LightSail 2 Mission Log

SpaceX Falcon Heavy Launch

June 25, 2019: LightSail 2 Launch

Following a three-hour launch delay, the Space Test Program 2 (STP-2) Falcon Heavy lifted off pad 39A at Kennedy Space Center at 2:30 AM EDT. Following deployment of a set of CubeSats into a lower orbit, the Prox-1 spacecraft deployment was confirmed by SpaceX at 07:49 UTC (3:49 AM EDT). Prox-1 was delivered to a near-circular orbit with the following orbital elements in the Earth-centered Earth-fixed coordinate frame:

UTC time of state estimate: 176:07:49:20.977 (ddd:hh:mm:ss.fff)
Apogee Altitude: 725.017 km
Perigee Altitude: 721.711 km
Inclination: 24.014 deg
Argument of Perigee: 52.035 deg
Longitude of Asc. Node: -185.975 deg
True Anomaly: 309.570 deg

Prox-1 will complete 100 orbits over the next week before deploying LightSail 2 exactly 7 days after the initial Prox-1 bootup upon separation from the launch vehicle. The timer-driven deployment should result in LightSail 2 being ejected from the Prox-1 P-POD at 07:49 UTC on July 2. The LightSail 2 antenna will be deployed five minutes later, and transmission of telemetry beacons will begin 45 minutes after P-POD deployment, at approximately 08:34 UTC (1:34 AM PDT).

The LightSail 2 flight team will converge on Cal Poly San Luis Obispo for initial mission operations. The first possible opportunity for the team to receive the radio signal from LightSail 2 is a tracking pass over the Cal Poly tracking station on orbit 102, currently predicted to extend from 01:26:12 – 01:38:48 PDT. The following three tracking passes during the subsequent four orbits will be over the Kauai Community College tracking station.

Cluster of Satellites

July 2, 2019: LightSail 2 Deployed

LightSail 2 was deployed into orbit by the Georgia Tech Prox-1 spacecraft at 07:35 UTC on July 2, and 45 minutes later the LightSail 2 radio signal was received by flight team via the Cal Poly tracking station.

A lot had to go right in order to receive the signal. Ten satellites were launched by the Falcon Heavy into a similar orbit (see figure below), and it was not clear which vehicle in the cluster was Prox-1. Based upon the expected turn-on time for the LightSail 2 radio signal, and the overflight timing of the tracking station, it would only be possible to receive the signal from the satellites that were near the trailing edge of the cluster. So the operations team programmed the Cal Poly station to track spacecraft “B” in the cluster. The radio signal was received from LightSail 2 at exactly the predicted time. The fact that the ground station antenna pointing was accurate enough to receive the signal indicates that we guessed correctly, and Prox-1 is one of the trailing satellites in the cluster. The 18th Space Control Squadron at Vandenberg Air Force Base provides tracking of space objects. They will likely confirm the appearance of a new object, LightSail 2, later today.

The radio signal was confirmed to be LightSail 2 based upon a morse code transmission that is embedded within the signal. The carrier wave for the radio signal (the carrier wave is the tone that is modulated to convey telemetry data) is programmed to transmit in a series of short and long bursts, representing dots and dashes per the international morse code convention. The LightSail 2 call sign assigned by the United States Federal Communications Commission is WM9XPA. The morse code signal transmitted by LightSail 2 corresponds to this call sign. The recorded downlink signal showing the morse code pattern is below.

Morse Code

Updated orbit predictions will allow more accurate antenna pointing, and the acquisition of telemetry data from the radio signal. The next tracking opportunity for flight team tracking of LightSail 2 is over the Georgia Tech tracking station is at about 00:30 UTC (July 2 5:30 PM PDT).

View of Mexico from Lightsail2

July 3, 2019: LightSail 2 Telemetry Decoded

Telemetry beacons from LightSail 2 were decoded by a Riyadh, Saudi Arabia ground station, part of the SatNOGs amateur satellite network. The initial health and status assessment showed that the spacecraft was in good condition, with the batteries well charged. Temperatures were within the expected ranges. The attitude control subsystem was in detumble mode, angular rates were low (< 2 deg/s), and torque rod currents were low (z-axis torque rod current was zero). The +Y solar panel was shown as deployed in telemetry, however, this is most likely due to slack in the Spectraline that restrains the solar panels.

