This page serves as some sort of a build log. |
posted Apr 6, 2009 7:59 PM by Todd Berk
The crucial components of the system are now all working together!
Updates: - The accelerometer is providing accurate (sub-degree once calibrated) pitch and roll, heavy filtering was required.
- Compass is providing heading (no yet tilt compensated, working on it).
- Motors are accurately maintaining desired speeds
- Control Systems for pitch and roll are working (just need to be tuned)
|
posted Mar 12, 2009 8:10 PM by Todd Berk
A lot of time and hard work went into designing a new frame to overcome the shortcomings in the Rev1 Frame. The goals of the Rev2 frame were to create a frame which would not only support vertical loads imposed on the arms of the frame by the user, but also the much larger transverse loads imposed by the user (I call them "whoops!" loads). The final design ended up being cut out of high strength .08" 7075-T6 aluminum using a water-jet service. Simon over at http://www.bigbluesaw.com water-jet cutting was helpful and prompt in fulfilling my request for a custom material (6061 is their standard aluminum offering). Overall I was impressed with their service, and the frame arrived in a timely manner after It was ordered. The cut quality was decent, and a little filing on the mating tabs using a leatherman was required due to the tight corners, and the waterjets kerf. The parts were cleaned up using scotchbrite pads to give the aluminum a "brushed" look, and provide a clean mating surface. The frame feels extremely strong when assembled, yet still lightweight. Images below:
The parts after they arrived and cleaned with scotch-brite pads
assembled Rev2 frame
|
posted Mar 5, 2009 12:05 PM by Todd Berk
[
updated Mar 5, 2009 12:18 PM
]
I spent the weekend working on the Pose Sensor, which I just recieved the prototype PCB from BatchPCB (It took about 2.5 weeks). A lot of good progress was made, however, there was one major issue.
The pose sensor contains a relatively new, groundbreaking chip from Honeywell which is used for compassing. It is the 3-axis HMC5843 I2C Magnetometer. The chip integrates 3-axis magnetic sensing, as well as all associated circuitry in a convenient (though tough to solder) package with simple I2C Connectivity. The largest perk is the unit cost, which is much less than any other 3-axis compass I've seen.
I was able to easily communicate with the HMC5843, and verify the identification registers, but the issue I had with working with this sensor was it was outputting bogus ADC readings. After a full day (8+ hours, seriously) I was unable to figure out the issue, which I thought was simply me overlooking something. After some discussion directly with Honeywell, they concluded I had a faulty sensor, and put me on the list to get a replacement one (a few weeks my guess). Major bummer, and due to my impatience I had to order another one from our old stand-by: Digi-Key.
More to come about this once the replacement shows up. On the bright side, the MMA720QT 3-axis Accelerometer works well! |
posted Feb 15, 2009 1:59 PM by Todd Berk
[
updated Feb 17, 2009 12:27 PM
]
A lot of progress has been made over the last two days and the propellers are now spinning! Along with this the Hall Senor RPM feedback has been verified, and elementary velocity control system tests are underway.
However, with success comes a few minor setbacks, mostly due to my klutziness:
- Set tricopter behind me while sitting in spinny chair - first broken motor arm.
- Repeated #1 - second broken arm
- Not what you think. Tested limits of motor from stop to 100% speed with max acceleration, and poor temporary gluing cause motor to thrust out of mount, result? - dent in prop, chopped off motor arm3, broken wires.
Overall a good day. Now back to gluing. (This was with the Revision 1 Frame)
|
posted Feb 4, 2009 12:30 AM by Todd Berk
[
updated Feb 7, 2009 2:45 PM
]
This is the brand new site for the TresCopter Project. It utilizes google sites. Google sites was chosen to utilize a number of features and conveniences it provides over standard content management systems (CMS).
Pro of Google Sites over standard CMS- Ability to focus on content quality rather than site design & programming
- Passive Security & Reputation: Google has a good security track record, and there isn't a requirement to constantly monitor the server for updates to various security elements, i.e. the constant updates required for open source CMS, PHP, Linux, etc...
- Powerful easy to use WYSIWYG editor
- Wiki-like Layout
- Integration with Google Docs (Which I use profusely)
- Ability to share editing capabilities & permissions (though it is basic)
- One less password everyone has to have
Con's of Google Sites |
|
