MAVRK Into Near Space

Blog Post: A system

Alex Kozitsky — Thu, 01 Mar 2012 05:06:00 GMT

I have been working on this project again for almost a week. I think that it is coming together quite well.

Here is a photograph of parts that are working and parts that are still in development.

These are almost all the parts that are currently planned to be on the capsule.

There modules starting on the top left and going across:

1.) Garmin GPS (wired to the RF breakout board)

2.) The RF micro debug board that main purpose is to enabled to log the parameters on a microSD card.

3.) The CC1120 radio with a 1 watt amplifier with an attached antenna.

4.) Pressure sensor (wired to the AFE breakout board and sitting below the Spot Personal Tracker)

5.) The next AFE breakout board is used to power the Garmin GPS from the first module

6.) Ericsson Phone (wired to the AFE breakout board)

7.) The big module on the bottom is the PMU-BAT. It allows to power the board from a single cell lithium battery. The battery is shown inside the black box (like the term) on the left. It has a 2200mAh capacity. I am still debating on whether to use the rechargable battery as shown here, or not use the PMU-BAT and go to a non-rechargable batteries. However with that setup it will be much harder to tell the state of charge ( the PMU-BAT has a feature where it will tell how much amp-hours are remaining).

8.) The ContourRoam which is the HD video recorder that will be functioning autonomously.

9.) Spot Personal Tracker, a satellite GPS tracking tool. Which also will be functioning autonomously and report the GPS coordinates to a web page, or mobile phone every 10 minutes.

10.) Not shown here is a photo camera (used to take the picture) that also will be taking a picture every 10 seconds or so autonomously.

The theory of operation:

The MAVRK system will take a

pressure reading,
inside temperature reading (outside temperature reading may also be developed),
a GPS reading that also contains the altitude, coarse heading and speed

This on a period of 1.6 seconds and send it over the radio. The radio will use frequency hopping per FCC regulations for radios operating at 1 Watt power.

The cell phone will only be used when the system detect that is near landing or has landed and will send out a SMS message to certain people. This functionality has already been developed.

The ground stations ( there may be as many as are necessary) will consist of a MAVRK motherboard with a CC1120 radio with an antenna. It will be connected to a computer that will display all the information that the nearcraft is sending back. This will allow a user to track the nearcraft. An optional item for this setup would be a GPS device that to locate where the person is in relation to the nearcraft. I have decided to not use a live video transmission from the nearcraft to make the system simpler and more feasable on the first attempt.

I am currently working on the frequency hopping for the nearcraft. This will be used to transmit the parameters to ground stations. I estimate that the range of the radio should easily be atleast 20 miles and to as much as 50 miles. We will need atleast 20 mile range to be have a shot at constant connection.

The next goal is to have the radio working and with finishing the nearcraft system should be easy. After that will be work on the ground stations and the computer application displaying the status of the nearcraft.

Blog Post: Using Canon Hack Development Kit on my Canon Camera

Alex Kozitsky — Sat, 11 Feb 2012 06:03:00 GMT

CHDK stands for Canon Hack Development Kit and it is a tool that allows you to add much more functionality and control to your Canon camera. The way that it works is that you write their software to your SD card before using it in your camera. When starting the camera with the SD card with the CHDK code, it takes control of the camera and boots their software. CHDK has a lot of interesting and complex features including controlling the camera using a written script. To return the camera to the normal functioning state, simply delete thd CHDK software or use a new SD card with the camera. In our case we want the camera to take a picture every so often. Also it would be good to have the LCD be turned of during that time to save on battery usage.

First I could not find a pre-written script that would do those tasks. There were scripts on their website that claimed to do those tasks but they would not work properly. The scripts would very easily take pictures every so often, but turning the LCD was just not possible with them.

After researching their Wiki, mainly on how to write in Basic I was able to put together a script that would turn of the LCD (actually the LCD backlight). The script is attached to the blog post if it is needed to be used by any one. Just as warning this has only been tested on a Canon SX230 HS. My best guess, considering it is using simple functions is that it will work on most cameras.

The script starts the camera with the LCD on. To toggle the LCD display press repeatedly on the "left" key (the one on the wheel) until the LCD changes state. Currently the timer is setup to take a picture every 20 seconds, but that can be changed easily on the script or on the camera in the script parameters section.

What remains to be seen is how long the camera will last with taking pictures continuously. It would be preferred if the time would be atleast several hours. That may require extending the picture timer to maybe 30 seconds. A test is needed and that may be done another day. Also a test is needed to a make sure the pictures a taken in the correct focus, though that may be done automatically by the camera.

The script is attached.

File: Avionics System Overview

Alex Kozitsky — Mon, 19 Dec 2011 03:31:00 GMT

Here is a current design for the avionics in the MAVRK-INSPACE nearcraft. This is an early design and this may change overtime as more research and development is done.

Blog Post: Avionics System Overview

Alex Kozitsky — Mon, 19 Dec 2011 03:28:00 GMT

I decided to provide a system overview of the avionics. Here is a picture (this picture may also be found in the files section).

