Microsat Motion and Passive Magnetic Stabilization

MICROSAT MOTION, STABILIZATION, AND TELEMETRY

(and why we care about it)

By Jim White, WD0E

[Note: This is an updated version of this paper which was originally published in the AMSAT-NA Journal in September 1990 and again in the AMSAT-NA International Space Symposium Proceedings in November 1990. A few typographic errors have been corrected, along with incorrect labels on a few charts.

jw 9-26-91]

Introduction

This paper discusses the stabilization method used with the four Microsats launched in January, 1990, and telemetry related to Microsat motion. Topics covered include stabilization theory, construction related to stabilization, telemetry useful for determining motion, results of work done so far, and topics needing further study. It is written both to document what is known to date and in the hope that others will become interested in this fascinating area of Microsat investigation and assist in solving some of the mysteries presented.

Construction information related to stabilization

In order to understand the stabilization method and the motion of the Microsats, we first have to know a bit about their construction.

The 'top' of the Microsats is the +Z surface (see figure 1). This is the one with the 2 Meter antenna projecting from the center, and is normally 'up' in photographs. It is also the surface that is up when the spacecraft are mounted on the launcher. On the opposite side of the spacecraft is the -Z surface, the one by which the spacecraft were mounted to the launcher. A line drawn up and down through the spacecraft, passing through the center of each Z surface, is called the Z axis. The sides of the spacecraft are named the +X, -X, +Y and -Y surfaces. When looking down on the top surface (+Z), and arbitrarily assuming the +X is at the top of your view, the +Y is on the left side, the -X at the bottom, and the -Y at the right. Lines drawn through the spacecraft and the centers of these surfaces are called the X axis and the Y axis.

There are four magnets mounted in the Z plane (up and down), one at each corner of the spacecraft. In PACSAT, WEBER, and DOVE, they are arranged so that their -Z end is attracted to the earth's north pole, in LUSAT they are opposite polarity.

The downlink antenna blades are painted white on one side and black on the other. These are canted turnstile antennas and are mounted around the edges of the -Z surface, centered on each side. On PACSAT, DOVE and WEBER if you are looking at the -X side, you see the white side of the blade on the left (below the +Y side) and the black side of the blade on the right (below the -Y side). The blades below the X surfaces are oriented similarly. When looking down on the +Z, this means that the white sides all point in the 'counter clockwise' direction. On LUSAT, the blade sides are oriented in the opposite direction (see the Bibliography for photo reference).

There are seven lossy iron rods (hysteresis rods) mounted in the bottom of the battery module, aligned in the X plane.

Expected stabilization mode

The earth's magnetic poles are not located at the north and south poles about which the earth spins or from which our latitude and longitude measurements are made. The north magnetic pole is actually located near the northwest coast of Greenland at about 78.8 degrees north and 70.9 degrees west. The south magnetic pole is near the coast of Antarctica at about 150 degrees east.

The lines of force of the earth's magnetic field are nearly perpendicular to the surface of the earth, except between about plus and minus 20 degrees of the magnetic equator (see figure 2). The equation for determining the slant of the field lines to a line perpendicular to the earth's surface is:

b = 90-atan(2*tan(g)) (Equation 1)

where

b = the angle of the magnetic force line to a line drawn from the center of the earth through the surface at latitude g.

Note that g is the geomagnetic latitude, not (necessarily) the geographic latitude.

This is an important angle because it indicates the Z axis tilt from perpendicular. This means, for example, that at 60 degrees geomagnetic latitude, the downlink antennas will be off pointed by about 16 degrees. Since the magnetic poles are offset from the poles of rotation, the magnetic equator is also offset from the geographic equator. At the longitude of the Americas, about 100 degrees west, the magnetic equator is about 15 degrees below the geographic equator. Hence it does not cross South America in Colombia, but about 15 degrees further south, in Brazil. Thus, the offset from perpendicular at the latitude of the northern border of the US (about 70 degrees geomagnetic latitude) is about 7 degrees, and at Florida (about 60 degrees geomagnetic latitude) it is about 16 degrees. Directly over the magnetic equator, the lines are parallel to the earth's surface. The following table lists these angles for several latitudes.

