Astrofriend's homepage


Ad / Annons:

Debris finder, PolCor-2
An instrument to detect debris around young stars


When performing research you finds often that the instruments needed not be available as standard product, especially in the field of astronomy. I have previously written about the astrograph at Observatory Saltsjöbaden, 1930s solution to this problem, I now continues with a new article about an another special camera. The Saltsjöbaden astrograph was manufactured about 80 years ago, this camera which I will talk about now is an ongoing construction and building (2008).

First some background:


Technology development is progressing, you have probably read about it since 1992 when the first exoplanet was detected, today (2017) we have detected over 3500 planets around other stars than our Sun. One technology is based on studying the Doppler shift arising from the star as the planet "pull" on it by gravity when they go into orbit.

There are other things than studying the planets indirectly that is of interest. A very interesting area to research and study is the surrounding of the stars, including how our Sun's planetary system formed and evolved.

Stars dust disc:

The planets are (reasonably) built up by the dust that is in orbit around a young star, usually in the form of a disk. Studying this dust disk in various stages of development around different stars contribute knowledge that can be used to build better mathematical models of how this development takes place. And the need for the new planets we've discovered with the Doppler method does not fit at all into the older mathematical models. It is predominantly very large planets, like Jupiter and bigger, and in many cases in an orbit very close to their star. That there are giant planets discovered today is because today's insensitive technology, the major planet's are easier to detect. But even they must be able to fit into the mathematical models.

Our own solar system as seen from a long distance would look like a dust disk with a hole in the central part where the planets are, a donut. That follows from studies from Earth of our own dust disc from "inside" and try to see the "hole", but this is almost possible to see or at least very difficult. The dust is there, we can see, even without instruments, that is the Zodiacal light. Zodiacal light which can be seen in dark places after sunset, just the Sun's reflection in the dust.

The dust in the same plane as the orbits of the planets, but much more can not be said about the form.

Seeing the dust disk and its shape around other stars can be done with the right technology. The hole has thus emerged as the planets formed from this material, and as vacuumed clean in this area when they grown up. Around our sun should this hole may have a radius of about 40 a.u. (a.u. = astronomical units, equal to the distance from the Earth to the Sun).

Our most distant planet is Pluto with a distance of about 40 a.u. from the Sun. Some say today that Pluto is not a planet, and it may not even not formed in the inner hole in the dust disc, the track that it goes in heavily leaning and more elliptical relative to the other planets.

The contrast between the star's emitted light, and what is reflected in the dust is huge. One way to reduce this problem is to study the thermal radiation that dust emits instead, which still has great contrast, but is somewhat more manageable.

It is this dust disc you want to study around other stars in various stages of development. We dive now into the technology that makes it possible to make these observations.

Debris finder or planet hunter

Image from a test of PolCor-2 (the instrument I will talk about here), a giant star with clear dust counter, the picture is 1' across. The object is no young star but a star with slightly larger mass than our sun, and throwing out the dust at high speed. The object is selected from previous radio observations where the object of dust clouds appear as two peaks (the Doppler effect from the parts of disc that move toward and away from us).

Albanova, research center in Stockholm:

To make this article more interesting I have investigated what is being done at the University of Stockholm at Albanova in this research area. I myself studied, among other things, astrophysics there, one of my teachers was Professor Göran Olofsson I still have some contact with him. It is also Göran who introduced me to the CCD technology in the mid 1990s.

Debris finder or planet hunter

Albanova, this keeps the Stockholm University astronomers after moving from Saltsjöbaden to here in 2001.

A phone call to him and talk about what's going on led to a later visit. Göran is not only theorists but also do practical design and build some of the instruments needed for the research. At his side, Göran also have H.G. for programming control and measurement systems, designing electromechanical and mechanical systems. I had always only heard H.G., but now know that behind that letters is Hans-Gustav Florén.

One of the later instrument that is under development is a camera, among other things to be used to study this particular dust debris around other stars. Technology and theory together make it much more exciting. In addition, this camera operates in the very difficult visible wavelength range and the near-infrared range. We'll see how Göran and H.G. solved this. Let's take a closer look at how this instrument is designed and built. I met Göran in his office in June 2008. Göran has plenty of work and is always stressed. I really appreciate that he has time to meet me. We walk briskly off to the workshop where the camera is completed. Before Göran dare open the sealed box containing the sensitive optical parts, we need to take on a hair net and gloves. Dust and greasy fingerprints is really nothing that belongs in the optical world.