First let me mention that this an early design configuration, this may change over time.

First the cameras will not be connected to the motherboard and there will be no way to stream the photography that they provide live because of that. The images will be extracted from the memory sticks after nearcraft retrieval. They will run until the batteries die, the memory space is used up or the nearcraft is retrieved.

As you can see from the diagram, there is a third camera. This one will be connect to the flight computer. Imagery from this camera will be streamed to ground stations using the radio in another slot. Currently it is expected that about 1 picture a minute will be beamed back to earth.

In a future test I will attempt to measure and document the radio range that the CC112X-2442-MVK module provides. I expect the range to be enough for tracking purposes (as long as the tracking vehicle is nearby) and possible enough range for telemetry (position, data...) to fixed ground stations for most of the trip, but it is likely that live images will only be available to fixed ground stations ( by that I mean a station at launch site) for a limited period of time due to the radio range when using high bandwidth transmissions. The low bandwidth transmissions have a longer range and should be available for most if not the entire trip.

Th RF debug board has a SD card connecter that will be used with an SD card to log all the data from sensors and also from the GPS device for the entire trip.

An AFE breakout board will be connected to a GPS receiver and an AT&T Sony Ericsson T226 cell phone which has a serial AT command interface. With this connection we should be able to send SMS messages from the flight computer. I have done this before a while ago just by experimentation. The messages are expected to be sent when the nearcraft has landed. The phone's SMS capability may be important if the landing location cannot be determined from the radio transmissions. But it too may be limited if the nearcraft lands in a non-covered AT&T area. If that case occurs, it will be up to SPOT tracker to save the day.

In future blog posts I will document an item or feature that I am researching or developing and it is likely that each item will have an in-depth review.

Blog Post: Testing a Garmin 18 LVC GPS Module

Alex Kozitsky — Sat, 17 Dec 2011 06:50:00 GMT

Today I received a Garmin 18 LVC GPS module in the mail. I ordered it through Ebay after an unsuccessful attempt at ordering it from Amazon. This product is hard to get because it has gone obsolete. It is replaced by GPS modules that are not too friendly to nearcraft. First they tend to have a USB interface. This one has a RS-232 TTL logic interface. This makes it much easier to interface with a MCU. Second even as important is the newer GPS modules (possibly not all) stop working above 60,000 feet. As we expect the balloon to go up as high as 100k feet this is a problem. This particular module does not have that limit.

After modifying MAVRK software to work with 4800 baud (the default baud rate for this module) and programming the MCU with a UART pass-through program, I checked the output on the computer. It was giberish. I eventually discovered that the RS-232 signal coming from the GPS module was inverted. Since the MSP430F5438A does not support inverted UART signals, I had to create an inverter. I used a 74HC00 NAND gate and set it up as an inverter.

That fixed the problem, I started getting the expected GPS sentances. It has two formats one starting as $GPGGA which I think will be sufficient for our purposes as it has latitude and longitude coordinates as well as elevation. It also has information on the current time. I checked the coordinate with Google Maps and it had pinpointed the apartment that I was in. It did take it about 5 minutes to start outputting correct info. But that is acceptable.

I really wanted for this project to not have to create any specific PCBs, but in this case I think that I will have to create one using the inverter, but it should be really simple.

File: A balloon with a payload

Alex Kozitsky — Fri, 16 Dec 2011 02:41:00 GMT

So this is an example picture of a helium balloon with a typical styrofoam payload. Descriptive labels with return contact information are typically posted on the payload so that if the team loses track of the balloon they have some hope that if someone finds the balloon they will know how to whom to return it.

Forum Post: Questions/Comments

Alex Kozitsky — Fri, 16 Dec 2011 02:30:00 GMT

If anyone has any questions or comments feel free to post them here. I will try to respond to them in a timely manner.

File: Picture taken at approximately 100,000 feet above Oregon by Justin Hamel and Chris Thompson

Alex Kozitsky — Fri, 16 Dec 2011 02:29:00 GMT

Here is an example picture (not by me) that was taken from a high altitude balloon.

Wiki: MAVRK Into Near Space - Wiki

Anonymous — Thu, 15 Dec 2011 07:49:00 GMT

Blog Post: MAVRK High Altitude Balloon Project Description

Alex Kozitsky — Thu, 15 Dec 2011 03:58:00 GMT

I will try to go into a little more detail here about this project than was outlined in the project proposal.

First why did I choose this project.

I think that to put something together that can travel into near space - or really almost out of this world is something that is really exciting. Not many things exist at 100,000 feet high. The pictures that past people have obtained of near space have been breathtaking. You can see the curve of the earth at that altitude.

Also this project will have plenty of electronics tinkering, software development and engineering things that I enjoy to do anyway with my free time.

First in case you are not familiar with high altitude ballooning let me give you a quick summary.

