Re: [Paparazzi-devel] ESC Startup

[Martin Mueller]

Hi Neil,

> Some of you may have heard that Gumstix is developing a UAV board
> based on Lisa/L and RoboVero. The rumours are true!
>
> First of all, thanks to everyone who has offered input so far. I have
> a couple more questions - hopefully this is the right forum.
>
>   * It seems like the Lisa/L makes use of several LDOs - should we
> beware of SMPS in UAV applications?

except for the Lisa/M all Paparazzi boards have a switching power supply 
to get the voltage down to 5V followed by one or more linear 3.3V 
regulators. If there is more than one its mostly to separate "listeners" 
(sensors) from "loud parts" (e.g. transmitter).

>   * I'm told that ESCs need to have a PWM signal when power is applied
> for calibration. How can this be achieved if the control electronics
> are powered by the ESCs?

It is not a question of ms until the ESCs complain about not having a 
signal. Paparazzi initializes the servo/controller PWM outputs with a 
default value very early after reset (modem initialization can take up 
to 2.5s) and then outputs the wanted signal when everything is up.

> Thanks in advance. Will keep you posted on our progress.

Did I understand correctly that you want to use a LPC1769 for the board? 
Is the NXP CMSIS driver library you use open source (GPL) compatible?

There is nothing wrong with that processor (family) but please take into 
account that writing the low level drivers and adapting it to the 
Paparazzi driver layer took a huge amount of time/effort in the past. We 
had that from ATMEGA128 -> LPC2148 -> STM32F10x and are not completely 
finished with the last step as the STM license turned out not to be open 
source compatible (yet sounds very similar to the NXP license).

What keeps you from using one of the well known STM32 processors (with 
GPL V3 libopencm3 support)?

Martin