# How Hubble Was Deployed: STS-31 and the Release from Discovery’s Arm

> Hubble was lifted out of Discovery’s bay by the shuttle’s robot arm and let go at 19:38 UTC on 25 April 1990, one orbit late because a solar array stopped half unrolled. The published clock, and what a 3D replay has to guess.

By Space Radar. Published 2026-11-13. Category: Missions & Spacecraft. Tags: Hubble, STS-31, Discovery, Canadarm, mission replay. 7 min read.
Canonical: https://blog.spaceradar.ai/hubble-deployment-sts-31-release/

Hubble was deployed on 25 April 1990, the second day of Space Shuttle Discovery’s mission STS-31: the crew used the shuttle’s robot arm to lift the telescope out of the payload bay, unrolled its solar arrays and unfolded its antennas, and released it at 19:38 UTC, 31 hours 4 minutes after lift-off, into an orbit about 613 km up. The release came one orbit later than planned, because one of the two solar arrays stopped part-way and had to be freed.

That 19:38 is almost the only clock time of the day that can be sourced. [Space Radar’s replay of the deployment](https://www.spaceradar.ai/#moment=hubble-deployment) is built around that fact: it uses the published release and launch times, and every other time in it is the app’s own, with captions that say so. This post lays out what is published, what is guessed, and the steps in between.

## The published clock

- **Launch:** 8:33:51 am EDT, 12:33:51 UTC, on 24 April 1990 from Launch Pad 39B (NASA’s STS-31 archive page). The replay counts mission seconds from that moment.
- **Release:** 19:38 UTC on 25 April, in Jonathan’s Space Report no. 36. That is 111,846 seconds, 31 hours 4 minutes, after launch. America Space gives 3:37 p.m. EDT, which is a minute earlier than 19:38 UTC; the replay quotes both.
- **The plan:** release on the 19th orbit; it slipped by one orbit (NASA’s press kit and Jonathan’s Space Report).
- **The orbit:** about 330 nautical miles up at 28.45 degrees (NASA); the replay draws it as a circle of 613 km at 28.45 degrees.
- **The telescope:** about 13.2 m long and about 10,800 kg at launch (NASA).

## What the replay guesses

The sources give the release and not much else, so the replay’s opening caption says it outright: where Discovery is over the Earth, and every clock but the release, is the app’s own. That covers the times of the grapple, the lift, the unrolling, the stop and the freeing of the stuck array, and the place along the orbit, which is chosen so that the shuttle is in sunlight.

It also covers the arm’s joints, drawn as a two-link arm that reaches each place; they are not telemetry. The latches that held the telescope in the bay (two on the sills and one on the floor) are the app’s in shape and in time. The telescope’s own arrays and antennas are drawn from photographs, since no drawing of how they were stowed was found. One check holds at every second of the replay: a test checks the arm’s end effector against the telescope’s grapple fixture to within 0.2 m, so the arm is always holding it.

![The Hubble Space Telescope held above the open payload bay of Space Shuttle Discovery on the long white robot arm, with the blue Earth below and its solar arrays still furled.](/media/hubble-deployment-sts-31-release/arm-lift.jpg)
*Soon after the lift out of the berth: the arm holds the telescope over the open bay. The arm’s posture is the app’s drawing, not telemetry.*

## The steps, in order

The replay opens on flight day 2, 613 km up, with the bay open and the telescope sitting in its berth. Its chapters follow the steps.

1. **In the bay.** The arm’s end effector closes its snare on a grapple fixture beside one of the telescope’s antenna drives. Hawley works the arm from the aft flight deck.
2. **Out of the bay.** The telescope switches to its own power and the shuttle’s supply to it is turned off. The arm then lifts it straight up out of the berth.
3. **The arrays.** The solar arrays, rolled on drums and folded along the tube for launch, are swung out from the sides. The first unrolls along its booms to about 7 m by 2.5 m. They were golden-yellow, and they were replaced in 1993.
4. **One stops.** The second array comes out part-way and stops.
5. **The antennas.** The two high-gain antenna booms spring out, downward and outward.
6. **To the release point.** The arm swings the telescope upright over the bay. Hubble’s pointing systems are switched on and tested while the arm still holds it.
7. **Let go.** At 19:38 UTC Hawley opens the snare.
8. **Backing away.** Discovery pulses its thrusters twice to raise its orbit a little, so that it drifts steadily behind.

