Re: [Paparazzi-devel] HW

[nisma]

Hi to all.
I will raise a little discussion or hear you'r feeling about
price/functionality of a UAV similar to paparazzi.

All the price discussion leaves out the RC TX/RX and laptop equipment 
required to operate the uav.

It's possible to build a full featured UAV (paparazzi) for 500$ having TV
quality live images with options to carry minimal other payload.

With more then the double price, 1000 Euro, it's possible to build a
advanced UAV with spread spectrum RF modems, higher imaging quality,
possibility to add visual processing, having redundand sensors.

With 1500 Euro, it's possible to have better sensor resolution, real gyros
+ accelerometers and so one.

Using LIPO batterys, 40 minutes flight is not a problem. 
This gives a theoretical 10-15 Km range.
Only this fact is amazing, having real time video.

I know, that if building some components itself, the 1000 Euro UAV can be
made at 700 Euro. It's the question, if one should use COTS parts or 
developing the subsystem itself. 
Hacking a CDROM motor to a brushless motor gives big cost savings,
for other thing, the story can be different. One example can be the air
frame.

Dont misunderstand me, i'm not telling, building a low cost UAV
with `limited' functionality is wrong. Building UAV is funny. I 
seriously ask myself and you, if it's better to build a 400 Euro UAV
, than building one for 1000 Euro with many more possibilitys.
I'm not interrest in building a UAV bussiness for lowest cost UAV.
Building a low cost UAV, expanding it later to higher capability, better
sensor processing involves developing the UAV two or tree times.
The hard part is sensor fusing and controlling the UAV.

Annother thing is, there is a high risk, that this project spin off like
the autopilote one. Ok, the open source autopilot project don't work
reliable, the paparazzi make the job a lot better.

As example, adding pressure sensors will increase the stability noticable.
Using the usual motorola part, means a price addition of 20%.
The stability is not only in the Z axis. The atmospheric pressure can be
used to evaluate the overall gps precision. If there is a noticable
error on the Z axis, it's highly probable, that the x and y axis are not
accurate. As a result, with more gps error, a minor weight can be added
to the error value. CA grade GPS updates the Z axis only at 1HZ interval.
Additionally, this gives stability, if the GPS signal is absent for some
time, and simple failsafe, in order to don't go down at to high speed.
 
Using the pressure sensor have influence to the altitude software.
In order to recover the UAV, if less than 4 GPS satellite are in field,
using 10 degree apogee instead of the default 5 degree for higer precision,
the altitude must be measured relative to the start altitude, in order to
land without GPS altitude information.
The simple approach of storing the start altitude don't work, because the
pressure is not absolute and must recalibrate during flight at certain 
intervals. If the uav, after one hour return and must land using the 
pressure sensor, the software is never able to recalculate the pressure
from the stored height value. Storing the initial pressure don't takes
in account the recalibrations.

Depending on the planned usage, each of the list member has other
constraints and needs. It's nice, if the base UAV can handle most of the
features wanted. 

As example of different features for the same goal.
You want higher GPS update rates in order to improve the overall GPS
performance . On higher update rates, it's possible to apply better filter
in order to get good coordinates.
I want better precision from gps unit with minor update rate.
The same goal, better GPS precision, opposite features wanted resulting
in different firmware. What solution is better ? i don't know the response.

chris






currently i review my uav.
I have found 19k2. full approved bidirectional RF Radios for 100$ each ( 40$
in 1K quantity). The range is more than 14Km.
Low cost 1Hz GPS with code for areal mode that go into 2D mode, if only 2-3
GPS satellites are visible and capable of generating DGPS infos. 100$ /
each. It's capable of giving 3D velocity values.
3D compass, pressure sensors, little MCU , IR sensors 100 Eur
bigger cpu 100-200 Euro
high resolution camera with webcam function 100-150 Eur

communication: 200
position       200
sensor+MCU     200
Imagery        200
               ----
               800

building two, one without imagery acting as RF relay for short range RF
modem = 1100

600 Euro / UAV
200 Euro / Control station + laptop/computer.
200 Euro / more processing power.

I can build one for 1000, two for 1500

No DGPS, 200+200+400+200=1000 Euro.

UAV:
100 GPS
100 Sensor
120 2Mb USB camera
140 200MH Xscale CPU
200 Model + servos
100 Modem
60  Battery
-----
800 Euro

Base:
100 Modem
100 Support logic + Joystick or RC interface for manual flight
    Computer
-----
200

Optional:
+50 400Mh Xscale CP
+100 DGPS
+100 3M camera
+160 4M camera with triple optical zoom.


This is twice the target price.
Removing the higer resolution camera and higher processing power, 800 Euro
remain.

Plain paparazzi solution:
UAV:
200 Frame + servo
100 GPS
100 Sensors+MCU
150 TV trasmitter/receiver
---
550 

Base
100 RC TX

Without RC TX/RX the 500$ target is archieved.

-- 
"Sie haben neue Mails!" - Die GMX Toolbar informiert Sie beim Surfen!
Jetzt aktivieren unter http://www.gmx.net/info



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