Here it is! Finally, the PixHawk Fire Cape (PXF) is ready for a wider audience and we've launched a crowdfunding campaign so that everyone can get theirs. We've been iterating through different generations over the last year and with the help of many, we are finally proud to announce the PXF starting at $100.
This board was born out Philip's hands and pushed by members of this community so it's just reasonable to try making it as affordable and accesible as possible. Some technical details:
Sensors
- MPU6000: 3-axis gyroscope, 3-axis accelerometer and temperature sensor.
- MPU9250: 3-axis gyroscope, 3-axis accelerometer, 3-axis magnetometer and temperature sensor
- LSM9DS0: 3-axis gyroscope, 3-axis accelerometer, 3-axis magnetometer and temperature sensor.
- MS5611-01BA03: Barometer that includes pressure and temperature sensors.
Connectors
- 3x LED indicators (Green, Amber and Blue)
- 2x serial UART ports (ttyO0 is not active)
- 1x CAN connector and transceiver
- 3x I2C ports
- 1x Buzzer
- 1x Safety switch
- 9x PWM output channels
- 1 PPM/S.Bus in
- 1x Spektrum
- 1x Power brick connector
- 1x Battery backup (1 LiPo cell)
- 1x ADC
- 2x GPIOs exposed (IO)
- 1x analogic pressure sensor (AIR)
Mechanical characteristics
- Size: 88.6 x 54.73 x 20.69 mm
- Layers: 6
- PCB Thickness: 1.62 mm
- Weight: 31 grams
Support the crowdfunding campaign of the PXF here.
Comments
@robert yeah, indeed PRUs code is particular for the units but I'd point out that RCOutput and RCInput (which is what the PRUs are taking care) has been ported to different boards and architectures (source code pointer).
Code is being structured in a smart way so that portability is always favoured. In that sense i'm pretty happy with the AP_HAL_Linux.
To put it in context, take Mirko's commits and his recent to support BBBMINI. I believe the code is pretty portable.
Victor - I signed up for early bird.
It will be interesting to test this out in a real world application, i work with a company that is using pixhawk for commercial mapping using planes.
>We generate PWM signals for the motors with one of them and process RC receivers PPM signals with the other.
interesting to look at.
but it is not portable at all ;-)
On a side note, it's amazing the amount of guys/companies that get in touch with you offering their advertising services! Totally overwhelmed by private messages.
Thanks for the support guys!
@Julian, NuttX is great and Greg does a great job keeping it constantly updated and bug free. Having said that, the learning curve of it (such an RTOS) is really steep. Not even thinking of hacking APM on the STM32F4 microcontrollers. Really amazing work what Tridge, Randy and rest of the developers do given the complexity of the code.
Our work through the last year proved that it was possible to "tune" Linux to respond to the needs of an autopilot and provide great chances of scalability (with frameworks such as ROS). Running APM on Linux is still experimental but I'm truly convinced it's the future for drones and it's improving everyday.
@Wilhelm, we have been flying with a single IMU and barometer so it's indeed possible however using redundancy makes the estimation much more robust when it comes to vibrations and errors in the sensing.
@Wilhelm, It does :)! We generate PWM signals for the motors with one of them and process RC receivers PPM signals with the other.
yes, backed up.
Hi Victor,
Just signed up for an Early Bird, hope you make your minimum.
In response to the queries above, the Pixhawk Fire cape was developed as a sort of joint project by DIYDrones, 3DR and Victor's people.
It is open source and open hardware.
At this point in time it has Pixhawk compatible Linux code that will let it do (pretty much) everything the Pixhawk does.
In addition there is a port of ROS (Robot Operating System).
The cape actually has a bit more capability than the Pixhawk.
But the real pluses are the ability to have a much more powerful processor and potential uses well beyond the Pixhawk capability without requiring an auxiliary computer as well as being Linux programmable.
It should be considerably easier to design in extensions and new capabilities than is the Pixhawk and should have sufficient headroom for them.
This is really the first serious implementation of an autopilot in Linux.
And while it makes programming generally easier, Linux is not optimized as a real time operating system.
There can still be some odd timing issues that crop up every now and then so at this time, this combo is better for developers and early adopters who are able to deal with glitches.
That said, the potential for adding serious extra capability and ease of programming makes it very attractive.
My take anyway, please feel free to correct or put in your own thoughts.
Best Regards,
Gary
Let's hope this cape will use BBB's pru units.
Why so many accelometers, gyroscopes and temperature sensors?
Wouldn't one of each be sufficient?
ha but thats not going to stop me getting one and testing it out :)
-
2
-
3
-
4
-
5
-
6
of 6 Next