From SEDSWiki

The map of the earth's magnetic field refered to in this thesis can be found here: http://www.ngdc.noaa.gov/IAGA/vmod/igrf.html the idea is to compare magnetic field measurements taken by the satellite to this magnetic field map to help identify the satellites orientation.

ION CubeSAT, University of Illinois, Urbana-Champaign

was destroyed in 2006 due to launch vehicle failure http://cubesat.ece.uiuc.edu/index.html

this one used:

-a magnetometer and sun sensors (no specifics found on the number) for detection

-magnetic torque coils for pointing the spacecraft

-4 micro arc-thrusters for rotation in x-, y- and translation in z-axis (something I find very interesting!!!)

other interesting facts:

-they wanted to process all their attitude controll data on ground due to limited capacity on s/c

Note 0 (I think I wrote everything I found on the site)

AAU CubeSat, University of Aalborg

http://www.cubesat.auc.dk/

launched in 2003

A short abstract of their adcs:

"To determine the satellites attitude two types of sensors are used. A magnetometer, build up with components from HONEYWELL, to provide information on the direction of the magnetic field of the Earth, and sun sensors. These sun sensors are basically planar photo diodes placed on each side of the satellite to measure the intensity of the incoming sunlight. The solar sensors and coils are shown on the figures below.

To control the satellites attitude in orbit three coils are used, which are mounted on three of the satellites sides perpendicular on each other. These will generate magnetic fields, which interact with the Earth's magnetic field, and hereby change the attitude of the satellite." (from: AAU Cube Sat: http://www.cubesat.auc.dk/)

there's a master theses about the project, but the acds section is not that detailed:

http://www.cubesat.auc.dk/dokumenter/acs_report.pdf

detailed documentation on attitude determination is available here: http://www.cubesat.auc.dk/dokumenter/ADC-report.pdf

documentation on attitude control: http://www.cubesat.auc.dk/dokumenter/acs_report.pdf

really informative is the appendix of the control report, giving an overview of many problems we will have to deal with in the future, like the appropriate coordinate system...

overall, the information is very detailed and should guarantee everything we need so far…

if you are interested in documentation of other parts of the AAU Cubesat: http://www.cubesat.auc.dk/doc.html

AAUSAT II, University of Aalborg

http://aausatii.aau.dk/homepage/index.php?language=en&page=home

will be launched on June 30 (??)

a. determination:

-3-axis magnetometer from Honeywell

-6 rate-gyro chips from Analog Devices

-6 sun sensor using photo diodes

control:

-3 magnetorquer flat coils

-3 brass momentum wheels

Note 0

DTUSat-1, Danmarks Tekniske Universitet

http://dtusat1.dtusat.dtu.dk/group.php?c_gid=1

This Cubsat was put in orbit in 2003 (same launch as AAU), but never heard of it again whithout knowing what went wrong....

attitude determination:

-5 sun angle sensors (one on each side except payload side) and a magnetometer

control:

-magnetorquers

Note: 0

Cute-1.7, Tokyo Institute of Technology

http://lss.mes.titech.ac.jp/ssp/cute1.7/subsystem_adcs_e.html

this was not a cubesat!!!! dimensions: 20 x 15 x 10 cm!!

determination by using:

-gyro sensor, magnetic sensor and sun sensor

and controlled by:

-magnetic torquers

Note: 1 (I think I included everyting, but on the page given abouve, we can find some specifics on the components used)

SEDSat-1

one interesting fact which I didn't mention in my last email is, that our predecessors were using a different approach for attitude detection than the other cubesats I found (if I got the information right). Instead of a combination of magnetometers and sunsensors, they were using a panoramic annular lens camera, pointiong to nadir.

For attitude control they were using again magnetic torquers, which could also be used to roughly determine the erath's magnetic field due to the induction that took place. More details can be found in their detailed technical report

http://www.seds.org/sedsat/papers/techvol.doc

Therefore: Note 2

Explanations:

-Notes: For every mission there will be a rating of 0, 1, or 2, for whether the site contains enough information to be worth checking out some more. 0 means that all the information available is included in the summary, and 2 means that it would take a small army to digest all the information available.