The instrument, PolCor-2:

The actual camera is only one part of this instrument if we talks details. The name of the instrument is PolCor-2, what it stands for? We answer that later. The mechanical device, or the chassis, can be studied in image below. The chassis is built of aluminum and has mounts for mounting on a telescope and is primarily meant to be used in the NOT, Nordic Optical Telescope.

The optical beam from the telescope passes in PolCor-2 unit parts successively as follows (this is an early drawing and differ in details from today's chassis):

Debris finder or planet hunter

The chassis of PolCor-2 holding the camera, the mechanics and the optical parts.

  1. Mechanical fastening plate against the telescope.
  2. Revolver with coronagraph. Here is the telescope's focus.
  3. Plane mirror. Its function is to "fold" together the light beam to save space.
  4. Segment of a parabolic mirror. Make the divergent beam parallel.
  5. Filter holder (not illustrated).
  6. Lyot Stop, Fourier optics.
  7. Polarizer.
  8. Segment of the parabolic mirror, focus the beam again, now towards the camera.
  9. Plane mirror. Folding down the beam toward the camera. The camera is located on the underside.

Ad / Annons:

The mechanical box's surfaces:

Here is the instrument cabinet opened. First Göran shows the instrument's overview and outer parts that are included, in the image below we can discern the internal parts.

Debris finder or planet hunter

Here Göran had just taken the lid off. Look to the right and see the revolver for coronagraph stick up and the camera (grey) to left underneath the instrument.

The entire chassis and the internal parts are coated in a black color / anodizing to dampen reflections. Reflections brings unwanted stray light into the detector and must be reduced to a minimum. The camera has its range of sensitivity from the visible and well into the infrared range (infrared, above 650 nm). The color that appears to be black to the eye does not necessarily mean that the camera see it as black. IR can behave so that the could be reflected in black surfaces and one must carefully select the right type of surface treatment to obtain adequate suppression of reflections over the entire wavelength range to be used, not easy when they directly to the eye not appears.


In the image below you see the coronagraph and here you did have part of the explanation of the name of the instrument, Cor of PolCor-2.

Debris finder or planet hunter

Here appears the coronagraph's revolver with the various masks (beam blockers) and mirrors.

Its function is to block the star's direct light, just as when the sun's corona to be studied have the blinding light from the central part blocked. The turret contains a full set of masks to suit different object sizes. As you may recall, I have in a previous tutorial said that all the stars are point objects in normal telescopes. But optics, atmospheric and other things means that the observed angular diameter gets larger. Depending on circumstances, this diameter can be unequal, that's why several sizes of mask are needed.

The mask consists of a very thin layer of metal evaporated on a glass substrate, it is like a small dot in the middle of the filter. The sizes that can be selected on the masks is 1.5", 3" and 6" (when instrument is mounted on NOT). 3" means 3 arc second and one arc second is 1/3600 part of a degree. The mask does not block the light completely, a very small part of the light penetrates to give the opportunity to set the camera's and telescope's, focus and tracking. Three different magnitudes of damping for each size can be selected: 5, 8.75 and 12.5. The blocking dot also has "softed" edges to reduce edge effects. Notice the "pits" in the outer edge of the revolving wheel, they have the function to fix the various filters in the correct position. A motor controlled by a computer, select the correct mask.

Mirror set one:

The mirrors that are included in the system has a very high reflectivity, better than 99.5% Göran enlighten me. They are of two types, the flat one is used to redirect the light beam, and the parabolic divergent and focuses the light beam. The mirror together serves as a relay lens with 1: 1 ratio.


Debris finder or planet hunter

Filter holder to the right.

Image above shows the filter holder in the system. This is activated when one wants to study the narrow wavelength ranges, 0.3 nm to 0.5 nm. Some interesting resonance lines FeI 386 nm, CAII 393 nm, 589 nm Nal and KI 767 nm. Even the more normal filters U, B, V, R and I are available. The filter holder can be loaded with two filters at a time.

Lyot Stop:

We follow the light path ahead and change thus chamber in the instrument box.

Debris finder or planet hunter

At the top shows the unit that compensates away diffraction from the secondary mirror holder.

