Comment by Stephen Carlson on August 30, 2012 at 12:58am

A related note: Some Baro Sensors have the MEMs element open to ambient light. Just having the light levels change, as in more or less sun pouring into the sensor port, is enough to cause errors of several meters. This may not be so much an issue now, but I know that in 2005 when Procerus was just getting started, they spent many hours trying to figure out why the plane would rise and fall through an orbit with pretty fair consistency. They found that the light striking the element inside was causing a charge buildup, thus throwing off the readings as the plane turned into and then away from the sun.

Comment by Drone Savant on August 30, 2012 at 6:14am

Thanks for sharing! This explains and unexpected near fly away that I had a year or so ago on a replacement Oilpan that I did not tape up before I tested it on my APM1. 

Perhaps a bit of extra text in the Wiki is in order at http://code.google.com/p/arducopter/wiki/AC2Assembly in the "Covering the barometric pressure sensor" highlighting the end result? Explaining what actually happens due to failing to tape it up would be helpful. Obviously this would be useful to be in the APM2.5 documentation as well. "badly throw off the altitude readings" does not exactly highlight what can occur as a result of said bad readings. Not all of us can put two and two together right off the bat. 

Cheers to adding the following: 

Note: Because the APM 2.5 board does not have a daughterboard covering the barometric pressure sensor (the small silver rectangular sensor in the image above), it's important that the board be mounted in an enclosure to protect the sensor from airflow, which can badly throw off the altitude readings. If you choose to mount the board out in the open, make sure to cover the sensor with some cotton padding or some tissue.

 

You've literally answered something that has been in the back of my head for about a year now. =] I actually haven't tested ALT_Hold since then for fear of my gear trying to fly away. 

Comment by Joseph on August 30, 2012 at 6:15am

This answers so many questions. My quad kept trying to fly away to when swapped to Alt hold. Thanks for the information.

Comment by R. D. Starwalt on August 30, 2012 at 6:25am

I've got quad frame envy Chris, please tell me which one your 2.5 is on.

Comment by Drone Savant on August 30, 2012 at 6:54am

Looks like Chris is flying a DJI F330 or F450 based on the picture.  http://www.dji-innovations.com/products/flame-wheel-multi-rotor/fea... 

Given the hole pattern (4 bolts vs 3) I'd go with F450 http://www.readymaderc.com/store/index.php?main_page=product_info&a... vs. F330 http://www.readymaderc.com/store/index.php?main_page=product_info&a... as my guess. 

They are pretty solid frames, you can nerf em pretty hard and they won't crumble. $26 - $31 for frame including PDB. The arms do have an interesting weak point but again the whole frame plus PDB is <$35 

3D Robotics Comment by Chris Anderson on August 30, 2012 at 7:53am

Nope, just the $17 Hobbyking frame. (Flamewheel clone) Works great!

Comment by LanMark on August 30, 2012 at 8:43am

Chris.. why is your quad missing screws where it connects the arms?

3D Robotics Comment by Chris Anderson on August 30, 2012 at 9:26am

Mark: because the Hobbyking frame is poorly made and half the screw holes don't line up. But it works fine with just half of the screws. It's a pretty solid design. 

Comment by Marooned on August 30, 2012 at 10:36am

I also have Q450 frame but without built in PDB and all screws fit very well. One issue is that screws are made from soft metal and has small hex socket. Because of that it's easy to smooth this hex into "circle" making them unable to be removed. Be delicate to those screws.

Comment by Jack Crossfire on August 30, 2012 at 5:39pm

Not unlike this guy, who went overboard with it, 3 years ago:

Hanging the barometer under the fuselage, out of sunlight helped, but really a better position sensor is needed.

Comment

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page