| v1.1 - draft 19/11/1998 v1.1a - additional link to CCDUSER.HTM in Task 3 J.M 28/12/98 |
author r. Easto |
|---|
This document is a training guide and checklist to help someone operate the CCD camera on the Kenley Telescope.
This document is still in draft mode as it was written from memory, please let the author know of any errors or improvements that could be made.
Prerequisites: this document assumes that you are familiar with the operation of the 18 inch telescope. In order to store images on a computer for processing and saving you will need to operate the PC computer in the annexe at Kenley and thus will need rudimentary computer skills.
The CCD camera is a very sensitive electronic camera that is mounted on the 18 inch telescope to take deep sky and planetary images. The camera may be mounted on the f4 or f20 focusses or onto telephoto lenses for wide angle shots.
Go through this checklist ticking every task as you complete it. You may need to ask an experienced observatory person for clarification on some of the tasks.
| Your Name | _________________________________ |
| Date | _____ / _____ / _________ |
|---|---|
| find the CCD Hardware User Guide. It can be found on the Computer in the Annexe under the Object 'Internal Documentation'. | At the moment it can't so you need to get a paper copy from the author. |
| First we need to locate all parts of the CCD system. Find the CCD camera, it should always be in the black box in the observatory annexe | |
| There are three sockets into the camera head, a UHF lead for the image to come out of; a 9-pin power-in socket and a 15-pin socket for signals to drive the read-out electronics in the camera. A 4th object is a plug for putting drying crystals into the CCD camera - do not unscrew this. | |
| The CCD is cooled to make it more sensitive, the heat sink on the back of the camera radiates the heat away from the CCD chip | |
| The CCD camera should be attached to its focusing mount. The two are connected by a Pentax camera screw thread. Separate the two halves. | |
| Find the set of three camera lens extension tubes in the eyepiece cupboard (they are probably all screwed together). The middle one of these can be attached to the camera and a lens attached to the other side for wide angle images. Put the extension tubes back into the cupboard. | |
| With the focusing mount still unattached look at the dust on the glass surface that protects the CCD chip. This dust will appear out of focus on the images that we take later. | |
| Through the front of the camera locate the CCD chip. This is the detector, it is expensive and sensitive to mistreatment. Behind the chip is the active electronic cooling device that connects to the heatsink but you cannot see this. To the side of the chip is the electronics that reads the image from the CCD and transmits it to the frame store. Connect the CCD back onto the focusing device. | |
| Next find the frame store, it is in the rack of electronics in the annexe. | |
| Toggle the power switch on and off to ensure that power is connected. Switch on the mains power to the frame store if necessary. Now ensure that the power switch is OFF. We are going to connect up the CCD camera to the frame store. When the CCD was bought we saw the manufacturer connect up with the power on and this blew the CCD chip costing several tens of pounds and these days may be irreplaceable. This is the only important point to note with the camera. | |
| Take the CCD camera into the observatory dome to the newtonian focusing position of the telescope. | |
| locate the ccd head driver box. This is mounted on the telescope to the your right as you look at the f4 focus position. It is a box measuring about 5 x 4 x 2 inches. A cable from the frame store goes to this box. There are two LEDs on the box, they should not be illuminated, when illuminated they show that the power to the frame store is on, these are just precautionary lights. When lit they do not affect the imaging. | |
| follow the cables from this box. You will see it go under the observatory floor. This cable is attached to other cables that will connect to the CCD camera. | |
| with the telescope in horizontal position. Remove the eyepiece focuser. Place this focuser in the eyepiece cupboard in the annexe. | |
| Fix the CCD camera to the vacant focusing position. It is very important that this is done securely so put all of the lights on first. Check that the camera is properly fixed. | |
| now connect up the three cables to the camera. The order does not matter. They all have different connectors so cannot be connected wrongly. Take great care with the cables, they are delicate and often the cause of problems with the CCD camera. | |
| all of the hardware is now set up. Go down to the annexe and power on the frame store. | |
| Find the temperature monitor on the frame store (at left) and write down the current temperature reading here: | degrees C |
| Power on the black and white monitor on the top shelf using the push switch at bottom right (left of the two switches). | |
Now set up all of the switches on the frame
store as follows from left to right:
|
|
| The frame store is now continually downloading images from the camera and they are displayed on the monitor. You should see the screen flicking every 2.5 seconds. Don't worry if nothing appears on the screen, just as long as it is flicking every 2.5 seconds. | |
| The camera is imaging but could be out of focus, could be pointing at the dome, could be pointing at a place in the sky with no bright stars. It could also be daytime and/or cloudy. If it is daytime wait until it gets dark and clear. We first need a good target object, if possible point the telescope at the moon if not then use a bright star. Use the 1-inch and 4-inch finders on top of the telescope to put the object in the centre of view of the 4-inch finder. | |
| Return to the annexe, has the object appeared on the monitor? | Yes / No |
| If no you could first try altering the contrast and brightness on the monitor. | |