--Dadai 10:09, 25 March 2007 (MDT)

Baseline Requirements

Potential Components

Magnetic Components

Solid gold: http://www.cubesat.auc.dk/ "Attitude Control Report (pdf)" -This link contains extensive information on all aspects of attitude control,

including how to build magnetic torkers, and with to build them with, and how much they will weigh, volume, power consumption!

first of all: magnetic control and attitude determination makes sense where the earth's magnetic field is strong --> LEO would be optimal

Determination

Geomagnetic Reference Field

IAGA V-MOD Geomagnetic Reference Field, already put into a fortran code, to use with known coordinates for deteriming the orientation.

fortran code does not seem to be allowed for uploading, so just the link here: http://www.ngdc.noaa.gov/IAGA/vmod/igrf10.f

Magnetometers

Honeywell

The price for a 3 axis magnetometer from Honeywell (HMC1053: data sheet: http://www.magneticsensors.com/datasheets/HMC105X.pdf) is in Austria (Germany) 70,52€ (about 96US$) + 6€ for shipping, but prices may vary depending on the country. This information is actually not from Honeywell but the local distributor Sensorwell GesmbH… The information I got also pointed out, that to actually get this magnetometer here in Austria, you have to buy at least 3 of them! I am currently trying to get more usable information on that magnetometer from someone living in the USA...

a much cheaper option would be 2 axis magnetometers: Honeywell HMC1052 is available for $8.95 at SparkFun (http://www.sparkfun.com/commerce/product_info.php?products_id=708)

MicroMag

3-Axis Magnetometer

- 500uA @ 3.3V DC

- Field measurement range +/-1100uT

- Resolution as low as 0.015uT

Price at sparkfun: $59.95 http://www.sparkfun.com/commerce/product_info.php?products_id=244

for real details: http://www.sparkfun.com/datasheets/Sensors/MicroMag3%20Data%20Sheet.pdf

seems to me as a real alternative to honeywell

Billingsley Aerospace - not a candidate (size and price)

They describe themselfs as the "world leaders in fluxgate magnetometers" and after looking at their page, it doesn´t seem to be that wrong.

they have a bunch of magnetometers, with a comparison table at: http://www.magnetometer.com/products/mags/magCmp.html

all of this would look really good, but looking at the sizes, a big problem appears: the smallest one, TFM100G2, is already over 8cm and would bust our mass limit by it's weight of 100grams. so i am afraid that´s it for billingsley magnetometers

I just got the price for TFM100S: EUR 7.950,00, so I guess Billingsley is out ;-)

Bartington

even I am afraid theses magnetometers are rather useless because of there rather huge size, I'll put them here. they seem to be used in geophysics only most of the time: http://www.ascscientific.com/mag03.html

main idea:

-usually 3 orthogonal magnetometers (1 for each body axis of the satellite) usually fluxgate-magnetometers (I hope everyone is familiar with the concept... in a few words: the coil is harmonically driven into saturation, while a second coil is used as a pick-up coil, receiving the date from external fields...) because of their accuracy...

-the output current from each pick-up coil is proportional to the external field of the pointing direction...

-you use a model of the earth's magnetic field (usually: the IGRF's map) and compare...

PRO:

-small and relatively cheap (I guess)

-3 magnetometers give you a lot of information, gives you vector information

-Performance is 0.5deg-5deg (according to the PhD Thesis of Thomas, Bak: Aalborg University Denmark), which should be enough for our purpose

-no moving parts

-small power consumption (<1 Watt)

-determination all the time, can not loose contact to reference field

CONTRA:

-the earth's magnetic field is complex and small (30-60nT on the surface) which means: hard to detect

-Disturbance fields due to spacecraft electronics, etc..

-To determine attitude, you need a model of the earth's magnetic field

-Some computing capability to determine the attitude, combining the data of the magnetometers and the model

-Model errors

-External disturbances, like ionospheric currents

-(I have not found that in the books yet, but I guess it should be a problem:) if we are in a polar orbit, there could possibly be some areas where magnetic determination is blind (I am speaking of the regions near the endings of the dipol)??

over all: magnetic determination is widely used on satellites, but not that easily to do. we will need some knowledge in electrodynamics ;-)

also we need to know, if there will be enough computing capabilities onboard (I also read about picosatellites, that did the computing and modelling at the ground, which means: you need 24h contact to the satellite)

also, the initial startup, I am thinking of the calibration, can get tricky I think...

I guess, trying to determine and control the attitude can be a not really easy way to go... with a lot still to learn and an interesting sidestep on data processing...the magnetic instruments would guarantee to maintain the satellite stabilized all the time, regardless of the actual pointing direction (there's no pointing in the wrong direction like it could happen with star trackers, sunsensors...)

--Dadai 07:22, 7 April 2007 (MDT)

Control

Magnetic Coils

main idea:

-similar composition: 3 coils on the three body axis of the satellite

-the magnetic fields induced in the coils interacts with the earth's field -> torque is produced

PRO:

-cheap (I think)

-there are no moving parts...

-if wheels are used additionally: the magnetic torquers are often used to de-saturate reaction wheels...

CONTRA:

-slow

-again there can be some interaction with magnetic fields produced by the satellites electronics

-(again just an idea from me:) can there be control problems near the polar...

One big probem that might occur, if magnetic adcs is used only: the interaction between torquers and magnetometers. Therefore, often a combination with reaction wheels is used, that means: the torquers only work to de-saturate the wheels, which do the actual control...

