Orbital Focus - International Spaceflight Facts and Figures
carousel image
North Korea

Tyneside, UK
2024 May 20
Monday, Day 141

Curated by:

Common Threads

Mission 5

Mission 4

Mission 3

Rocket Trajectories

Kwangmyongsong 3 - The Satellite

What do we actually know about the design of Kwangmyongsong 3, the satellite orbited by North Korea on December 12. The answer is actually "very little".

Kwangmyongsong 32012 April, visiting journalists were shown a satellite that was described as being Kwangmyongsong 3.

The object orbited in December is referred to by North Korea as Kwangmyongsong 3 Flight Unit 2. The general assumption is that it is a second model of the same satellite but, as yet, no photographs have actually been released so it is presently unconfirmed.

In the photograph, an engineer is pointing out salient features to the the camera operators who seem to outnumber the jounalists. He also provides a general indication of the object's size.

What follows is not a particularly technical review but it picks out some of the satellite's features. Comments and corrections will be much appreciated.

The April Satellite - Physical Description

Kwangmyongsong 3Shown to journalists was a cuboid device with solar cells on at least three of the long sides. It has the feel of an engineering model rather than an active satellite.

Scaling up from photos that include a North Korean engineer, it is about 0.65m square and 1m long. The upper square face, that would probably be Earth-pointing in real life, carries sensors and aerials. The opposite face is attached to a circular adaptor that may be an analogue of the mounting for attaching it to the launch vehicle. Several web sources say North Korea indicated that the mass is about 100 kg.

The general feel is of a satellite designed to operate with its long axis vertical where the end that carries sensors points down towards the Earth.

It appears to be an unpressurised design where the components are exposed to the space vacuum. If it were pressurised, it would more likely be cylindrical and possibly have domed ends. It is conceivable that the box hides a pressure vessel but it is not too likely.

As regards a detailed description by the engineer to the assembled journalists, it seems either the presentation was very basic or language difficulties got in the way. There are no web sites carrying reports in detail of what was said but the gist of the presentation seems to have been that it is an imaging satellite and is three axis stabilised. Some reporters came away with the impression that it was the actual vehicle that was to go into orbit but, again, it's an unlikely scenario seeing that launch was only hours away.

Solar panels

Kwangmyongsong 3 hingeIn photos, and to the journalists invited to view it, the three visible long sides were covered with solar cells. The fourth of the sides was never revealed. It is not possible to say whether it too was a solar panel because, even though the satellite was mounted on a wheeled trolley, it was never turned round.

Of the three visible panels, the two lateral ones are hinged on the two long edges visible in photographs. There is a single hinge at each end of the join. One interpretation is that they open out once on-orbit to form a three-segment flat panel.

An alternative explanation for the hinges is that the two sides are doors than can be opened to allow engineers to reach inside the satellite in order to work on internal equipment. There are what could be a door catches holding it closed.

Radio Aerials

Kwangmyongsong 3 aerialsThere are at least four aerials on upper surface of the model and possibly as many as six.

There are four rod-like structures, two pairs of differing lengths. The longer, silver coloured, ones are for an uplink at around 400 MHz, and the shorter gold coloured pair is for transmission, maybe at the hinted frequency of 470 MHz.

There is also a cylindrical housing on a stalk that resembles an X-band (around 9 GHz) aerial in an enclosure. What others have interpreted as a waveguide slot on the side of the cylinder is, in fact, a black label with silver or white printing but there could be a slot on top that would point vertically downwards when the satellite is properly oriented in space.

A fifth device is a small disc on a stalk. It may feed a GPS receiver but that conclusion is uncertain.


Kwangmyongsong 3 camerasTwo are visible. In the centre of the upper surface is a short cylinder that could be the lens of an imaging camera. It points directly out of the body and would need to point directly downwards to obtain images of the Earth's surface.

There is one mystery item, a tapered cylinder with a feed cable leading into the body, that looks like another imaging device pointing at right angles to the longitudinal axis of the body.

If it is an imager, it could be for sensing the Earth's limb, or it could be for capturing pictures of star fields to help determine where the Earth imaging camera is looking. It looks away from the surface of the cuboid that carries hinges for the side panels. If the solar array does actually fold out into a flat triple panel arrangement then, when the panel is facing the Sun, this sensor faces the dark sky in exactly the opposite direction.


Tsikada satelliteInitial stabilisation would probably involve permanent magnets interacting with the Earth's magnetic field to push the satellite towards its designed orientation. Once there, two options present themselves - three-axis orientation or, maybe, a gravity gradient boom.

Three axis stabilsation is usually achieved either by the use of thrusters, or by momentum wheels (gyroscopes). There is no evidence of thrusters so gyroscopes are a possibility. This form of orientation control is a 'must' if the intention is to point one of the cuboid's long faces permanently downwards towards the Earth.