| On a piece of paper draw (freehand) a circle 50mm in diameter. This represents the 50 arc minutes field of view of the 4-inch finder when used with a 25mm eyepiece with a 50 degree apparent field. Now draw in a horizontal diameter though the field, this imaginary line will always be parallel to the top of the box section of the telescope tube. Now draw anywhere in the field a rectangular box measuring 12mm by 8mm, this represents the size of the field that the CCD is capturing when used at f4. | |
| If the object is not in the field, the problem is probably that the 4-inch finder is not aligned to precisely the same point in space as the main telescope. This is quite normal. Looking through the finder move the star upwards half way to the edge of the field. Can you see the object on the monitor now? | Yes / No |
| If not move the telescope to the other seven points in the field corresponding to directions of the compass NE, E, SE, S, SW, W and NW, the same distance from the centre of the field. If the 4-inch finder has the 25mm eyepiece fitted then one of these positions should show the object in the field. Have you found the object | Yes / No |
| If you still have not found the object then you cannot proceed. Go and ask an experienced user to look at the problem. We now assume that you can see an out of focus image of the object on the screen. Don't worry that the object is out of focus, we are going to centre the image on the screen. The telescope drive must be switched on or the object will be drifting across the screen with each successive download. Locate the 4 telescope move buttons on the electronics in the annexe, make sure that the speed control is set to slew (or fast) rather than fine guide. Push the top button for ten seconds and see the object move on the screen. How far did it move approx.? | _______ mm |
| Now push the rightmost of the four buttons for ten seconds. Note the direction of movement and distance of movement on the screen. Did it move about the same distance? | Yes / No |
| Now use all four buttons to centre the object on the monitor. | |
| With the object in the centre of the monitor go back up to the 4-inch finder and look at the position of the object within the field. Make sure that the cross hairs of the finder are parallel with the square box of the telescope tube. Draw the field, the crosshairs and the objects position in the field in the box at right. This will help in locating other objects for the rest of the night. Another option would be to realign the finder, this is not too easy and it may have already been aligned for the f20 focus so this is best left alone. | |
| Finally move the telescope to another object using the 4-inch finder and place this object in the same position in the field. You should be able to locate this object straight away. Congratulations you can now use the CCD to find objects in the sky. | |
| From the frame store compare the current temperature with that which you recorded above. Has the temperature gone down? | Yes / No |
| On a sheet of paper, again freehand, rule two lines around 300mm long in the form of a letter 'X', the two lines should form an angle of around 15 degrees. The lines represent the cone of light coming from the secondary mirror converging at the precise point of focus and diverging again. Draw a line down the centre of the centre of this cone this is the on-axis line. Draw a 15mm long line at right angles to the on-axis line to represent the CCD with half above the on-axis line and half below, this line should be drawn 5mm to the right of the point of focus. | |
| We see that light from our object is spread out over an area on the CCD. To focus we need to alter the focusing mount until the CCD and the point of focus converge. At an f4 focus, in order to focus light onto a single pixel (or picture element), we have to set this position to 0.06 mm. We cannot see this position we have to blindly guess at the position and observe the output. | |
| look at the focusing mount on the ccd. It was made by removing the lenses from an old 50mm camera lens. It has a mechanical advantage of 10. i.e. move the focusing ring around by 1cm and the CCD moves in or out by 1mm. This means that fine focus will eventually be achieved by moving the ring by less than 1mm. With a ruler, measure the rough size of the out of focus star on the monitor and write it here | _______ mm |
|---|---|
| Write down the current setting of the focusing ring | _______ feet |
| On a sheet of paper draw a scale to match the scale of the lens that is used as the focuser, all the way from infinity to 2 feet. | |
| Draw a line from this scale the size of the out of focused image on the screen. | |
| Turn the focuser by perhaps a quarter of a turn in either direction, read the focuser setting and make another mark on the scale the size of the out of focus star. Repeat this perhaps 10 times until you have readings over the whole of the scale. | |
| Can you see a point of convergence on the scale with zero image size? | Yes / No |
| now set the focusing mount to this point. Is the star in good focus? | Yes / No |
| At this point you may wish to try a few minor adjustments on focusing. Note down the focus position setting. | _______ feet |
| As a final test, move the telescope to a bright deep sky object, perhaps a globular cluster, set the exposure setting on the frame store to 11 (for a 40 second exposure) and wait. Can you see your deep sky object? What was it? | Yes / No
Object _________ |
| Try to find the focusing mask in the observatory. It is an 18-inch cardboard disk with three 4-inch holes punched into it. Place this over the end of the telescope with the centre hole mounting the cardboard on the support for the secondary mirror. Defocus the focuser. | |
| Try to focus with this mask in place, do you prefer this method to the standard method? It is just a matter of choice. This task is now complete. |
Before you attempt this task please find and read the 'ccd software user guide'. You need not bother with the section called 'Software and Installation' because the program is already loaded onto the Kenley computer.