The flight team established two-way communications with the spacecraft and began housekeeping activities. The recurring autonomous antenna deployment process was disabled, and a file listing was downlinked. The error in the spacecraft clock was determined, and an updated two-line element (TLE) orbit state was uplinked to the spacecraft. The spacecraft clock was then updated, which resulted in an expected spacecraft reboot. Detailed attitude determination and control system information was downlinked for assessment. Commands to downlink stored telemetry files were unsuccessful, possibly due to errors in antenna pointing and uncertainty in the TLE. The positioning of the Cal Poly tracking station will be verified during daylight hours. It is expected that the 18th Space Control Squadron will publish a TLE specific to LightSail 2 within the next day, which will improve antenna pointing performance.

First LightSail 2 Image

July 4, 2019: First LightSail 2 Images Downloaded

Spacecraft telemetry indicated that LightSail 2 continues to be in a good state of health. The flight team attempted to turn on gyros that are mounted on the payload interface board (PIB). The PIB gyros are higher fidelity than the gyros that are included as part of the Intrepid flight processor board. The PIB gyros did not respond to power-on commands; while the commands appeared to be received by the spacecraft, there was no increase in the power load and PIB gyros continued to show zero angular rates. Six test images were acquired from each of the two cameras mounted on the solar panels. The attitude control mode was changed from Mode 0 (detumble) to Mode 1 (Z-axis alignment). Mode 1 aligns the longitudinal (Z) axis of the CubeSat with the Earth’s magnetic field vector, which is a favorable attitude for communications. The command was successful, and the torque rod currents and Intrepid gyro angular rates changed in response. Notably, the Z-axis torque rod transitioned from a persistent zero reading to non-zero values, resolving a concern.

Troubleshooting of communication protocols fixed an ongoing problem with file downlinks, and the flight team was successful in downlinking stored telemetry files and thumbnail test images. A test image thumbnail is shown above. Overnight, the Cal Poly student operators continued to downlink stored telemetry files and began the downlink of a full resolution test image.

A LightSail 2 two-line element has not yet been provided by the 18th Space Control Squadron. The flight team continues to use the TLEs for object A.

Goal for today’s July 4 shift include a retry of the PIB gyro on command, and an autonomous momentum wheel test.

Anomaly List:
1. +Y solar panel shown as deployed in telemetry. Resolved: Beacons on July 3 showed all solar panels as stowed.
2. Spacecraft not responding as expected to file downlink commands. Resolved: Communications protocols update allowed multiple successful large file downlinks.
3. Z-axis torque rod showing zero current. Resolved: Transition to ADCS Mode 1 (Z-axis alignment) resulted in non-zero Z-axis torquer current.
4. Cal Poly (Marconi) tracking station unable to decode beacons, despite strong CW signals. Resolved: Ground system connections checked, and several beacons received in subsequent pass.
5. Solar panels indicate spacecraft in eclipse, ADCS algorithm flag (based on TLE) indicate illuminated. Open.

Low Resolution LightSail 2 Image

July 5, 2019: LightSail 2 Activates PIB Gyros

The LightSail 2 spacecraft continues to be healthy. Communications with the spacecraft was sporadic, due to degrading accuracy in the two-line element (TLE) orbital state estimate. We have not yet received a TLE specific to LightSail 2, so we are tracking object 44339 which is thought to be Prox-1.

During the evening of July 4, the flight team was able to activate the Payload Interface Board (PIB) gyros, resolving a concern from the previous day’s activities. Due to the poor antenna pointing resulting from TLE errors, we were unable to complete the planned momentum wheel checkout. The attitude control mode was transitioned to Mode 0 (detumble), to keep the spacecraft angular rates low.

In the current configuration with the solar panels stowed, the magnitude of the magnetometer readings are several times higher than the expected magnitude of the Earth’s magnetic field in the LightSail 2 orbit. This is likely due to magnetic field noise generated by the spacecraft components. As a result, the spacecraft attitude determination is inaccurate. Attitude determination should improve when the solar panels are deployed, and the panel-mounted magnetometer sensors are extended away from the spacecraft components. The magnetic field bias in the deployed configuration has been measured during ground testing and is accounted for in the attitude determination software (based upon a configurable gains file). The planned test of attitude control Mode 2, the mode used for control when solar sailing, will be rescheduled to occur after solar panel deployment.