SWOT Analysis

Strenghts	Weaknesses
> Often used by other CubeSats (well tested) > cheap > no moving parts	> interaction with magnetometers > not really innovative
Opportunities	Threats
> .	> .

Momentum Wheels

http://www.utias-sfl.net/docs/canx2-iac-2004.pdf
http://www.vfct.com/satellites/wheels/wheels2.html

Gravity Gradient Boom

A potentially helpful site: http://sei.tamu.edu/cubesat/cubesat.htm

Sun and horison sensors

Dalsa, manufaturers of mineature sensors: http://www.dalsa.com/pi/products/products.asp

Micro propulsion systems

Overview

-history of rocket propulsion, electric and otherwise: http://www.fathom.com/course/21701743/session1.html

Image:Lecture-17.pdf

Image:ASPS script.pdf

µPPT

see http://en.wikipedia.org/wiki/Pulsed_plasma_thruster for general information

Design Goals for µPPT by ARC Seibersdorf (we would get fly-ready prototypes for SEDSat):

- Impulse bit: 10-20µNs, which would guarantee a pointing accuracy of <1 degree
- Spec. impulse: 500-1000 s
- Power: 0,5-1,5 W
- Weight: < 30g

possible configuration using just 6 thruster for 3 axis stabilization (Stanford University):

SWOT Analysis

Strenghts	Weaknesses
> Innovative (I think the first use with CubSat) > Very Accurate (<1°) > Electr. Prop: Can almost never run out of propellant > non-toxic propellant (Teflon) > professional manufactured devices	> High electr. Power Consumption (maybe up to 1,5 W per Thruster) > relatively heavy (<30g) -> only 6 thruster configuration to use
Opportunities	Threats
> Could also be used for Orbit Changes (in limits), with additional thruster(s) > PR for using an innovative new technology for CubeSat	> Not available before Fall 2008 > New Technology (first ones to test)

Specific pages

http://www.busek.com/micro.html
http://www.mdatechnology.net/techsearch.asp?articleid=519
http://www.st.northropgrumman.com/capabilities/space/propulsion/technologies/micropropulsion.html
http://www.micro-a.net/products.htm

Specific component suppliers

Honeywell (magnetometers):

http://www.magneticsensors.com/prod_syst.html

http://www.tdc.co.uk/magnetic_sensors/magnetic_honeywell1.htm#hmc1053

SPARKFUN (more a online store that sells accelerometers, MEMS gyros and compasses, ...) (thanks to Brynjar from Payload)

http://www.sparkfun.com/commerce/categories.php?cPath=51_29

Key questions of relevance to ADCS

-What it the maximum likely spin rate of the satellite upon release?

-How important is it to attain and maintain an exactingly precise orbit?

-ie, can we implement an attitude control system that is more efficient,
 even if using it results in an (incremental) change in orbit?
 For example, we will have the choice of installing micro-thrusters either
 on one side or on opposite sides of the cube.  Using only one thruster can 
 counter the satellite's rotation, but will change the orbit slightly, where
 simultaniously using thrusters on opposite sides of the satellite, rotation can
 be countered without changing the orbit.

-Do we need three-axis stabilization or can we allow a spin around one axis? (eg.: for taking pictures, it won't matter if the satellite spins around the nadir axis, but if we need a certain side pointed to the sun because of a limited number of solar panels we would need a three axis stabilized satellite)

Proposed ADC Sub-systems

Ongoing discussion of possible ADCS solutions

Mike:

What I think we should do next is focus on attitude control methods. Establish "once and for all" what our options are. there are a million and one ways we can do attitude determination, and I think we understand them much better, and have a decent idea of how we will be going about it.

For attitude control, I see there being three options. Gravity gradient boom, magnetic torkers, and thrusters. I have severe reservations about thrusters, namely that I don<t think we will find any small enough, reliable enough, etc. Gravity booms I also have doubts about for mass and volume and complexity reasons. However, we should look further into each to be able to really justify to the others why we have dismissed them. Magnetic torkers seem to be the industry standard. Stepan is supposed to be looking into them. Would you like to look into one of the other two?

David:

in short: i don't know why, but my thruster guy somehow seems to refuse giving me information about his thrusters, as he doesn't seem to be very happy with my proposal about ginving us the thrusters for free :). i also have serious doubts about thrusters in general, also thinking about the large numbers of thrusters we would need (in comparison to just 3 coils)

the grav grad boom would be a nice idea since it'S kind of simple physics, but thinking of our camera as a primary payload: am i wrong or would the boom point in the same direction as the camera, which would make it hard not to have the boom itself on every photo... right?

so for me, i am starting to really like the magnetic idea, just one thing that came to my mind: the deal is: in the momant we start interacting with the coils, our magnetometers are blind... which could make it a little confusing if we want do supervise the pointing, but that'S something we should be thinking later on, i guess...