Offset From Naidr Pointing

(Table 1)

Geomagnetic

latitude (g)

Offset (b)

90 (pole)

0

75

7.6

60

16.1

45

26.6

20

53.9

0 (equator)

90

The Microsats are in a polar orbit that is sun synchronous. To each spacecraft it is always 10:30AM, as it stays in the same place relative to the direction of the sun, and the earth rotates under it. To achieve this, the orbit is tilted slightly, so it doesn't pass directly over the geographic poles, but misses them by about 10 degrees or 500 miles. Note that the orbits miss the magnetic poles by even further on some orbits, and cross directly over them on others (see figure 3).

The magnets in the Z plane of the spacecraft line themselves up with the lines of the earth's magnetic field in a mode generally called 'passive magnetic attitude stabilization' (with the addition of the 'solar propeller' described below, the phrase 'with photon assisted spin' is added to this nomenclature). This holds the spacecraft with the Z axis nearly perpendicular to the earth's surface, except when it is near the magnetic equator. Either the +Z or the -Z will be 'up' or away from the earth, and the tilt of the Z axis to the north or south will match the angle of the earth's magnetic field lines at that latitude. During an orbit as the spacecraft nears the magnetic equator, the top surface will be pulled toward the direction of motion until the spacecraft is flying with the Z axis parallel to the earth's surface. After passing over the equator the leading surface will be pulled down until it is pointed at the earth, and the spacecraft has completed a 180 degree roll. The opposite Z surface is now down and the Z axis is again about perpendicular to the surface, varying by the tilt of the force lines as described in Equation 1 and Table 1.

While the Z axis stays locked into the earth's field lines, the spacecraft spins about the Z axis because of the pressure from solar radiation on the downlink antenna blades. Solar pressure is greater on the white sides of the blades than the black, so the spacecraft spins in a counter clockwise direction when viewed from the +Z side (except for LUSAT, which is opposite). This spin mode is officially known as 'photon assisted spin' but is commonly called the 'solar propeller'. The spin rate was expected to be about 3 minutes per revolution for PACSAT and LUSAT, about 3 times that for DOVE, and slightly slower than that for WEBER. DOVE should be faster because it's painted antennas are for 2 meters and are about 3 times the length of the 70 cm antennas on the others. WEBER should be slower because of it's larger moment of inertia in the Z axis. The hysteresis rods damp forces acting on the spacecraft and prevent excessive motion. It was expected that the speed of the spin would increase until a stable rate was achieved that was a balance between the torque generated by the solar propeller and other forces acting on the spacecraft. The increase in rate would at first be quite fast and would decrease exponentially until that stable state was reached. During sun-lit parts of the orbit it was expected that the spin rate would increase slightly, then would decrease when the spacecraft is in eclipse. Some seasonal variation in spin rate should also occur as the percentage of the orbit that is in sunlight changes. An additional reason for the spin is to equalize temperatures throughout the spacecraft.

All of these factors must be carefully balanced. Too much spin can introduce enough gyroscopic effect to prevent the spacecraft from rolling over at the equator. Not enough spin and the spacecraft may tumble. If there is not enough strength in the magnets the spacecraft will not lock on the field lines and will also tumble. Many factors can upset this balance and cause very complex motions, all of which are undesirable. This stabilization method is strictly passive, no adjustments can be made once in orbit.

The intent of keeping the 70 cm (2 meters for DOVE) antennas down when over favored areas is to maximize efficiency on the downlink, which is expected to be the weaker link.

Telemetry measurements in the Microsats

Except for the computer, each module in a Microsat has one or more sensors (voltage, current, temperature, etc.) that is part of the telemetry system (see figure 4). Many of the signals from these sensors are wired through a resistor voltage divider in the module before being connected to the input pin of an analog multiplexor (mux) chip located in each module. The onboard computer selects which mux chip and which of the signals connected to that chip is switched through the mux to the single analog sampling lead that is part of the spacecraft wiring bus. That lead is connected to the analog to digital converter chip (ADC) in the computer module. The digital output of the ADC is made available to the computer and can be sent in real time as a telemetry (TLM) channel, stored for later download, etc., all under onboard software control.

The voltage dividers in each module are necessary because the input range of the ADC is 0 to 2.5 volts, so all signals must be scaled down to that range. The published TLM decode equations do two things. First, they undo the scaling of the voltage dividers and second, convert the directly measured voltage to it's appropriate engineering value (amps, degrees centigrade, etc.). Some channels have no units, so we refer to their values in 'counts'. The infrared (IR) sensors are a good example.