Today’s activities will be focused on the autonomous momentum wheel test.

Anomaly List:
1. +Y solar panel shown as deployed in telemetry. Resolved: Beacons on July 3 showed all solar panels as stowed.
2. Spacecraft not responding as expected to file downlink commands. Resolved: Communications protocols update allowed multiple successful large file downlinks.
3. Z-axis torque rod showing zero current. Resolved: Transition to ADCS Mode 1 (Z-axis alignment) resulted in non-zero Z-axis torquer current.
4. Cal Poly (Marconi) tracking station unable to decode beacons, despite strong CW signals. Resolved: Ground system connections checked, and several beacons received in subsequent pass.
5. Solar panels indicate spacecraft in eclipse, ADCS algorithm flag (based on TLE) indicate illuminated. Open.
6. Beacon lines 284-286 (computed angular rates) appear to be in rad/s, while the units in the beacon are deg/s. Resolved: This is an error in the ICD & beacon units, but the attitude determination algorithm is correct and internally consistent.
7. Gyro ON command failed. Resolved: Gyron ON command successful on July 5.
8. File uplink for the autonomous wheel test failed. Apparently file uplinks are having the same problem as file downlinks. Open: this may be resolved by the communications protocol update, but not yet confirmed.
9. Norms of +X magnetometer measurements are too high compared to the Earth magnetic field model at the spacecraft orbital altitude (>250 µT compared to the known value of ~50 µT). Open. Possible magnetic field bias due to spacecraft components. Behavior likely to change once solar panels are deployed.
10. Mode 1 behavior not as expected. Angle between +Z axis and magnetic field vector is oscillating around 135 deg +/- 20 deg, based upon computed attitude quaternions. Should be oscillating around 0 deg. Open. Possible attitude determination error due to significant magnetic field bias introduced by spacecraft components in stowed configuration.

LightSail 2 Image of Earth and Sun

July 6, 2019: LightSail2 Solar Panel Deployment

Based upon the Doppler shift of the received radio signal from tracking station passes, the flight team derived an orbit estimate that indicates that LightSail 2 has drifted approximately 35 seconds in its orbit relative to Prox-1. An updated two-line element (TLE) was produced based upon this orbit estimate, and the new TLE was used for tracking station pointing during the July 5 PDT passes. Tracking station performance was significantly improved at Cal Poly. The Kauai Community College tracking station came online, and received a number of telemetry beacons.

The automated momentum wheel test was successfully performed. The script-driven activity commanded the momentum wheel to wheel speeds of 500 and 2000 rpm in each direction, followed by commanded torques of +/- 0.001 Nm. The momentum wheel axis of rotation is aligned with the spacecraft Y-axis. The angular rate history about the Y-axis during the wheel test is shown below. Review of the test data indicates that the momentum wheel is performing as expected.


The flight team then proceeded with solar panel deployment. The panel deployment command was sent, and it was confirmed in telemetry that the panels had deployed. A set of three images was obtained from both cameras, and the image thumbnails were downlinked. Images from Camera 1 (left) and Camera 2 (right) are shown below. The images were taken at different times. Both images show the limb of the Earth as the spacecraft is about to enter eclipse. The Camera 1 image shows objects in the foreground that are thought to be remnants of the restraint line that was burned through to initiate panel deployment. The flight team will attempt to downlink one or both high resolution images.

Following panel deployment, the panel-mounted magnetometers returned magnetic field measurements that are much more consistent with the Earth’s magnetic field model.

Through analysis of earlier attitude control performance in Z-axis alignment mode, the flight team determined that the magnetic dipole moments commanded by the attitude control software are not being properly commanded to the torque rods. This is a subject of ongoing investigation.

Today’s commanding will prioritize the downlink of imaging and stored telemetry. Tomorrow will be focused on attitude determination and control testing.