In image above, at the top is the optical device that reduces the diffractions spikes. The diffraction spikes we've talked about earlier causing trouble here, just like in the amateur astronomer Newton telescope. The spikes gives disturbing light on the vague light from the dust. It lowers the contrast and something must be done about it. Using a technique called Lyot Stop based on Fourier transform, these spikes is compensated away or at least reduced. One thing that complicates the matter is that NOT (the telescope normally used) does not have an equatorial mounting without an azimuth. A good technique for larger telescopes because it simplifies the mechanics but with the disadvantage that the image field rotates. Field of rotation compensates for the telescope itself but spiders (secondary mirror holders) are not. It must be dealt with within this instrument, it is the computer's task to control the servo motor and the rotating filter (Lyot ​​Stop) so it is in the correct position relative the spiders and the spikes it caus. A disadvantage of the Lyot mask is that it screens of some light.

Servo motors:

Apparently this instrument contain a lot of motors. To design and build new motor system from scratch is time consuming and expensive. Göran, who is a very clever person, based on the designs in other areas which can be advantageously used to this. Believe it or not, but part of the servo motors are of the same type used in radio controlled airplanes!


The system includes a polarizer, here did you also had the other part of the instrument's name, Pol of PolCor-2. Its task is to filter out the polarization angle that is most beneficial. Often the angle that give as high a contrast as possible. By turning this filter and analyze the data (image) obtained as a quality measure can be used to obtain the optimum angle.

Polarizer is visible in the image above as the device that follows the Lyot Stop filter. Also this is provided with a servo motor that transfers torque through a timing belt. The servomotor sets quickly filter into the angles of 0, 45, 90 and 135 degrees. It also has a position where it blocks the light path. The final image may consist of several combinations of angles. The polarizer is designed to let through the wavelength range from 410 nm to 750 nm.

Parabolic and flat mirror set two:

After passing the polarizer remains to focus the light beam back and align it with the camera that is mounted on the outside of the cabinet.

Debris finder or planet hunter

The parabolic mirror to the right and the plane mirror (tilted) to the left.

Also here, an off-axis parabolic mirror. See it as a mirror in a Newtonian telescope where only one circular outer part of the mirror is used. The great advantage of this is the output light beam from the mirror is not moving in the same direction as the incoming. In other words, do not descendants equipment sit in the way of the input light beam. A major drawback is that the optics imaging becomes more deformed than the normal parabolic mirror which already severe coma. But you can not get everything, but choose the one that is most optimal for what you want to achieve. Using mirrors instead of lenses also means that the instrument can handle a wider range of wavelengths.

Lucky Imaging:

If you want to study dust accumulations around other stars you study faint objects with very small angular spread. See the introductory image, the entire image field is only 1' (one arc minute) and details of only fractions of the field. Total system resolution thus becomes a very important parameter. A major limiting factor is the atmosphere. NOT telescope located at a carefully selected spot, at 2400 meters hight at La Palma in the Atlantic Ocean, where there is very good properties in this regard. The atmosphere is changing constantly and cause bad seeing. If you can keep the exposure time very short, fractions of seconds you can be lucky and get a picture just when the atmosphere gives small contributory distortion. If you take a lot of pictures with short exposure you can get several sharp images. The technology is called the Lucky Imaging.

Now, this technology is excellent on our own planets around the Sun, which is very bright and one can therefore take short exposures, typical 1/100 second, and many images, shifting to lie exactly against each other and add them up (stacking). The dust emission around other stars, however, is very weak and noise arises whether to keep exposure short. It can be compensated by adding together more than (100 to 1000's) exposures. However, it requires that the signal is significantly higher than the camera read-out noise, so unfortunately it is not for the normal camera under these circumstances.

Ad / Annons:

The camera:

How do you solve this? You have probably seen all of these night vision googles, the Russian with the image intensifier. They have an internal valve which electronically amplifies the signal. That's one way to go, at the binoculars at least. Other they have to much drawbacks to be useful.

Now there is a more refined method where the image light reinforcement is built into the CCD circuit. The technique is called EMCCD, Electron Multiplying Charged Coupled Device, in which an internal electric field detaches a burst of electrons for each detected incoming photon. That in itself gives no less noise but the signal comes up at a higher level so that the camera read-out noise impact is reduced. The camera dynamics however reduced accordingly as the gain is increased. One of the manufacturers who have such cameras is Andor Technology and is also the camera Göran has chosen for this instrument.

