be something for "Mad scientist"-types with experience in soldering or reflowing the 0.5 mm pitch found in some of these sensors. Looking at SCP1000 I see it probably cant be soldered by hand, but would have to be reflowed!

Purpose should be to arrive at a cheap i2c 6-9DOF that would integrate the best sensors available and be competitive with wii sensors and sparkfun breakouts! All compatible with 328-based arduinos and current hardware! This we put into production tru seeedstudio or similar!

I have started work on this, and hope to be able to show some designs for it soon!

@ Andrej, I would be very interested in sourcing a small qty of the MPU6000 when it becomes available.....

Branch 2 is the "advanced" version based around an ArduPilot Mega. Although a bit expensive, the APM offers all the functionality we need and more- including very handy onboard PPM multiplexer and failsafe. It has 8 PWMS brought out so okto could become an option with this!

Based around the design for the basic version we then design an all-out i2C IMU board for APM - perhaps also with options for popular analog sensors.

This branch would cater for the people wanting to add "advanced" functions like GPS and waypoint navigation! And be a cheaper alternative to an oilpan!

(I am looking at the option to make a shield that can plug with both the 1.11 shield or an APM)

Since boards are pretty cheap to have made tru seeedstudio in small qtys I think we should just go ahead and get some designs flowing here! No reason not to try and knock ourselves out!

I have no intention of "monopolizing" this process, the decisions or making money on it. I hope we can make it a group effort and use the best of everyones abilities to arrive at a cheap and well-performing alternative to Sparkfun breakouts. I know some of you guys have much more experience and come from engineering backgrounds so your contributions will surely be invaluable!

This is already more work than I can manage in the foreseeable future.
JussiH is offline Find More Posts by JussiH
RCG Plus Member
Old Dec 26, 2010, 09:37 AM
Registered User
Alexinparis's Avatar
Paris, France
Joined Jun 2009
756 Posts
Quote:
Originally Posted by andrejk View Post
I have an inquriry on Invensense on getting samples and production quanties. As long as the accel is good and the price is right, it's a no-brainer swap for separete gyro and accel. For me, it would free up space for a baro sensor. If there's enough interest I could make a simple breakout (in Sparkfun style).
Thats sounds interesting !
I would be very interested in testing such a breakout board.
I don't know what will reply Invensence, but if the cost of a final breakout could be around $25-$30, I think we could have a replacement for Wii parts at the same price and with probably the most important argument: 100% predictable.
Alexinparis is offline Find More Posts by Alexinparis
Reply With Quote
Old Dec 26, 2010, 10:32 AM
OS Sponsor
kinderkram's Avatar
Germany, NRW, Dusseldorf
Joined Jun 2010
5,220 Posts
Way to go, guys!

Starting with something simple (I2C connections and stackable 5x5 sensorboards) to get a flyable combo could be the kickoff for something, most Arduino/ATMega based boards can benefit from.

An all-in-one ATMega1280 with gyros & accels, nearly plug&fly and extensable through additional sensor boards would be a hammer!

Maybe I'm just dreaming... But if someone manages to design such a board and keep the end user price below 100$ I promise they'll be ripped off his hands!

Another board with baro, magneto and GPS for another 100$ would be heaven...
kinderkram is offline Find More Posts by kinderkram
RCG Plus Member
Reply With Quote
Old Dec 26, 2010, 01:15 PM
Registered User
mgros's Avatar
Espaa, Aragn, Zaragoza
Joined Dec 2006
184 Posts
Quote:
Originally Posted by reverendrichie View Post
Absolutely Freaking AWESOME!!!!
Richie, tell us if my Mega porting works fine.
mgros is offline Find More Posts by mgros
Reply With Quote
Old Dec 26, 2010, 02:11 PM
Registered User
rimshotcopter's Avatar
United States, CA, Los Angeles
Joined Aug 2009
11,692 Posts
Quote:
Originally Posted by mgros View Post
Richie, tell us if my Mega porting works fine.
I am going to load it up today!!

Thank you greatly!!!
rimshotcopter is online now Find More Posts by rimshotcopter
Reply With Quote
Old Dec 26, 2010, 02:23 PM
Registered User
USA, OH, Mason
Joined Jun 2010
136 Posts
Let's brainstorm a bit... I wonder if there's a compact modular way to build an IMU. Sensors are evolving quickly and it seems like everyone wants a different combination of gyro, accel, baro, gps, etc, etc.

Some sort of stackable socket would be ideal. Since I2C is a bus, the pins could just be connected together in most cases.

Something like:
Code:



Baro  -->                   |---------| 
Gyro  -->                   |---------|
Accel -->                   |---------|
APM   -->      -------------|---------|---------- 
Plain old .1" headers like the Arduino family use are probably too bulky, but a compact SMT connector could be very dense and stack very thin. Then the whole stack could be secured with standoffs/screws.

Or instead of stacking you could put 3 or 4 of these sockets on a board. This has been done with .1" headers and right-angle sensors but that makes for a bulky configuration. Doing something flat would make this very compact.

I'm just thinking out loud here. Has this already been done? Any thoughts?
andrejk is offline Find More Posts by andrejk
Reply With Quote
Old Dec 26, 2010, 02:54 PM
Registered User
rimshotcopter's Avatar
United States, CA, Los Angeles
Joined Aug 2009
11,692 Posts
Quote:
Originally Posted by andrejk View Post
Let's brainstorm a bit... I wonder if there's a compact modular way to build an IMU. Sensors are evolving quickly and it seems like everyone wants a different combination of gyro, accel, baro, gps, etc, etc.

Some sort of stackable socket would be ideal. Since I2C is a bus, the pins could just be connected together in most cases.

Something like:
Code:



Baro  -->                   |---------| 
Gyro  -->                   |---------|
Accel -->                   |---------|
APM   -->      -------------|---------|---------- 
Plain old .1" headers like the Arduino family use are probably too bulky, but a compact SMT connector could be very dense and stack very thin. Then the whole stack could be secured with standoffs/screws.

Or instead of stacking you could put 3 or 4 of these sockets on a board. This has been done with .1" headers and right-angle sensors but that makes for a bulky configuration. Doing something flat would make this very compact.

I'm just thinking out loud here. Has this already been done? Any thoughts?
Here is a picture of the I2C bus connection on this DFRobot.com Nano shield I purchased. It can handle five I2C devices.
rimshotcopter is online now Find More Posts by rimshotcopter
Last edited by rimshotcopter; Jun 06, 2011 at 10:47 PM.
Reply With Quote
Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion MultiWiiCopter (previously TriWiiCopter) Alexinparis Multirotor Electronics 29927 Yesterday 01:13 PM
For Sale RTF QuadCopters (MultiWiiCopter board) Sid3ways Aircraft - Electric - Helis (FS/W) 8 Feb 12, 2011 04:12 AM
Question Designing RC Control Sets for LEGO Bots volg Robotics 1 Oct 28, 2010 01:34 PM
New Product Cheap(ish) standalone 12v DVR - records to dvd $100 volto Hot Online Deals 0 Oct 05, 2010 03:24 PM