Open Anomalies:
1. Solar panels indicate spacecraft in eclipse, ADCS algorithm flag (based on TLE) indicates illuminated. Open.
2. Norms of +X magnetometer measurements are too high compared to the Earth magnetic field model at the spacecraft orbital altitude (>250 µT compared to the known value of ~50 µT). Open. Much improved following panel deployment. Awaiting analysis prior to marking this resolved.
3. Mode 1 behavior not as expected. Angle between +Z axis and magnetic field vector is oscillating around 135 deg +/- 20 deg, based upon computed attitude quaternions. Should be oscillating around 0 deg. Open. Possible attitude determination error due to significant magnetic field bias introduced by spacecraft components in stowed configuration.
4. ADCS_Status command reports zeros for magnetometer bias regardless of what the actual biases are in the gains file (determined via BenchSat testing). The DGmag command provides the correct magnetometer biases. Open. Correcting this would require an ADCS flight software update.
5. Torque rods are not being commanded to the proper current levels to provide the magnetic dipole issued by the attitude control software algorithm. Open. This is related to anomaly 3, above.

High Resolution Earth Image

July 8, 2019: LightSail 2 Successfully Updated Torque Rod Control

Last night, we successfully updated torque rod control and confirmed the update in ADCS Mode 1 (Z-alignment). We also downlinked additional images from the sunlit portion of the orbit, and verified that the PIB interface with the deployment motor is operational (done via changing the motor count to 10, then back to 0).

Today, we will address the magnetometer disagreement, and evaluate the updated attitude determination in Mode 1. The next step is to test Mode 2, the attitude control mode that will be used for solar sailing.

High Resolution Earth Image 2

July 9, 2019: LightSail 2 Attitude Determination and ADCS Updated

During prime shift on July 8, the flight team updated an attitude determination and control subsystem (ADCS) configuration table, to provide a positive current to the Z-axis torque rod during Z-axis alignment (ADCS Mode 1). This will result in the +Z axis aligning with the Earth’s magnetic field when in Mode 1.

The +Y magnetometer was disabled, leaving the +X magnetometer as the sole magnetometer used for attitude determination. The total angular rate threshold for ADCS Mode 2 (solar sailing mode) was increased to 8 deg/sec. If the root-sum-squared of the angular rates exceeds this threshold while in Mode 2, the ADCS system will autonomously change to detumble to reduce angular rates.

A command was sent to re-enable battery charging, producing the desired result.

The flight team determined that the x- and y- torque rods are not being properly activated in response to commands from ADCS. This was traced to a magnetic torquer initialization error in the spacecraft flight software. The problem will require a software patch to fix. It was also determined that the attitude determination algorithm was not correctly incorporating the sun sensor measurements. Additionally, the sun sensors report invalid values for the Sun direction when the Sun is not in the sensor field of view. These invalid measurements are not being properly filtered from the attitude determination process.

Based upon these new findings, project management has decided to delay solar sail deployment to no earlier than July 21. This will allow the team to methodically work through the ADCS issues, perform the necessary flight software updates, and compare flight system performance with simulations to ensure proper performance.

The spacecraft remains healthy. LightSail 2 has been identified by the 18th Space Control Squadron, and has been given the identifier 44420.

High Resolution Image

July 10, 2019: LightSail 2 Software Update Released

Today, the flight team developed and tested flight software updates related to the attitude determination and control subsystem, as summarized in the table below. The team is working toward completion of testing of all software updates for version 7, and will begin the uplink process this evening. Uplink of the full software patch is expected to take approximately two days. These changes address all of the critical anomalies that have been observed to date for ADCS modes 0 (detumble) and 1 (Z-axis alignment).

Software Update Tested on BenchSat? Software Version
Fix torquer initialization error Yes v7
Change to feed sun sensor angles to ADCS Yes v7
Update sun vector calculation from sun sensor angles input Yes v7
Change to sun sensor voting v7
Addition of sun sensor voting to ADCS_Status v7
Enable passive telemetry reporting for disabled ADCS sensors Yes v7
Longitude encoding fix Yes v7
Updated ADCS gains file to correct sun sensor and magnetometer coordinate transformation matrices v7
ADCS algorithm eclipse check fix No v8

Mission operators continued to downlink stored telemetry and images. A new two-line element (TLE) was uplinked to the spacecraft.

The team will complete the testing of all software to be included in the v7 update today, and will procede with uplinking the software patch.