Mike:

For the thrusters, we would really appreciate the information on them. For the moment, it is not strictly important that we get them for free. Find out how much they would cost. We might get a sponsor to pay for them, who knows. However, knowing the stats on them will help us hugely in evaluating other thruster systems. We know nothing about thrusters. At least I don't. Physicists don't measure energy levels. They measure differences in energy levels. We will learn a great deal about the thruster systems out there if we have this "ideal" thruster system to compare them to. Also, we have no idea how many thrusters we will need. We can't make decisions without knowing the mass, volume, etc. of given components. The dalsa sensors are less than a cc. So would be the thrusters I would be developing if I followed through on my originally proposed masters project.

Secondly, the gravity boom. Actually, when you think about it, the boom could point either toward the surface, or away from it. It would work in either case. Another thing about the gravity boom is that it is passive ac. Cool to be able to say that we built and successfully deployed it, but it is not active attitude control unless we complement it with another system. Active attitude control is what we are trying to learn how to do.

As for magnetics, as soon as we begin passing current through the coils, we in principle know what kind of magnetic field they will be generating, and can take it into account when trying to find the earth's contribution to the local magnetic field.

David:

Oh i didn't know that the boom works in both directions. sure it is passive, but in some way that's the thing that makes it appealing to me: simple and reliable. the thing we should know: is there the need to point our sat at anything other than towards the earth? seems to me that we will be looking at the earth all the time, since there won't be much else up there to take pictures from...

You're right of course about knowing about the magnetic field we generate, but thinking of how small the earth's field is compared to the one we will produce, i am afraid that it won't be easy to distinguish from the noise we produce... anyway, that's not something we should be worried right now.

Concerning the number of thrusters we would need, i guess for 3 axis stabilization and in order to avoid tumbling, we would need 6 thrusters (!), two for each direction. (2, fully equipped corners). I guess, even really small PPTs would make it hard to meet our mass limit, and we get great problems if some of them don't work (if one doesn't work, for example, the other thruster for this particular axis, would then not just produce a torque, but also some forces, causing tumbling and moving the sat)

Mike:

It's true that a gravity boom is the simplest way of maintaining the correct attitude once we get it. As for the nois of our magnetic torkers, this is something we can test (I understand your concerns and don't know the answer). However, I would just assume that when we actually activate the torkers we would have already achieved a decent tumbling model of the sat and that we would be applying a correction to it, to be measured again after...

As for the number of thrusters, we don't know yet that the changes in our orbit that would result would be critical. That would actually be a very interesting analysis to cary out. We should do that! (it wouldn't be hard...) Also, by your reasoning, we would actually need 16 thrusters: 3 axes, x 2 per axes, x 2 directions per axes. We need to be able to push in both directions! The other option would be 4 thrusters, each on a pivot with a motor.

David:

you mean 12 thrusters? the more I am thinking of how to actually do the determination, the more I like the idea of GPS. but would be be able to implement such a chip? I know that my knowledge so far wouldn't be sufficient...

David:

Sorry for beeing quiete for a long time, but I was really busy with moving to Chicago, and although I know it's last minute as we have been mostly focusing on magnetic torquers lately, I've got some news: The prof I told you about with the thrusters, offered me to write a diploma theses in his research company beginning in December about desining a microPPT thrusters which should be used especially for CubeSats. This project would run until summer 2008, so right in time. This would mean, we would have the possibility to try a really innovative attitude control system for these little sattelites, that no one ever used before, and which would be especially designed for our kind of sat

David:

correction: att. control can be done by 6 thrusters if arranged as in figure (sect. microprop).

Prefered Components and Sub-systems

Milestones

April 29th, 2007

The next task will be to choose a payload
(prelim date: 29 April, 5.5 weeks away). Your team will help to support this by understanding how to control the orientation of the spacecraft, including immediately after deployment when it may be very unsteady (of course, you can help pick the payload as well).

SEDSAT-2 ADCS Research to date

From SEDSWiki

Contents

New info

History (Other missions)

Other missions

Missions listing

Specific missions

Baseline Requirements

Potential Components

Magnetic Components

Determination

Control

SWOT Analysis

Momentum Wheels

Gravity Gradient Boom

Sun and horison sensors

Micro propulsion systems

Overview

µPPT

SWOT Analysis

Specific pages

Specific component suppliers

Key questions of relevance to ADCS

Proposed ADC Sub-systems

Ongoing discussion of possible ADCS solutions

Prefered Components and Sub-systems

Milestones

April 29th, 2007

Views

Personal tools

Navigation

other SEDS sites

Search

Toolbox