Re: [Paparazzi-devel] Discrete power supply for servos and autopilot

[Tilman Baumann]

On 05/04/12 11:11, Gareth Roberts wrote:
> Hi Tilman,
>
> What about something like this? http://www.national.com/pf/LM/LM2678.html
Does not quite go low enough I would say.
> Buck-boost should be ok as long as you have sufficient filtering on 
> the output.
That was my thought. But you automatically get inductors in your design. 
I hate sourcing them.

> The critical stuff is all fed off the 3.3V reg anyway, which is linear.
Exactly. I will even try to get a smooth 5V supply. For all the other 
electronics. Telemetry, GPS and so on.
> The RC receiver will be pretty noise tolerant as it's designed to run 
> off cheap BECs in ESCs, most of which are switched-mode anyway.

I was thinking 5A 5V unfiltered here. More might perhaps be better, for 
complex planes with more servos?
What do experienced RC-ler expect from their BEC?
> I've used ppz with two cells before, although it usually implies an 
> aircraft so small you wouldn't want a mount like this.
Still, I like the idea to be universal. I will consider Buck-Boost again.

TI has a nice powersupply design tool. I can simulate a lot there.
> As for the simplest option:
> Looking at this: 
> http://paparazzi.enac.fr/wiki/Lisa/M_v20#Powering_the_Board
> you drive the servos seperately from the autopilot normally.
> If you close the jumper JP1 you just need to connect 5V straight to 
> the servo side of the Lisa (the BEC in your ESC would do fine, 
> assuming it's a stock Bixler).
> That way you don't need any external components or additional wiring 
> at all.
I would like to keep them separate.
> Diydrones also do something similiar to your proposal w/o electronics: 
> https://store.diydrones.com/Bixler_SkySurfer_Chassis_No_Pan_Tilt_p/bx-chassis.htm

Hehe, that is exactly where I got the idea from. :)


Cheers
 Tilman