LightSail Deployment Image

July 12, 2019: LightSail 2 Mission Update

Over the past two days, the flight team has developed and tested attitude determination and control subsystem flight software updates to address problems seen in flight and in lab-based testing using BenchSat. Software patches have been transmitted to the spacecraft, and version 8 of the ADCS flight software was activated on board last night. This version corrected several identified errors in the ADCS software, including torque rod initialization, eclipse entry/exit prediction based upon the on-board trajectory propagation, and scaling of sun sensor readings so that they can be properly utilized by the attitude determination software. Also, software was installed that allows attitude determination sensor readings to be included in telemetry, but not utilized in attitude determination calculations. A correction was also made to the longitude reported in telemetry.

Parameters in the ADCS gains file were updated, which should allow proper torque rod commanding. The new gains file will be uplinked later today, and the attitude control performance in detumble mode will be assessed.

Additional testing using BenchSat is ongoing to demonstrate proper sun sensor voting (a process that eliminates spurious sun sensor measurements), and to assess whether torque rod activation is being properly phased so that the magnetic fields introduced by the torquers do not interfere with the ambient magnetic field measurements using the magnetometers.

The team is working toward having all identified ADCS issues addressed by the end of shift on Sunday, July 15. If the on-orbit performance shows that the corrective actions have been successful, the team will then move on to assess ADCS Mode 2, which controls the spacecraft orientation relative to the Sun direction using a combination of the momentum wheel and torque rods.

LightSail 2 In Front of Sun

July 13, 2019: LightSail 2 ADCS Gains File Activated

During last night’s passes, the updated ADCS gains file was uplinked and activated. Following activation of the new gains, the attitude rates dropped as expected in detumble mode, as shown below. In particular, the Z-axis angulr rate dropped from over 7 deg/s to 1.25 deg/s. Torque-rod based attitude control is now behaving as expected in detumble mode.

adcs file

An increase in torque rod current was not verified in telemetry. This could be due to a telemetry reporting error, or it could be that the torque rods were not active when the beacon packets were generated. The flight team will download stored telemetry to see if there are torque rod current evident, and they will also send a payload_status command to the spacecraft to get higher resolution torque rod current data.

The team continues to investigate the sun sensor voting scheme. This algorithm is part of the attitude determination process, and is used to exclude sun sensor measurements that are out of family with the other sun sensors. Using BenchSat, the team constructed a test with three sun sensors and light sources that should have resulted in agreement between the sensors, once the unit vectors to the light sources were converted to the body-fixed frame. The voting process did not return a positive vote on BenchSat. Code review has not uncovered any problems, so this issue remains unresolved. All five sun sensors are now operating in “passive” mode on the spacecraft. In the passive operating mode, the sensors report data, but the sensor measurements are not used in the ADCS process.

During tonight’s tracking passes, the higher-fidelity gyros on the payload interface board (PIB) will be activated, to provide additional insight into the spacecraft attitude dynamics.

On Sunday, an ADCS software patch will be uploaded to update the transformation of sun sensor measurements (angles in the sensor x-z and y-z planes) to unit vectors in the sensor frame, fixing a discrepancy found during testing. The team plans to transition the spacecraft ADCS to Mode 1 (Z-axis alignment) and assess performance. A hard reboot of the spacecraft will be performed to reset the long-duration watchdog timer.

LightSail 2 Deployment Thumbnails

July 14, 2019: LightSail 2 Hard Reboot Performed

The LightSail 2 spacecraft was allowed to perform a hard reboot based upon its long-duration watchdog timer on July 14. During the July 14 prime shift, the flight team commanded the spacecraft into ADCS Mode 1, Z-axis alignment. Proper torque rod commanding was observed, and +X magnetometer readings showed that the measured magnetic field was within 20 deg of the spacecraft’s Z-axis.

adcs Mode 2

The ADCS mode was then changed to Mode 2, solar sailing mode, in order to demonstrate the functionality of the attitude actuators. In Mode 2, the momentum wheel is used to rotate the spacecraft about the +Y axis, while the torque rods are used to provide supplemental torques. LightSail 2 successfully transitioned into Mode 2, and the momentum wheel was seen to spin up. The Mode 2 test duration was 20 minutes. Stored telemetry data was downlinked covering the test duration, and is currently being analyzed. The wheel torques commanded by the ADCS algorithm, and the actual wheel torques implemented, are shown below. It is noted that in there is significant chatter in the commanded torque, and the wheel is not very responsive in keeping up with the commanded torque. Additional analysis is being done on the ADCS algorithm to evaluate the behavior.

