Comment by Bill Nesbitt on June 30, 2011 at 7:47am

Maximus: All waypoints are in 3D coordinates

 

ThomasB: I think any STM32F2 code will wait for v6

 

CW: Wind - I can't say I know exactly.  I live in a valley between two mountains and the wind is typically very gusty.  I'm fairly sure the gusts were over 10mph, but I can't say with any certainly how much more.  Take a look at some of the trees or bushes in the background.  Some of them are moving quite a bit.  The horizontal hold can handle a lot of wind and gusts, but they effect the vertical more as vertical is tied to air pressure which changes as the gust moves through.  The error is tempered by the accel integration, but you have to rely on the pressure reading to correct them.

 

vis.asta: v6 information will be posted in another blog entry when I've had a chance to test.

Comment by Giovanni Esposito on June 30, 2011 at 8:40am

..and you did it, again! :-)

Arducopter guys still struggling to release a decent stable version on the old atmega, and you test high speed navigation :-)

By the way, the last slow piroette in position hold is really good (at least compared to what you can do with MK in position hold in windy condition). What are your plans, now? Will you release the new code and hardware specs? Will you produce the board or sell a kit?  

Comment by Roy Brewer on June 30, 2011 at 9:23am

Congrats, Bill! Great work, as always.

 

So, do you think its worth it to go from SMT32 F1 to F2, or would it make more sense to wait for Cortex M4? Also, do you know if the F2 series corrects the bug with hardware I2C (I'm not sure if you're even using I2C)?

 

- Roy

Comment by Bill Nesbitt on June 30, 2011 at 9:36am

Roy: I can find no reason to use the old STM32F1 instead of the F2. They have improved too many things. I was going to wait for the M4, but until when? And if it is missing the optional FPU (like Freescale's Kinetis) then there would be little advantage in waiting. I am not using I2C but I don't remember seeing anything about it in the errata. Of course it is still too early to know all of the problems with the F2, but I have proven that it does at least as much as I need.

Comment by robert bouwens on June 30, 2011 at 11:12am

the m4 will be available from nxp in october.

it is a dual core mcu :)

Comment by Lew Payne on June 30, 2011 at 10:30pm

Sort of like this blog announcement from November 2010...

Comment by Roy Brewer on July 2, 2011 at 6:47am

Thanks, Bill.

 I was excited when I learned of the Kinetis, only to be disappointed when I found out it currently lacked the FPU (although I think their latest series will have one). I did acquire a Piccolo Control Stick from TI. It has an FPU and a "control law accelelerator", which can run floating point control loops (but can only be programmed in assembler). I might try to use it in a quad.

 

- Roy

 

Comment by valentin colon on July 2, 2011 at 8:55pm

awesome!!!

Comment by valentin colon on July 2, 2011 at 8:55pm

awesome!!!

Comment by John Hanson on July 2, 2011 at 11:30pm

Wow!!!!!

Comment

• ‹ Previous
• 1
• 2
• 3
• 4
• 5
• Next ›
• Page