For some TLM channels a single bit change results in a larger change in the calculated (or engineering) value than you might expect. This is because the ADC is 8 bit and can only separate input voltages into 255 parts, and because of constraints in the voltage and current capabilities of the sensors. For example, a one bit change in raw value of temperature TLM channels from DOVE results in a .6 degree change in the calculated temperature.

The TLM decoding equations were derived from calibration measurement made in the AMSAT Boulder Lab and in Kourou. Those results were plotted and then software was used to fit a curve to the results and derive the TLM decode equations. In some cases few data points were available to derive the equations. In other cases a quadratic equation will not fit the calibration data. Therefore there are some built in errors. Adjustments may be made for those at some point in the future through either changes in the existing equations or completely new decoding methods such as table look up. In the meantime, what we have is reasonably accurate and quite useful.

Telemetry useful in motion studies.

The primary source of motion study data to date has been the solar array currents. An additional valuable input on DOVE, PACSAT, and LUSAT is the IR sensor which looks out the +Z surface. On WEBER, there are two IR sensors mounted on the +Y surface, the same surface the camera looks out of. There are also two magnetometers in WEBER, measuring magnetic forces in the XY and YZ planes.

The +Z IR sensors are designed to show when the +Z surface is pointing at the earth. Their scale is in units and is logarithmic: The larger the count, the more light is striking the sensor. This data is somewhat noisy. It sometimes varies from the actual value by a count or two due to slight bus voltage variations, AC on the source voltage, RFI, or other reasons yet to be determined.

WEBER has no +Z IR (see figure 5). It's +Y IRs look slightly to the left and right of the direction the camera points (perpendicular to the +Y surface). They form an angle between them of 22 degrees and their conical field of view is 10 degrees. They are positioned so that the only time they are illuminated simultaneously is when the +Y is facing the earth. The intent of this design was to allow their use to assure pictures were taken only when the camera was pointed at the earth. Their readout is opposite the other spacecraft's IRs: Illumination decreases their count. This data is also somewhat noisy.

A temperature sensor was mounted on the inside of the +Y solar array on each spacecraft. One of the reasons for this positioning was to provide an additional point that could be used for attitude and motion determination. However, it has turned out that the solar panel is a good enough insulator that the temperatures as measured by these sensors only reflect the general warming and cooling of the spacecraft during the sunlit and shadowed portions of the orbit, and do not change noticeably as the spacecraft rotates.

The temperature sensor mounted on the top module cover (+Z side) can be used to roughly determine when the +Z is in the sun. Temperatures here will lag sun illumination by a few minutes because it takes a while for heat to be transmitted through this surface when the spacecraft comes out of eclipse.

The WEBER magnetometers were designed to allow measurements of the earth's magnetic field. Magnetometer 1 is oriented to sense flux lines in the YZ plane of the spacecraft, and magnetometer 2 measures flux lines in the XY plane. Each of these sensors was biased with a small permanent magnet mounted near it to cancel the effect of the stronger spacecraft attitude control magnets. At present we do not know if these sensors are functioning properly. Work to validate their function and correlate their readings to a known map of the earth's magnetic field and to spacecraft motion remains to be done.

By far the most useful telemetry channels have been the solar array currents. Since the current from the solar arrays changes essentially instantaneously as a function of the angle of sunlight on the panel, that angle can be calculated with reasonable accuracy for every surface that is in the sun every time a TLM sample is taken. This allows us to determine the angle of the sun on either two or three surfaces every x seconds, where x is the TLM sample interval. Since we know the angle of the sun to the plane of the orbit is 22.5 degrees all the time, that amount can be factored out and the angles of the surfaces to the orbit plane can be determined. The equation for calculating the sun angle from the array current is:

a = acos(Inow/Imax) (Equation 2)

where

a is the angle

Inow is the current measured

Imax is the maximum current the panel is capable of producing.

Imax will be 351 ma plus or minus 2 ma for every panel on every spacecraft except the -Z panels, which are exactly half, since they contain half the number of cells. The result of this calculation using many computer spread sheets will be in radians and must be converted to degrees by multiplying the result by 180/Pi. Using Equation 2, when there is no sun on the panel the calculated angle will be 90 degrees, and when the sun is directly on the panel (perpendicular to the panel surface) the angle will be 0 degrees. Since that is not the way we usually think of angles to surfaces, we can take the complement of the angle (subtract it from 90) and the result will be the angle between the direction of the sun and the panel surface. Thus the more complete equation is:

a = 90-(acos(Inow/Imax)*180/Pi) (Equation 3)

See Figure 6 for a chart of this relationship

This formula is not accurate at shallow sun angles because of scattering from the solar cell cover slides and array shadowing as explained below. That is why the result of the telemetry decoding equation is negative when there is no sun on the panel. Work needs to be done to further refine this equation, but it appears to be quite accurate above about 15 degrees of sun angle.

