To Sebastian and others, Mikrokopter guys also moved away from I2C due it's not reliable enough.
We also have been having a lot of problems with I2C. There seems to be some fundamental problems on Atmel I2C implementation and after you add certain amount of I2C devices to the bus, it hangs with random time.
If PWM esc dies especially on hexa, it does not matter so much due others can compensate. If PWM/Motor dies on Quad, you are in trouble as we know. Ok if you are skilled pilot you can still land your quad with minimal damage/no damage at all. But it's a big IF....
I2C is good but I don't trust it anymore so much as I earlier did.
@I.S.: these cons are not as big as PWM ones (sloooooow)
you can not use those 2 outputs as fast outputs because they are using a different hardware timer (just a guess, may be wrong), so you can't make them go faster
Having built an Okto based on the early AC code I am not sure about these issues with the I2C bus. Yes it can and does fail but so can there be many other sources for failure. Yes if it brings the whole bus down then all motors stop and the craft plummets to earth. With a quad or hexa what happens if a pwm ESC fails?
I have looked at converting the new ACM code to I2C and okto but it is nowhere near as easy as I had hoped it would be and it was in the early AC code.
Comments
To Sebastian and others, Mikrokopter guys also moved away from I2C due it's not reliable enough.
We also have been having a lot of problems with I2C. There seems to be some fundamental problems on Atmel I2C implementation and after you add certain amount of I2C devices to the bus, it hangs with random time.
If PWM esc dies especially on hexa, it does not matter so much due others can compensate. If PWM/Motor dies on Quad, you are in trouble as we know. Ok if you are skilled pilot you can still land your quad with minimal damage/no damage at all. But it's a big IF....
I2C is good but I don't trust it anymore so much as I earlier did.
Hi Sebastian,
thanks for the link´but I already have the code, but I am not sure if I can use it on my configuration.
@ Peter Seddon
Do you think I can use your Okto code?
Is there a primer on how to design and build a conning tower, or canopy as it's called?
@I.S.: these cons are not as big as PWM ones (sloooooow)
you can not use those 2 outputs as fast outputs because they are using a different hardware timer (just a guess, may be wrong), so you can't make them go faster
I have looked at converting the new ACM code to I2C and okto but it is nowhere near as easy as I had hoped it would be and it was in the early AC code.
How about this http://www.aplanding.com/forums/showthread.php?t=5033
Peter
@Sebastian
Both have big cons.
I would really like to know why we cannot use more Fast PWM available on the ATmega
you could use I2C (like the Mikrokopter guys do)
or UART, like Jani said
Damn it! joint coax looked so easy.
So, how can we get X8 counter coax with APM hardware?
@Jani
What issues the micro has at output5-6?
even the ATmega2560, just the 1280, or both?
what FastPWM (further literature)?
Thanks