Re: [Paparazzi-devel] HW

[address@hidden]

> Hi Chris,
> I do agree with most of your points regarding
> increasing the budget of the UAV.  If we can achieve
> considerable improvements in design and functionality
> with a reasonable cost increase then so be it.  
> 
> My concern is that many of us have fairly busy lives
> with work and family etc.. The thought of one crash
> forcing us to be grounded because we must re fabricate
> custom components does not sound appealing.
You use four sensors. FMA autopilot has switched to using 5 sensors,
because a lot of airframes are crashed. This happens because the autopilot
has no information, if the model flyes inverted or not.
This is critical, if the model has 40% inclination and a wind approaches,
or the model climp, and the motor has a failure.


> 
> I would hope that we do set standards for hardware so
> that all of us are working together to improve one
> system.  As with any project there will be divergence,
> is it possible to build it in components so that those
> of us who would like to spend a little extra can.
> 
> While those who do not wish to do so can use the basic
> system.  I will let Antoine speak with respect to base
> cost and how we should proceed.
> 
I don't see this with the actual HW.
The smart, cost saving thing is sensor fusing.
Reimplementing/changing the SW is a cost/risk factor.
Making the SW modular requires more SW/HW resources.

> I for one would also like to know if there is an off
> the shelf or kit available for the infrared sensor. 
> If those amongst us who are talented with a soldering
> iron could produce units for a modest fee, it would be
> much appreciated.  This way we can get more people up
> and running and contributing to the project.
I can make it for 35 Euro including shipping (5 IR sensors) + FMA licence
fee with digital outputs (RS232/I2C/SPI), but i need to sell 10-20 units.
I have wanted to do it, but the FMA patent has stopped me after asking
my lawer and the FMA for the needed licencing.
> 
> I'm not sure if any of you have seen this website yet 
> http://www.cyber-flyer.com/
> but if you click on the telemetry part you will see an
> antenna that tracks the guys RC plane.  He can
> actually track his plane while driving.  Very
> interesting, maybe he can provide us with the software
> and setup he used.  I will e-mail him.


Seriosly, what is the difference of the actual project status, and
putting the servo controller with GPS receiver piloting the servos
with simple pids using angle/speed limitations.
The FMA autopilot is placed between the servos and the controller.
This is the simplest solution.  Instead of the FMA autopilot, 
a RC gyro can be used. Someone will already own this type of equipment.
In my eyes, there is not a difference.
The two CPU solution is a too heavy for such a simple solution, even
if the IR sensors are handled directly by the servo MCU.

I don't know the final plan for the paparazzi project.
Using Cots parts, this is the price.

100  GPS receiver
150  FMA autopilot or gyro
200  Airframe, batteries ... 
50   Servo controller.
----------------------------
500 

For a RC activist, the cost is only 150 for the gps and the controller.
The servo controller can be made for less. This is only a really big
estimation.

Reimplementing the FMA autopilot don't save mutch, because the additional
Mega128 cpu and PBC. Annother thing is, that this type of UAV don't can be
selled without a FMA autopilot for patent issues.

Adding more sensors implies redesign the software or adding a
high processing power board like this one. www.gumstix.com.

I don't know the real goal of the paparazzi project.
If is to support cheap UAV project letting them to add more sensors or
payload, the way to make a generic gps controlled pid/servo controller, that
can be used as servo controller and that can switch/fly autonomous the UAV
is better. If the goal is sensor fusing and developing control strategy's,
sensors are important and this should be discussed. Beside the prestige of
having a uav, most of the UAV builder have a goal. This can be areal
photography, boarder control, archeological (magnetometer) measurements and
so on. The goal is important for choosing the right sensor package.

Chris


-- 
+++ Jetzt WLAN-Router für alle DSL-Einsteiger und Wechsler +++
GMX DSL-Powertarife zudem 3 Monate gratis* http://www.gmx.net/dsl