The effect of the shadow of the antennas falling on the solar panels is noticeable. The antennas are made of 3/8" Stanley tape measure material, therefore they will cast a 'thin' shadow and a 'thick' shadow depending on their orientation to the sun. Their long side is parallel to the X axis. The uplink antennas will shadow the +Z array all the time the +Z is in sun, since they project upward perpendicular to the +Z surface from the center of it. The downlink turnstile antennas will cast a shadow on the -Z surface at times, and on the side they are mounted below at other times. The effect on array current outputs when shadows are cast on the various panels, both for a nominal attitude and for situations when the spacecraft is wobbling, remains to be studied.

The spacecraft spin rate can be easily determined by examining array currents. The best way to do this is to plot them with a computer. See Figure 7 for an example of an ideal case. Note that the vertical (Y) axis of the graph is current and the horizontal (X) is time. Points on the X axis are 10 seconds apart. With nominal conditions we expect each of the side panels to rotate into the sun and out again as the spacecraft spins around the Z axis. A plot of the X and Y surface array currents will show the current from one surface start out at zero, slowly climb to a maximum, then drop to zero again, as it rotates into then out of the sun. As an example, let's assume the +X panel is fully in the sun. It's array current will be about 340ma. This is the expected current with a sun angle of 22.5 degrees, the nominal angle of the sun to the orbit plane. No current will be generated by the other X panel because it is in full shadow on the 'back' side of the spacecraft. The Y panels also will not be generating current because they are parallel to the suns rays. As the spacecraft rotates (counter clockwise as viewed from the +Z), the current from the +X will slowly drop and the +Y will start to generate current as it comes into the sun. When the spacecraft has rotated 45 degrees the +X and +Y will be receiving equal sun and be generating equal current. 45 degrees later only the +Y will be in sun and the +X will drop to zero because it has passed into shadow. This pattern continues as the spacecraft rotates. To calculate the spin rate, count the number of seconds between the peeks of two successive panels and multiply by four.

If there is wobble about the Z axis, and the period of the wobble is short compared to the rotation rate, the current plots from each array will be a wave superimposed over the nominal rise and fall described above. They can be pictured as a higher frequency cosine wave imposed on top of the positive half of a lower frequency cosine wave.

Initial motion

The spacecraft were mounted on the launcher with a single explosive bolt extending from the center of the -Z surface through a spring to a ring (the ASAP) around the bottom of the payload section of the Ariane IV launcher (see Bibliography for photo reference). The +Z surface was pointed straight up as the vehicle sat on the pad. Each spring was a slightly different strength. After SPOT 2 and the UoSats were deployed, the Ariane third stage and payload section was pointed backward (top opposite the direction of flight) and the explosive bolts were fired, allowing the springs to pop the Microsats off the ASAP. Over time the spacecraft became separated because of the different spring strengths. The torque generated by the springs, which were not perfectly aligned with the center of gravity of the spacecraft, caused the spacecraft to initially tumble in orbit. After a few days the magnets had locked into the earth's field, and some of the tumbling inertia was translated into spin about the Z axis. Initial spin rates about the Z axis were about 20 minutes per revolution, tumble rates were apparently much higher. The change from tumble into magnetic lock has not been studied in detail and is of great interest to the spacecraft builders.

All of the above motions happened as expected except for the spin direction of PACSAT. PACSAT's spin is discussed in more detail below.

DOVE quickly spun up to a brisk 20 seconds or so per revolution due to its longer downlink antenna blades (see figure 8 for actual DOVE array current plots). The exact current spin rate has not been determined because the real time telemetry sample rate is too slow to assure an accurate measurement. In order to obtain an accurate picture of a wave form, it must be sampled at a rate at least two times its frequency (Nyquist rate). Since the sample rate is 10 seconds and the apparent spin rate was 20 seconds, we believe we were sampling slower than the Nyquist rate, possibly resulting in aliasing, and the result is suspect. However, strip chart recordings of signal strength made by Junior DeCastro, PY2BJO, also indicated an apparent spin rate slightly greater than 20 seconds.