![Hubble held over Discovery’s open payload bay on the arm, one golden-yellow solar array fully unrolled and the other only part-way out, over the blue Earth.](/media/hubble-deployment-sts-31-release/array-stops.jpg)
*The array that stops: the first is out whole and the second has come only part-way. Which array stuck is told two ways, and the replay draws one of them.*

## The array that would not open

Here the sources are honest about disagreeing, and so is the replay. Jonathan’s Space Report says the starboard array was the problem and that the preparation of a spacewalk was cancelled after the final attempt to command it open succeeded. America Space says the first array unfolded normally and the second opened a short distance and stopped, and that Mission Control suspected a software fault in a tension-monitoring module, which an engineer had set to do nothing; the array then unfolded. Which array it was is told two ways, and the one the replay draws is the app’s choice.

While the ground tried again, McCandless and Sullivan waited in the partly depressurised airlock in case a spacewalk was needed to free the array by hand. The orbit went through the Earth’s shadow and out, and the release, planned for the 19th orbit, slipped to the 20th. How long the wait lasted is not published; the replay’s duration for it is its own, and it plays that stretch at up to ×600.

## After the release

Discovery stayed near Hubble for about two days as a precaution; the plan was to trail it at about 40 nautical miles in case the aperture door gave trouble. America Space reports that the flight deck, struck by the view as the telescope drifted away, shouted for a camera.

The telescope’s troubles were not over. Its main mirror had been ground slightly too flat at the edge, because a measuring device used while polishing it malfunctioned, and the flaw showed soon after launch. On the first servicing mission, in December 1993, astronauts replaced the first camera with WFPC2 and installed COSTAR, a set of movable corrective mirrors, which restored sharp focus. NASA counts five servicing trips in all.

## How to watch it

[Open the Moment](https://www.spaceradar.ai/#moment=hubble-deployment) and press play; it runs about seven minutes for some two and a half hours of flight, so the player prints the speed-up beside the clock the whole time. Chapters jump to each step. Links to a second use mission seconds from launch: [the array that stops](https://www.spaceradar.ai/#moment=hubble-deployment&t=105349) and [the release](https://www.spaceradar.ai/#moment=hubble-deployment&t=111824) each open the player paused on that second. You can also open the shuttle’s pad, Launch Complex 39B, or Hubble’s own card on the map and choose “Replay this moment”.

*The screenshots are Space Radar 1.7.9 (MIT licence) frames rendered by the app. The telescope and the Canadarm are built in code from published dimensions, and the orbiter model is the project’s own. No radio recording plays: NASA’s STS-31 air-to-ground audio was not found as a clearly licensed file, so the words are captions.*

**Fly there in the 3D map:** [Watch Hubble’s release from Discovery in 3D](https://www.spaceradar.ai/#moment=hubble-deployment). Free, in your browser, no account.

## Questions people ask

### When was the Hubble Space Telescope deployed?

Space Shuttle Discovery launched on mission STS-31 at 8:33:51 am EDT (12:33:51 UTC) on 24 April 1990. Hubble was released from the shuttle’s arm at 19:38 UTC on 25 April 1990, 31 hours 4 minutes after lift-off, in an orbit about 613 km up at 28.45 degrees to the equator.

### Who released Hubble from the shuttle?

Mission specialist Steven Hawley worked the robot arm from the aft flight deck, lifted the telescope out of the payload bay and opened the arm’s snare to let it go. The STS-31 crew of five was Loren Shriver, Charles Bolden, Steven Hawley, Bruce McCandless and Kathryn Sullivan.

### Why was Hubble’s release delayed by an orbit?

One of the two solar arrays opened part-way and stopped. Controllers tried again and again while the orbit went through the Earth’s shadow and out, and McCandless and Sullivan waited in the airlock in case a spacewalk was needed. A last command from the ground worked, the array unrolled, no spacewalk was needed, and the release slipped by one orbit.

### Did Hubble work right after it was deployed?

Not fully. Its main mirror had been ground slightly too flat at the edge, a flaw that showed soon after launch. It was corrected in December 1993 on the first servicing mission, when astronauts installed COSTAR and a new camera, WFPC2.

## Sources

1. [STS-31 (NASA Space Shuttle archive)](https://www.nasa.gov/mission_pages/shuttle/shuttlemissions/archives/sts-31.html)
2. [Jonathan’s Space Report no. 36](https://www.planet4589.org/space/jsr/back/news.36.txt)
3. [Somebody Get a Camera: remembering the deployment of Hubble](https://www.americaspace.com/2020/04/25/somebody-get-a-camera-remembering-the-deployment-of-hubble-otd-in-1990/)
4. [STS-31 press kit (NASA)](https://science.nasa.gov/wp-content/uploads/2023/06/sts-031r-press-kit.pdf)
5. [Hubble overview](https://science.nasa.gov/mission/hubble/overview/)
6. [The spherical aberration problem (ESA/Hubble)](https://esahubble.org/about/history/aberration_problem/)
7. [Space Radar](https://www.spaceradar.ai/)