A high altitude balloon consists of a payload that must be below 6 pounds in weight (more is possible if needed with certain restrictions). That payload usually has a flight computer that logs information like temperature, air pressure, and servo control positioning cameras. Most try to have radios but keeping in contact with the nearcraft (near space craft) but that has proved difficult for some projects that have attempted to use them. The nearcraft airframe is usually a styrofoam box with ports made for cameras on the sides. Styrofoam is chose because it is light, acts as an insulator, and can absorb energy on high impact landings.

A weather balloon is filled with helium and until it provides several pounds of lift. Once release the balloon will slowly rise to altitudes such as 80-100k feet. At those heights, the air pressure is 1/100th of the pressure at sea level. Thus the helium balloon expands until it bursts. A parachute is attached below the balloon and opens up once the nearcraft starts accelerating down.

The total flight takes about 2-4 hours from launch to landing. Almost always the balloon is tracked using the GPS telemetry from the balloon, or by the use of beacons. This is due to the fact that the payload is not expendable and has value including data and pictures stored on it.

The balloon can travel around 100 miles over a period of 2-4 hours. The balloon is chased using the GPS telemetry and hopefully eventually recovered for future reuse.

Temperatures in near space is around -40 C, which probably only occurs on the Earth poles.

So I have been working on an outline of this project. There are four main components of the payload. They are the camera, video camera, tracker (I will explain later) and a group of other components that work of the MAVRK motherboard.

Almost of all of them have not been acquired (other than the MAVRK motherboard and modules) and will be in the next month or so.

The camera needs to take picture periodically. It can be set up to be triggered by the flight computer or as we will try to do automatically.

There exists a Canon Hack Development Kit for Canon cameras that allows to program a Canon camera to do many things including taking pictures automatically periodically and storing them to the flash disk. The way that it works is that you store a piece of code and your script which has the control sequence for the camera on the flash card and insert into the camera. Upon power up the program gets loaded into the camera memory and it executes until you tell it stop (by pressing a button) or the camera runs out of power.

In our case all we need to write in the script is too take a picture every 10-20 seconds, and keep the LCD off to conserve power.

I have not decided which Canon camera to purchase, but it probably will be of of Ebay and have pretty good resolution.

The video camera that I have decided to use is the ContourHD which is HD video camera by Contour. It has simple operation, again does not have to be controlled by the flight computer and has battery power for about 4 hours recording time, which is enough for high altitude balloons.

The flight computer will be the MAVRK motherboard. It will have RF Debug module which will have a two-fold purpose. First a RF Debug module has temperature measurement and that will be used for internal temperature reading. Second the module has a micro SD connector that will be used for mission data logging. This log will contain data from all the sensors including the GPS location every second or so that can be analyzed after craft retrieval.

It will also have a module for exteranal temperature measurement. Another camera that is somewhat low resolution that will be directly hooked up to the motherboard with a UART connection. The data from this simple camera will be streamed via radio back to ground stations. I expect a picture every minute or so arriving to ground stations.

The radio will probably be a CC1120 (a MAVRK module) with a power amplifier and mostly operating in transmit mode, although receive for command purposes may be implemented.

Connected to the motherboard will be a GPS receiver which the flight computer will periodically send its info to ground stations for tracking.

And finally I also plan to use a cellular phone attached to the motherboard to message the location of the craft when it lands if the radios have been compromised in some way.

There is one other component that I will use. This is for the sake of redundancy and near-fail-safe purposes. It is the Spot Personal Tracker. It essentially is a satellite tracker that periodically sends a SMS message to a mobile device of its GPS location.

The reason that that is necessary is if something happens to the flight computer, software problem, the batteries for it die prematurely, that will mean that the radio will stop transmitting GPS information, the cellular phone will not make the landing location known and unless it lands in a populated area and it is discovered it may be lost. That would be bad. So this tracker is completely autonomous, it does not need any external control and it keeps transmitting every 10 minutes its location for 24 hours. That should be enough.

I have mentioned the ground stations and here I will provide more detail. I envision one ground station that will stay in place that will use a MAVRK board with a PC connection to view the pictures coming back from the craft. This can be used for public monitoring. The mobile stations can be placed in vehicles, also will use the MAVRK motherboard and will also receive telemetry from the craft, the most important is the GPS location for chase purposes.

I will go into more detail on each component as I develop them in future days.

Blog: MAVRK Into Near Space - Blog

Matt Lauer — Thu, 15 Dec 2011 01:49:00 GMT

Files: MAVRK Into Near Space - Media Gallery

Matt Lauer — Thu, 15 Dec 2011 01:49:00 GMT

Group: MAVRK Into Near Space

Anonymous — Thu, 15 Dec 2011 01:49:00 GMT

Let's build a weather balloon payload to take aerial photography and collect other data, and then beam this data back to Earth!

Forum: MAVRK Into Near Space - Forum

Anonymous — Thu, 15 Dec 2011 01:49:00 GMT

Wiki Page: MAVRK Into Near Space - Wiki

Matt Lauer — Thu, 15 Dec 2011 01:49:00 GMT