There was some concern that DOVE would spin so fast that the gyroscopic effect of the spin would overcome the magnetic lock and it would not roll over at the equator. This does not seem to be a problem as array current and IR sensor telemetry clearly indicate it's +Z is up in the northern hemisphere and it's -Z is up in the southern.

Wobble about the Z axis has been apparent to one extent or another on all spacecraft since magnetic lock was achieved. This is of interest because it affects RF links, has the potential to tell us a great deal about this stabilization method, and affects what can be done with the WEBER camera and other experiments. This wobble can be described as a line extending from the Z axis describing a cone pattern in space, and is sometimes called (reasonably) 'coning'. Initial measurements indicated PACSAT was wobbling up to 20 degrees either side of the nominal Z axis alignment, or a total of 40 degrees (see figure 9 for an early PACSAT array current plot). The wobble rate was about 56 seconds. That is, a complete circle from say, tilt in the direction of flight, all the way around until tilt was again in the direction of flight, took 56 seconds. Theory holds that the wobble direction will be the same as the spin direction, but this has not been proven with data. The wobble appears to vary a great deal from orbit to orbit. The early array current wave forms were fairly complex and it now appears that we were seeing a combination of the effects of rotation, wobble and antenna shadowing. LUSAT was also wobbling, but somewhat less. Data from WEBER prior to the commencement of picture downloading has not been analyzed, so its wobble amount is unknown. However, it is expected to be larger than the other spacecraft since it has less favorable moments of inertia. DOVE's wobble amount is currently unknown because the TLM sample rate has been to low. The wobble amplitude of PACSAT appears to have reduced since it began to spin up in the correct direction. Wobble amounts of PACSAT and LUSAT now appear to vary from nearly none to about 10 degrees, 20 total (see figure 10 for an example of PACSAT array currents showing little wobble). The pattern of this variation is not apparent, and the cause is unknown. However, it is theorized that the wobble now being seen is caused by the pull of the magnetic poles as the spacecraft passes over the top and bottom of the earth, but at a point far away from those poles. The effect is similar to giving the top of a spinning gyroscope a push to the side with your finger: A wobble is induced that eventually damps itself out. When the spacecraft orbit passes more directly over the magnetic poles, no wobble is induced because the pull is in line with the orbit plane. At present this is speculation and confirmation will require collection and analysis of additional whole orbit data (WOD). It has also been suggested that the variations in wobble are influenced by changes in the earth's magnetic field and that the spacecraft could be used to study those variations. The wobble amount for PACSAT has been confirmed by correlating the IR sensor data with the array current data (see figure 11). Plots of both have shown that the Z axis leans over far enough at times that the IR sensor in the +Z sees the sun. Sun angles calculated from the +Z array current match up well with the angle necessary for the IR to be illuminated by the the sun.

There was initially some confusion about the spin direction of LUSAT. However, once the opposite orientation of both the turnstile blades and the magnets is taken into account, it is clear that LUSAT spun up in the correct direction.

PACSAT spin reversal

Although it had never before been seen in a spacecraft using this stabilization method, there was a 50/50 chance that any one of the Microsats would start out spinning in the 'wrong' direction; that is, the direction opposite that caused by the solar propeller. As it turned out, this is exactly what happened with PACSAT. By early February 1990 it was clear it was spinning clockwise instead of counter clockwise, and was slowing down. By the tenth of February it had slowed down to about one revolution per hour, and on about February 17 it turned around and began to spin in the correct direction (see figure 12 and figure 13). While enough data has been examined to assure this interpretation is correct, much additional study is necessary to determine exactly when the turn around occurred, if there was any tumbling, if magnetic lock continued even though there was no spin, how long there was little or no spin, etc. PACSAT has continued to spin up and has now reached a rate of about 2 1/2 minutes per revolution, which is well within the design parameters (see figure 14).

