Dear Friends,
Holger the Creator of Mikrokopter have setup a new benchmark . We need to work hard to improve our project :)
Great Job Holger , if you have an Multipilot1.0 or MK FC 2.0-1 with navi you can check these new features.
http://www.virtualrobotix.com/page/multipilot-8-v10
He use on his fly control a 8 bit micro 644p the same of MP10 , the main difference is that he use only native code write in 'C' not use any kind of arduino wrapper ... so the code is very efficient .
The original Navy is based on a str91x the old version developed in 2006 of STM32 micro so with Multipilot 32 we have more power and possibility to break the barrier of 100 mhz in the next near future.
So now with MP32 we have more power available than in Mikrokopter platform , the code from str91x is portable because we can use the same operating systems and tools , but not source code available for Mk Navy so we need to reinvent the code , better , more original and powerfull i hope :)
Check this docs : http://developers.stf12.net/just-another-eclipse-demo-str91x.
All the member of Arducopter team will be recived in the next days the MP32 , so we're ready to start :)
http://www.virtualrobotix.com/page/multipilot32-1
Best
Roberto
Comments
Hi Michael,
I agree with you, than you say. The MP32 founded on the basis of my experiences on MP1.0 . It mounted the same microcontroller as the Flight Control 2.0. So i know their limit.
I think that already 'MP32 now considering as a simple platform stabilization, is better of FC 2.0-1 for a lot of reason , for example : PWM at 16 bit , Support of i2c/can bus for control ESC 9x-24x fast as main loop.
Now we need to evaluate our point of development with ACM vs MK FC 2.0 + Navi 1.1
Could be interesting to mantain a thread where day by day we can evaluate how is far , near or better the functionality of MK :)
Best
Roberto
The stability of the system is very good in POI follow me mode. MP32 performance may distinguish itself relative to 644p/1280 if its extra computational capability is used for the purpose of precision. Sensor fusion, airframe stabilization, camera stabilization will need to exploit algorithms out of reach of base atmel processors and perform more iterations through main loop per second. The goal should be to provide an increased envelope of performance, operate in higher winds with same or better stability, track ground targets with greater precision, provide data at increased rate and lower latency to future VROS system. The goal of MP32 will be achieved when it is evident that it is providing computional power to achieve an observable benefit that could not be accomplished as well using lesser processors, it will then have justified itself.