LightSail 2 Dark Knight Image

July 15, 2019: LightSail 2 Sun Sensor Test Successful

A successful test of the sun sensor voting algorithm was performed using BenchSat, and uplink of the associated ADCS flight software to the spacecraft was initiated.

Goals for July 16 prime shift are to place the spacecraft in Mode 0 (detumble), uplink the most recent two-line element, complete the ADCS flight software update, and transition the sun sensors to an active state.

LightSail 2 Over Mexico

July 20, 2019: LightSail 2 Solar Sailing Analysis Completed

The flight team completed their analysis of the July 17 test of attitude control in Mode 2, solar sailing mode. The Mode 2 test was performed with full attitude determination capability using five sun sensors and a single magnetometer. Sun sensor voting was enabled, and operated as expected. Analysis shows that the attitude estimate is generally good to within 25-30 deg, which is within the expected performance range.

A system reboot and transition to detumble were performed prior to the Mode 2 test. The Mode 2 sequence was then initiated and maintained for one orbit period. The attitude control performance during the test is shown in Figure 1. For the initial 3500 sec of the test, the attitude control performance was poor. At that point, the momentum wheel speed (Figure 2) changed from zero to approximately -400 rpm and from that point onward the desired attitude profile was followed.

Figure 1. Spacecraft -Z axis angle relative to Sun direction.

Figure 2. Momentum wheel speed during Mode 2 test.

The flight team conducted a thorough review of the test results on July 20, including an ADCS expert from NASA’s NEA Scout mission. A key takeaway from this review is that the momentum wheel performance appears to be compromised near zero wheel speed, and establishing a non-zero spin rate for the wheel is desirable.

The team will perform another test of Mode 2 on July 21. This test will incorporate an additional flight software update, incorporating a filtering of the momentum wheel torque commands to reduce jitter. In additional, the momentum wheel speed will be initialized to -300 rpm, and the spacecraft will detumble (damp angular rates) prior to entering Mode 2.

Sail deployment is scheduled for July 22 at approximately 18:30 UTC. Any additional ADCS updates will be made in the sail deployed configuration.
Acquisition and downlink of images has continued, including the image above showing the Baja peninsula.

Image of Earth

July 22, 2019: LightSail 2 Attitude Control Testing

On July 21-22, the flight team conducted additional testing of attitude control Mode 2, solar sailing mode. The team implemented filtering of the commanded momentum wheel torques to reduce jitter. Following the July 21 test, an error in flight software related to the momentum wheel maximum speed threshold was found and corrected via a software patch. The gains for the proportional-derivative control algorithm were updated to increase the expected magnitude of commanded momentum wheel torques.

Results from the July 22 test are shown below. The test was fully successful, and LightSail 2 correctly followed the desired profile, as indicated by the red line in the Figure. Three 90 degree slews were performed during the test.

The flight team is prepared for solar sail deployment on July 23. The nominal time for solar sail deployment is approximately 18:45 UTC.

LightSail 2 Deployed

July 24, 2019: LightSail 2 Mission Declared Successful

The LightSail 2 flight team successfully commanded solar sail deployment on July 23, 2019. At 18:46:11 UTC (11:46:11 AM PDT), in response to a ground command the deployment motor count began to increase, initiating the deployment of the four booms that pull out the four segments of the solar sail. At 18:49:55 UTC (11:49:55 AM PDT) the deployment motor reached the specified full deployment motor count limit. The motor came to rest at a count of 135,786, compared with the full deployment motor count limit of 135,728. The motor count as a function of time during the deployment event is shown in the figure below.

One orbit after sail deployment, the flight team commanded LightSail 2 to enter the solar sailing attitude control mode, Mode 2. In this mode, the sail will turn edge-on to the Sun direction when approaching the Sun, and face-on to the Sun when receding. The flight team observed the spacecraft properly initiate a rotation to the face-on attitude relative to the Sun.

During the final tracking pass opportunity of the day, the team attempted to command downlink of image thumbnails acquired during the solar sail deployment event. The command was not received or acted upon by the spacecraft.

The plan for July 24 mission operations is to downlink image thumbnails and a selected full resolution image. The team will also observe the spacecraft’s attitude control performance in Mode 2.

Solar sail deployment complete.