The importance of understanding wobble

Understanding wobble is important for a number of reasons. Let's first look at RF link performance. For the purposes of this discussion we will ignore the slight tilt of the Z axis at high latitudes caused by the tilt of the earth's magnetic field lines. Assume for a moment the Z axis of PACSAT is stable (no wobble) and the orbit will pass directly over your station, which is in the northern hemisphere well away from the magnetic equator. As the spacecraft comes over the horizon the uplink antenna sticking out the top of the spacecraft will be pointed away from you at an angle of about 117 degrees from a line drawn between you and the satellite (see figure 15a). The downlink turnstile will be pointed at the earth, but away from you by about 63 degrees. We could say the pointing angle of the downlink antenna is 63 degrees. This angle will slowly decrease as the spacecraft approaches until, when it is directly overhead, the pointing angle will be essentially zero. It will then increase until at the opposite horizon it is again about 63 degrees. On the horizon the polarization of the downlink signal will be elliptical and of the proper handedness (right handed for PACSAT), overhead it will be nearly perfectly circular.

If the pass is not directly overhead the initial off pointing will be less, but it will not be zero at time of closest approach (TCA) or highest elevation. For a pass that never gets above 15 degrees elevation, the pointing angle never gets better than about 45 degrees, and the polarization is always elliptical. Of course, for stations near the magnetic equator the situation is quite different since the spacecraft is rolling over 180 degrees at that point.

Now let's assume that the spacecraft has a 40 degree wobble about the Z axis (20 degrees either side of the stable position). Remember that the wobble rate is about 1 minute per rotation. Now when on the horizon, the pointing angle will vary from about 83 degrees to about 43 degrees in 30 seconds (figure 15b). An 83 degree off pointing will cause the polarization to be nearly linear and at that point in the wobble cycle it would be horizontal. When wobble causes the spacecraft to tilt to the left or right as seen by the observer the polarization would be strongly elliptical and canted about 20 degrees from horizontal.

At this point we do not have radiation pattern measurements from the Microsat antennas, except S band. The engineering model may be used to complete that work. However it should be obvious that wobble such as we are seeing will have some effect on RF link performance. This will be especially true when using simple omnidirectional ground stations antennas. Remember also that the wobble is not always there; some orbits show nearly none. Since one of the objectives of the Microsat effort was to demonstrate effective use with simple antennas, and the most efficient simple design is probably going to be influenced by the wobble, understanding it becomes important.

S band performance on DOVE and PACSAT will also be affected by wobble. The S band antenna is a bifilar helix which projects from the +Z surface and is mounted about 1 1/4" in from the edge. That surface is generally away from the earth in the northern hemisphere, so for high elevation passes in that area the body of the spacecraft is often between the antenna and the ground station. However, when wobble is present the ground station will see a complex pattern of changes in polarization and signal strength as the wobble period and the spin period interact. The design of ground station antennas could be changed to most efficiently work with this pattern once it has been characterized.

It would be advantageous to be able to include the pointing angle of the Microsats in tracking programs. This would help those experimenting with antennas or just using the spacecraft; station adjustments could be made to achieve maximum efficiency. It may turn out, for example, that one type of simple antenna works best when the spacecraft is making a low elevation pass and a different type works well for high elevation passes. It may also be true that different antennas work best for different amounts of wobble. It will be necessary to more fully understand the cause of the wobble in order to mathematically model, and to some extent predict, the spacecraft motion in tracking programs.

Since the wobble is generally undesirable and was not expected, it is important to know its cause and characteristics so the designs of future similar spacecraft can be modified appropriately.

Areas needing further study and investigation

A number of areas of study falling into the general category of motion and stabilization are ripe for further investigation. Some will require research into fields such as antenna radiation patterns, or the application of knowledge of physics, orbital mechanics, etc. Some will also require access to sophisticated test equipment or the Microsat engineering model. Many can be undertaken with only a small amount of assistance from the engineering team. However, for most, all needed data can be obtained from Microsat telemetry and published documents. Following is a summary (in no particular order) of those study areas. It is not exhaustive, but is included here in an attempt to show the richness of further investigative opportunities, and in the hope that it will stimulate those interested in amateur satellites, or just in satellites, to participate in this fascinating field.

- Exactly when and why does a Microsat's Z axis wobble? What is the cause of the wobble? Can we understand enough about the cause to predict when it will occur and its amplitude? There is a component of precession in the roll over that occurs at the equator. Exactly how much does the Z axis deviate from the orbit plane during this event?

- What is the nature of WEBER wobble as compared to the others? WEBER's moments of inertia are different. How does that affect its susceptibility to wobble? Can WEBER wobble be correlated to it's magnetometer readings to validate those readings, or the other way around? Can the magnetometer readings be correlated to the array currents?

