Welcome

My virtual playground. Meer info

Micropilot MP2028 Review - Part 1

I’m lucky enough to have the opportunity to “play” a bit with the Micropilot autopilot module and software that comes with it. While going through all the steps before we have our MAV in the air, I’ll do my best to post some of my experiences on my weblog. Part one! A first glimpse at the autopilot.

The module itself looks very professional: All paths are in an inner layer, so none is visible. On top of that, a special coating layer is covering the PCB.

Some chips on the module caught my attention:

  • Sipex UART to RS232 on-board : Isn’t it odd that they use RS232 on the module and not just TTL to communicate with the modem (which also contains a RS232 to TTL chip) and PC?
  • MAX660 Switched Capacitor Voltage Converter I have no idea what they use that for
  • Cirrus Logic AD convertor This is just a normal AD-convertor
  • LCX14 Standard hex invertor
  • A 49.1uH inductor . I suppose that is for filtering? I’m not familiar with other uses of an inductor…
  • 3 gyroscopes and 2 accelerometers. The Analog Devices ones you find in most IMU’s.
  • One static and one dynamic pressure sensor. Also looks very standard.
  • XC9536XL High Performance CPLD This is a Xilinx FPGA. High performance, low power.
  • M410000025: This is RAM. I have on idea how much…
  • The main processor is a FreeScale one. I didn’t rip off the sticker on it to see which one :-)
  • Commonly used TIM GPS module from u-blox.
  • The servo board contains the 17HCT237 to demultiplex a 3-bit data input to 8 data output (to the servos)

Most of these components are pretty standard, but it is clear that this module has quiet some processing power!

The weight of the module is only 28 grams. Unfortunately, the standard GPS antenna that comes with it is 38 grams! The interconnecting cables that come with it aren’t made for weight-saving either. So if we want a really low weight, we’ll have to come up with some ideas ourselves. The manual refers us to the website for solutions for these kind of issues, but I haven’t gotten my Support ID yet to log in.

Speaking of the manual: it looks very complete (160+ pages) and there is even an instruction video explaining the basic setup procedures. Compared to some other autopilot modules i’ve seen, it all looks very complete and professional.
Every parameter of the numerous PID loops it contains is configurable, including min and max settings, which PID loop to use at what speed and some others I never even thought of myself (and still need to find the use of it ;-) ). For automatic take-off, there are also a lot of options.

The back of the module, with the servo board next to it:

The front of the module, with a small part of the cables mess to interconnect everything:

More to come! If you’d like me to cover a certain topic of the micropilot, let me know!

17 August 2007, 05:40 | Link | Comments [14]

Autopilot phase 1 (stabilization & servo controller)

I’ve been slowly working on my own autopilot system for more than half a year. I spent most of my time gathering knowledge and components. The last few weeks I worked a bit more intense on the project and it paid off: the stabilisation and servo controller are finished. They have been tested successfully and proved to be reliable!

Schematic:

Now, what can my module do:

  • Calibration on startup (defining min and max values for the sensors, and the neutral point)
  • Manual mode: Pass transmitter signals to the servo’s
  • Stabilized mode: Read the desired roll- and pitch angle from the transmitter sticks and stabilize the plane in those desired angles. This stabilization works with thermophiles. They sense the temperature difference between the ground and the sky.
  • (mode is determined by a slider on the RC-transmitter)

Other features:

  • Delta mixing is done in the microcontroller so it’s not required on the transmitter (makes it easier to read pitch and roll input signals seperately)
  • DSP filtering. The PPM signal is taken from the receiver before any DSP-filtering is done, so we need to do it ourselves to eliminiate glitches. Features:
    • Checks if every PPM frame has the same number of channels.
    • Checks if every pulse in the PPM frame is within a valid range (1ms and 2ms).
    • When a bad frame is received: keep the last valid positions. Go to failsafe positions after 2 seconds.

