new board

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

new board

nisma
Controller board with
 3xgyro(6adc,
optional)/3d-accelerometer(3adc)/2xpressure(2adc)/voltage(2adc)
using the dsPIC30F5011.
The developement SW is available for free (C compiler) and the micro
features one 12 bit adc. Using a range of 3.3V, this allows a dynamic
range of 200 or 100 degr/sec for the gyroscope using respective 300 or 150
deg/sec gyros.

The analog 3.3V and 5V is regulated with OP´s scaling the 2.5V reference
by a factor of two and routing the voltage trought a transistor in order to
get a 4.8V output and clamping the resulting voltage with 2 diodes
to 3.4V (si diodes). For the gyros, the 7v input voltage is doubled
externally, stabilized and applied to the gyros directly.
This allows to place the gyro to sleep
between the readings allowing a reduced power consumation.

The main idea is to have a coprocessor, that handles all the communication
interface (gps, mmc, servo, sensors) and is connected as slave to a
spi interface (max 5MHZ clock rate).
The main MCU (arm, avr, ... ) is connected trought spi and query every
(preprocessed) sensor data, makes some routing decision and command
the coprocessor.

>From the controller point of view, the idea is to use the gyro/accelerometer
as reference (100 HZ) and to reset it every (0.25HZ)
with a new gps coordinate. Further, the gps acceleration is used to
exactly divide the gravity and acceleration for the accelerometers
improvering at the same time the gyro performance.

When using dgps, the algorithm position estimation algorithm should
additionally recalibrate automatically the gyros in the same way the
gps recalibrate it´s internal clock for temperature compensations.


If someone will join in order to influence the specifications,
 please write a mail. As you can see, this design is not suited for
using the cheap thermopiles (5Euro/piece). Instead it´s should be a
platform for a pure gyro stabilized airplain.


--
Echte DSL-Flatrate dauerhaft für 0,- Euro*!
"Feel free" mit GMX DSL! http://www.gmx.net/de/go/dsl


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

AW: new board

Ethan Arnold-2
Hi,

Is the first part of your mail missing? I'm not sure who you are (a name
would be polite, in the least), if you are offering a finished product for
sale, or currently building something, or planning on developing something?

Regards,
Ethan

--
Sim-Zone Flight Simulation
Voltastr. 5
65795 Hattersheim
Tel: +49 (0)6190-975-374
Fax: +49 (0)6190-975-626
www.sim-zone.de


-----Ursprüngliche Nachricht-----
Von: paparazzi-devel-bounces+ethan=[hidden email]
[mailto:paparazzi-devel-bounces+ethan=[hidden email]] Im Auftrag von
[hidden email]
Gesendet: Donnerstag, 6. April 2006 11:42
An: [hidden email]
Betreff: [Paparazzi-devel] new board

Controller board with
 3xgyro(6adc,
optional)/3d-accelerometer(3adc)/2xpressure(2adc)/voltage(2adc)
using the dsPIC30F5011.
The developement SW is available for free (C compiler) and the micro
features one 12 bit adc. Using a range of 3.3V, this allows a dynamic
range of 200 or 100 degr/sec for the gyroscope using respective 300 or 150
deg/sec gyros.

The analog 3.3V and 5V is regulated with OP´s scaling the 2.5V reference
by a factor of two and routing the voltage trought a transistor in order to
get a 4.8V output and clamping the resulting voltage with 2 diodes
to 3.4V (si diodes). For the gyros, the 7v input voltage is doubled
externally, stabilized and applied to the gyros directly.
This allows to place the gyro to sleep
between the readings allowing a reduced power consumation.

The main idea is to have a coprocessor, that handles all the communication
interface (gps, mmc, servo, sensors) and is connected as slave to a
spi interface (max 5MHZ clock rate).
The main MCU (arm, avr, ... ) is connected trought spi and query every
(preprocessed) sensor data, makes some routing decision and command
the coprocessor.

>From the controller point of view, the idea is to use the
gyro/accelerometer
as reference (100 HZ) and to reset it every (0.25HZ)
with a new gps coordinate. Further, the gps acceleration is used to
exactly divide the gravity and acceleration for the accelerometers
improvering at the same time the gyro performance.

When using dgps, the algorithm position estimation algorithm should
additionally recalibrate automatically the gyros in the same way the
gps recalibrate it´s internal clock for temperature compensations.


If someone will join in order to influence the specifications,
 please write a mail. As you can see, this design is not suited for
using the cheap thermopiles (5Euro/piece). Instead it´s should be a
platform for a pure gyro stabilized airplain.


--
Echte DSL-Flatrate dauerhaft für 0,- Euro*!
"Feel free" mit GMX DSL! http://www.gmx.net/de/go/dsl


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel



_______________________________________________
Paparazzi-devel mailing list
[hidden email]
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: AW: new board

nisma
Hi,
sorry for the inconvenience, the first part of the mail is really missing,
in short i have designed such a system with a arm processor (lpc2000)
and tested/verified some things at the begin of the year.
Because i want/need better adc resolution at low cost, i´m currently make a
redesign with a different processor and new concepts derived from
lessions learned with the last version. Switching to the dspic have some
other commercial advantages, from RTOS up to the voice recognition library,
that don´t really matter for a open source design.

For usb triggering, low cost wifi connection, ... i will resell other
products that can be integrated with the one board.

I plan to release this product in this year in two different versions,
one with custom firmware and annother with sourcecode.

The difference between the two firmware will be the support and some
different features like bytecode interpreter, voice recognition features,
support for transparent communication channel encryption and similar
functionality.


I´m located in italy and i own a small company. More later.

Regards
Christoph.

--
Analog-/ISDN-Nutzer sparen mit GMX SmartSurfer bis zu 70%!
Kostenlos downloaden: http://www.gmx.net/de/go/smartsurfer


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
Reply | Threaded
Open this post in threaded view
|

Re: AW: new board

nisma
In reply to this post by Ethan Arnold-2
Hi,
i´m planning to produce a small uav board having 3d accelerometer,
3x gyro, 2x pressure sensor and one voltage sensor.
I´m currently unsure, if using the digital (spi) gyro sensors and
using the two spare adc´s/channel for pressure and one accelerometer axis
using the spare cpu power inside the gps
or using the lower costs analog gyros and a dspic processor. Both solutions
have 12 bit accuracy. Further for some sensors, like pressure
the resolution is increased adding a pwm generated bias for selecting
the measurement window inside the sensor range.
The board will be 4-6 layers. Because the gps sensor can output
the raw gps data at 50-100 HZ, without processing it.
The source code for processing and kalman filtering is provided, so i
actually search for a optional external cpu for the gps position extraction
at higher update rate. Any of you´r smart guys have a suggestion in this
area ? Actually my preference is to use the digital gyro sensors and
additionally using one digital spi inclinometer (mems - heat based - 1hz
update rate) because it´s cheap and helps fusing the two accelerometers
in order to reduce the drift and vibrations errors in case of non electric
engine. Actually, no active magnetometer is planned, because it´s possible
to use the gps with really better results, and as gps backup, the small
spi 3d-magnetometer from the pni corp can be used placing it far away from
the electric motors and servo.

Any comments will be welcome. If someone is interrested in this developement
or want to be a beta tester, don´t esitate to contact me directly.

Regards
Chris

--
Echte DSL-Flatrate dauerhaft für 0,- Euro*!
"Feel free" mit GMX DSL! http://www.gmx.net/de/go/dsl


_______________________________________________
Paparazzi-devel mailing list
[hidden email]
http://lists.nongnu.org/mailman/listinfo/paparazzi-devel