Of the options available, a gravity gradient boom is the simplest. Once an approximate orientation is achieved by the magnets, the boom with a mass at one ended is extended in the upward direction, driven by a motor. It will then keep the face with sensors pointing towards Earth. The longer the satellite body, the longer the boom can be and the more efficient the stabilisation.

Examples of an efficient gravity gradient stabilised satellite are Russia's Parus, Tsikada and Nadezhda satellites. The satellite in the photograph is a Tsikada.


On balance, it seems that Kwangmyongsong is designed to operate with the instrumented face pointing towards the Earth. A major pointer is that none of the aerials is repeated on the opposite face of the cuboid. Any other orientation would need aerials on two sides to avoid loss of signal when the satellite body blocks the line of sight from one of the sets to a ground station.

If the hinges on the model are there to allow the sides to fold out into a single large panel then three-axis stabilisation is essential as the panel must be constantly and steadily oriented towards the Sun.

With a simpler combination of magnetic and gravity gradient stabilisation, a fold-out solar panel could prove very inefficient. Unlike the situation with three-axis stabilisation, the satellite is free to rotate around the longitudinal axis so the panel could move to a position where the sun hits the 'wrong' side and the electical output drops to zero, no longer charging the storage batteries. With solar cells on all sides, one or two sides are always illuminated by the Sun.

Radio Transmissions

As of December 16, no reliable reports have been forthcoming of transmissions from Kwangmyongsong. Amateur tracking stations around the world have put much effort into monitoring frequencies between 400 and 500 MHz, especially around 470 MHz where signals were expected.

It is possible that the satellite is being commanded to switch on and off only while over North Korea. This is unlikely to be the case. For one thing, because of early mission tumbling (more on this later) there is no guarantee that commands will be received properly as the aerials could be masked by the satellite body.

If it were transmitting over Korea then the likelihood is that it would have been reported by the South Korean government and/or amateur radio trackers in Japan, for example.


Unfortunately, with this launch being politically charged, the 'opposition' is going out of its way to belittle the accomplishment. Soon after launch, a US source suggested Kwangmyongsong was tumbling "out of control". Journalists and reporters tend to pick on the word 'tumbling' to mean a violent motion. In satellite terms, tumbling is simply simultaneous rotation around two axes. The rotation, and hence the tumble can be quite gentle if the rate is slow.

December 15, Greg Roberts made a visual observation of Kwangmyongsong and the upper stage of its Unha 3 rocket. The rocket seemed quite stable but the satellite was indeed tumbling. He noted that its brightness varied with a period of about eight seconds but the pattern of the variation suggested a rotation (or tumbling) period of about 16 seconds. So the tumble is not violent but we will have to wait and see if it settles down or is indeed uncontrollable.

Politics and Posturing

Because of the secrecy and occasional sleight of hand by North Korea, and the politically charged responses from other countries, it is very difficult to tease out the facts.

The day before launch occurred, North Korea announced it had been delayed, and extended the window by seven days. At the same time, it appeared the US was wrong footed when North Korea made it look as though the rocket was being disassembled for the benefit of reconnaissance aircraft and satellites.

The Philippines accepted North Korea's launch announcement at face value and passed on navigation warnings for second stage re-entry of a satellite launch vehicle. South Korea issued the warnings also but referred to a missile launch, and Japan issued Notices to Airmen warning that they should avoid flying near the country's ABM sites.

Because of the radio silence and little information coming out of North Korea, a slight possibility exists that the satellite it orbited is not an operational vehicle at all and is a dummy. There is equally no certainty that what was shown to journalists back in April was a genuine satellite.

James Oberg, who visited Pyongyang in April, questioned whether mission control might be a dummy in itself. He commented on the large console type desks in the room having nothing inside them. Unfortunately such comments ignore other control room examples like Arianespace and SpaceX where the only equipment at each operator position is a small console supporting a keyboard, a monitor and a PC base unit. It's for others to speculate on why North Korea's are so large.

Put all these together and it leaves a lot of doubt and questions.

More to Come

The story is by no means at an end. More information may be forthcoming from North Korea and it is possible the radio transmitter may spring into action.

More visual observations are to come that may confirm whether Kwangmyongsong stabilises, and whether or not the solar panels have opened or remain wrapped round the body.


Thanks go to Greg Roberts for allowing me to quote from his observing notes and to Sven Grahn for providing opinions on some of the external devices on April's Kwangmyongsong 3 model. However, any major gaffes are mine, not his.

Page date: 2012 Dec 15

Copyright © Robert Christy, all rights reserved
Reproduction of this web page or any of its content without permission from the website owner is prohibited