Planned features:

  • Read input from UART instead of RC-transmitter (interfacing to other microcontroller or PC)

I tested the stabilization unit on a small delta wing (40cm span) because I see no use in testing it in an easystar :-)

Module on the bottom of the wing:

Sensors on top of the nose, slightly pointing up:

Now some showing off :-)
The stabilization works so well I can launch the plane with the transmitter on the ground! When I give full left, the plane understands this as “go as much left al long as you can still see the horizon”. Considering the IR-sensor lens of 100 degrees, this will probably be about 40-50 degrees.

15 July 2007, 13:34 | Link | Comments [10]

New inertial sensor

Analog devices has created a new inertial sensor, the ADIS16355. It includes a 3 axis high precision gyroscope, and 3 accelerometers. On top of that, it includes digital I/O with some on-board basic sensor filtering. It’s target market are dead-reckoning applications such as car navigation when a GPS signal is lost.

Being a cube with a side of 23mm, it seems kind of big. Unfortunately no weight specs on the datasheet yet. For the very small MAV’s this might be too big, but for most autopilot systems this sensor will improve navigation performance!
Samples will cost $350, +1000 will be $250 per unit.

4 July 2007, 16:23 | Link | Comments [2]

Connecting a GPS (EB-85) to a PIC microcontroller

Last month I finally bought a GPS module. I decided on the EB-85 because it’s cheap (thanks to mr. rc-cam!) and it has a 5Hz update rate!

However, this module has an UART that operates at 2.8V. My PIC runs at 5V. After some investigations I figures out that the TTL pins of my PIC would interprete the 2.8V as logical high, but the ST-pins (Schmitt-trigger) won’t. Because the hardware UART or PIC uses ST, I was left with 2 options: Use the TTL-pins and a software UART or continue looking for another solution :-)

What I need is a TTL chip that has en open-collector output. TTL accepts 2.8V as a logical high, and the open-collector port allows me to use any voltage as logical out if I use a pull-up resistor.
My solution uses the 7407 IC, which is a “Hex Buffers/Drivers With Open-Collector High-Voltage Outputs”.

This is the schematic:

On the left you see the PIC, on the right you see the connector to my GPS. In the center-top you see the pull-up resistor to 5V, in the center-bottom you see the pull-up resistor to 2.8V that consists of a voltage divider that yields 2.8V.

On my test-board it looks as follows:

This is what my test-program does:

  1. Startup
  2. Send a message to the GPS so only the RMC (Recommanded minimum) messages come through:
$PMTK314,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0*29\r\n
  1. Every time a byte is received on the UART, an interrupt is launched and the received data is decoded.
  2. The main loop prints the data in the LCD

The overhead of the UART and GPS decoding is very low on my 4Mhz PIC. Perfect!

Here are some nice sites for electronics newbies like myself:

Update:
There was a slight error in my schematic: I played the voltage divider before the buffer instead of after it. It is corrected now.
This should also work (but my 7407 chip has 6 buffers so I didn’t care):

30 June 2007, 15:25 | Link | Comments [11]

Extremely basic autopilot

I found an interesting link while surfing the net (well, doesn’t it always happen like this ;-) )

It’s a rudder-driven autopilot with waypoints hard coded, but works with a PIC16F877
Very basic, but comes with pic-code and a very simple circuit.
http://www.pages.drexel.edu/~weg22/gpsMagpie/gpsFixedWing.html
Interesting to read.

23 June 2007, 07:08 | Link | Comments

FPV!

I was lucky enough to use someones wireless camera for two days. So I put it in my easy glider an HOP in the air with it! The transmitter was a 10mW one, and together with the CCD camera the imagery was brilliant!

The camera and transmitter I used is this one
The original quality was obviously a lot better than this compressed youtube one

20 April 2007, 14:59 | Link | Comments

Choices...

What’s the most dangerous thing about looking for information for a project you’re working on? Finding new, exciting things you want to make and try out!!!
Look at this: Nice camera and transmitter in your plane, attach the camera to a servo (or 2), and add a gyro to your ├╝ber-cool VR glasses to drive the servo. It’s like you’re inside the plane!!

Dear Santa, ... ;-)

17 November 2006, 16:18 | Link |

Older articles | Newer articles