- What are the effects on the RF links of the wobble? Can predictions be correlated to measured signal strength and polarization changes? What are the radiation patterns of the antennas (70 cm downlink, 2 meter downlink, and 2 meter uplink)? What are the best simple ground station antennas to use when wobble is present and when it is not?

- Exactly how quickly did each Microsat achieve magnetic lock. What was the nature of the motion (tumble) in the first few days? How quickly did PACSAT lock as compared to the others, given that it was spinning in the wrong direction?

- Exactly when did PACSAT turn around, and how much, if any, tumble occurred at that time? After it turned around, what was its spin rate each day until it reached a stable rate? How did the temperatures in PACSAT change as its spin slowed down to zero about February 17th? How hot did the +Y surface get when it faced the sun for long periods?

- What is the DOVE spin rate currently, and how fast did it increase since it came out of tumble? A WOD collection of the array currents is planned using a 2 second sample rate. This should allow accurate determination of the spin rate, and also tell us if DOVE's roll over at the equator is delayed or affected at all by its higher spin rate.

- The conical view angle of the IR sensors in all but WEBER needs to be measured so the IR data can be more precisely correlated with other sensor data. What is a more precise method for relating sun angle to solar array current when the sun angle is low?

- What is the exact current drop from a panel when a thick and a thin 2 meter antenna shadow falls on it? What is the effect of shadowing of the S band antennas on the +Z arrays? Under what circumstances will the shadow of the 70 cm and the 2 m downlink antennas fall on the X or Y panels (with and without various amounts of wobble)? Can this be verified with telemetry?

Contact information

Experimenters interested in pursuing any of these areas should contact the author or Bruce Rahn, WB9ANQ. Bruce is lead command station for AO-16 and the coordinator of the Command Station Development Program (CSDP) for AMSAT-NA. Information about how to obtain microsat telemetry from the TLM data bank can be requested from Reid Bristor, WA4UPD. When requesting information please include an SASE, and for data include a disk, mailer and return postage.

Jim White, WD0E
Colorado Satellite Services
45777 Rampart Rd.
Parker, CO 80138
jim@coloradosatellite.com

Abbreviations:

TLM - telemetry

ADC - analog to digital converter

WOD - whole orbit data. Data collected for one or more entire orbit and downloaded in one package.

TCA - time of closest approach. The time at which the satellite comes closest to the ground station.

ASAP - Auxiliary Secondary Payload Adaptor

Bibliography

A Brief User's Manual for WEBERSAT'S Ancillary Experiments: Chris Williams, WA3PSD, Center for Aerospace Technology, Weber State University, Ogden, Utah, 84408-1805.

WEBERSAT, Stan Sjol, WKP, AMSAT Journal, V12 #3, Nov 89.

Microsat Telemetry Equations, AMSAT Journal V13 #1, March 90, pp24-25.

For a photo illustrating the mounting of the Microsats on the ASAP see AMSAT Journal V13 #1, March 90, pp28

For a photo illustrating the black and white painted antenna blades see AMSAT Journal V12 #1, May 89, pp8.

Acknowledgements

Thanks go to Jan King, W3GEY, for generously sharing his vast knowledge, Chris Williams, WA3PSD, for his patience in explaining the WEBER attic, Bob McGwier, N4HY, and Harold Price, NK6K, for their assistance with telemetry formats and WOD collection, Franklin Antonio, N6NKF for providing the math describing the earth's magnetic field, and Ron Cox, Stan Woods, Richard, VK7ZBK, & Grahm, VK5AGR, for data collection and reporting.

About the author

Jim White, WD0E, has been licensed for 37 years and currently holds an Extra Class. He played a small roll in the construction of the Microsats, assembled the ground stations used in the lab and at the launch, and is an AO-16 and DOVE command station. He co-authored a telemetry decoding program for the Macintosh and has been concentrating on Microsat telemetry, especially related to motion, since launch. He is also active as the sponsor of an Amateur Radio Explorer Scout Troupe, teaches the Extra Class with the Denver Radio Club, works with a high altitude balloon group called Edge of Space Sciences and with the Deep Space Exploration Society, and participates in the SKYWARN program.

Update 1999:  Jim retired from US West after 25 years in computer operations and software development and now owns Colorado Satellite Services, a satellite and ground station consulting company.  Most recently he completed the construction and testing of Falconsat-1 for the US Air Force Academy where he was Project Engineer.