Debris finder or planet hunter

Camera shown here mounted in the bottom of the instrument unit. Above the vents visible are the water connection for cooling.

Andor Technology data for his camera:    
Model: iXon, U897  
sensor format: 512x512  
pixel size: 16um  
Frame rate: 34.5 fps  
vacuum TE cooling    
-85oC air / water -100oC    
dynamic range 16 bit  
single photon detection    
back illuminated    
QE up to 95%  
Gain from off to 1000  
Sensor from: E2V  

This camera can be seen as a photon counter, it detects the individual photons. It may, however limit the range of wavelengths to achieve this. An impossible dream before, normal amateur CCD cameras feature a noise of 8 e- to 25 e-. The system's pixel scale mounted on NOT is 0.12" / pixel, with the option to install a Barlow lens with x2 or x3 focal extender.

Something that perhaps many people reacts to are the few pixels the camera have, 0.5 million only. But for professional astronomers, it is not the number of pixel, it's the quality of the measurement it performs that are important. If the camera cost anything? Well, 30,000 Euro you have to pay for it.

Debris finder or planet hunter

Here the camera is disconnected from its mount on the coronagraph.

The computer:

Debris finder or planet hunter

The computer that control the system.

This complex system is controlled by a computer. Here, a standard industrial PC, which provides a flexible and economical solution. It is for including this as H.G. comes in as software developer, many different components to be controlled, measurement data to be collected and processed.

Test of the instrument that has been done:

Göran tells:
"We typically get a resolution of 0.7" and with Lucky Imaging 0.4" (camera mounted on NOT, 2.5 meters mirror). You can also push the system up to 0.2" with image selection and deconvolution but it increases the noise and the dynamics are reduced. With this camera has basically the internal noise reduced to zero."

"One thing that must be mastered is the backlight (light pollution and the noise from this). One method that is common in IR is to move the image field (chopping and / or node) to the side where no object are and take a picture. This image is then subtracted from the object image. Proven technology, but more than half of the exposure time is lost in the overhead. The optical chopper carrying out this is not on the pictures we've seen here. Tests show that the instrument PSF (Point Spread Function) behaves well and the instrument is well suited for high contrast imaging."

The test has been performed including the red halo surrounding a compact blue galaxies and verified, object Mrk900. Simply put, the focused image is well composed. In this test the camera's gain is set to 100, and 12,000 images were collected respectively. Of these 85% were used. The frame rate was set at 10 Hz. To read the full report there is a link at the end where it can be downloaded as a PDF file, written in English.

Debris finder or planet hunter

Mrk900 in Visual filter (left) and Infrared filter (right).

The instrument has also been used successfully to detect AGB envelops (star set to become a planetary nebula), and more normal high-resolution images without the pole filter and the coronagraph mask that normally sits in the path. The instrument, however, is in an early phase and still has no measurements made on young stars and the dust ring that was written about in the beginning of this article. Watch NOT's website if something comes up there in future.

Ad / Annons:


Should we conclude PolCor-2 instrument we can say that it is an instrument cluster consisting of Lucky Imager, Polarization meter and Coronagraph with the following advantages over a more general instrument as ALFOSC (you can read about ALFOSC at NOT's website):

  • Sharper images, typically 0.7" to 0.4" with the opportunity to come down to 0.2"
  • High time resolution, milliseconds if necessary (sub-array)
  • Much better PSF (higher contrast)
  • Much better coronagraph performance
  • Much lower interference wings
  • More accurate sky subtraction
  • More accurate color index measurements
  • No time loss for reading

Göran says:
"If one were to name one drawback is that QE in the UV range is limited to about 30%, and that the field of view is limited to 1', see image in the introduction."

After this day at Albanova one can not be other than very impressed by the instrument that has been designed and built here at Albanova. And you shall know that Göran has done the most amazing things besides this!

End of demonstration:

After this tumultuous day in the technology front area concludes Göran to reassemble the lid.

Debris finder or planet hunter

Göran mounts back the protective lid again.

Many thanks to Professor Göran Olofsson of all the material and the time he dedicated to making this article possible.

Lars Karlsson

Useful links here:

When I google at PolCor-2 today (2017) I can't find any recent observation with it, maybe it has been replaced with a new version.