The prerequisites for this task are a little knowledge of computers. You also need to know the power-on password of the computer.
| On a piece of paper draw three rectangular boxes in a line. Label them CCD, Frame store and Computer. | |
| images can be stored in all three places. The image forms on the CCD. On request or periodically the frame store transfers the image from the CCD to its own memory where it is shown on the video monitor, in the process the CCD image is erased and a new image starts to build up. The computer can fetch an image from the frame store to its own memory where it can be shown on the computer monitor and saved to disk, when this is happening a snow storm effect if seen on the video monitor. If an image is being copied from the CCD to the frame store at the same time that the frame store is being copied to the computer then the images collide. No harm is done but the result is a mess. | |
| power on the computer. Do you know the power on password? If you don't you cannot get beyond this point. | |
| the computer boots into windows. Find the 'icon' labelled 'CCD Program' and quickly 'double-click' on this icon. This runs the program. | |
| Referring to the 'CCD Software User Guide', we are now going to copy an image down from the frame store. Ensure that the frame store has an image displayed. | |
| from the main menu hit key 'd' for disk menu. Then hit 'F' to 'fetch from framestore'. As you hit 'F' look at the video monitor. You should see the image flicker for a second as the image is copied to an area of memory in the computer. | |
| We have decided that we like out first image and we wish to save it permanently to disk so that the image will remain even after the computer and frame store are switched off. We have returned to the main menu so hit 'D' for disk menu once again and 'S' for 'save file'. Enter a filename of no more than 7 alphabetic or numeric characters. Such as 'delete'. The disk light briefly comes on on the computer as the image is saved to a file on the hard disk. | |
| we have returned to the main menu. Now we want to hit 'S' to 'Show Image' the image is displayed on the computer monitor. When the image is displayed a number of keys affect the image. These are listed in the 'Show Image' section of the CCD Software user Guide. Try out these keys in this order: autocontrast stretch, zoom, toggle smooth display, push to send this image back up to the frame store, toggle smooth display off again, the arrow keys, zoom out, photometry and exit. | |
| We are now going to load a file saved by someone else on a previous night. From the main menu select 'D' for disk menu and 'L' for 'Load File'. A list of images is displayed, you may note our image 'delete' in amongst them. Choose a filename that looks interesting perhaps 'M27A'. The file is loaded from disk into memory. Now go back to show image to look at the image. |
Store all files beginning with test, please also erase all of these files at the end. This will prevent the disk becoming full of unwanted files.
Seemingly shorter this task will probably take a number of nights but will leave you with a full understanding of the system to let you take the best images possible.
| Capture an image of a deep sky object with the gain on 3 and the exposure set to 11 (40 seconds) and label it something beginning with test | Filename: |
| Fit a mask over the point at which the light enters the focusing mount to shut out all light. Take a dark frame. Label the file d34k113 for temperature -34, exposure 11 and gain 3. Replace the temperature part with the current temperature. This image is called our dark frame. Since we used | Filename: |
| Record the current temperature reading here: | Temperature: |
| Retrieve the original image, show the image on the monitor and hit 'C' for an autocontrast stretch. Can you see a lightening in one corner of the screen. If not go to the contrast stretch menu and perform a normal stretch operation entering values of 0 and 1024. This brightens the image by a factor of 4, you should now be able to see a brightening in one corner, write down which corner here: | |
| This brightening is caused by the on-chip amplifier electronics on the CCD chip. The amplifier causes one corner of the chip to heat up more than the rest of the chip and create this apparent glow. This effect appears on dark fields and image fields alike. If we subtract a dark field from an image field then we are left with a pure image. Reload the original image, this is now stored in the memory. From the 'Flat field' menu choose 'Subtract Dark Frame' and enter the name of the dark field. Now go show the image and use autocontrast stretch. Has the 'glow' disappeared from the image? | Yes / No |
|---|---|
| Note: the dark frame and image must have been taken under the same conditions; i.e. same exposure, same gain and roughly same temperature for the results to be meaningful. They do not have to be taken on the same night. This why we have labelled the dark frame something meaningful, if cryptic. Usually you do not have to take a dark frame for every image as there is already a suitable one saved on the computer. Finally from the contrast stretch menu choose log stretch and enter ten. Go back and look at the image. Is the image very grainy? | Yes / No |
| Load your dark field from the saved file once again. Now from the 'Miscellaneous Menu' choose option 'Image Statistics'. Write down the noise measurement for this frame. (it should be about 1.1 - 1.5%) | % |
| This noise figure represents the average percentage difference between all of the adjacent pixels. The smaller this number the better the image and the more it may be enhanced before the noise ruins the picture. When we subtract the dark frame from an image frame we remove the background glow which is good but we add the noise in the dark frame to the image frame which is bad. In other words we have doubled the noise. There is a solution to this. You may have noticed that some of the dark frames are called dnnannn where 'n' is a number between 0 and 9. The 'a' signifies that this is an averaged dark frame. Choose the 'Flat Field menu' and 'Integrate Dark Frame'. Enter the correct dark field name according to the current temperature, exposure and gain settings. i.e filename d34a113 for temperature -34, exposure 11 and gain 3. Write this filename down here: | Filename: |
| enter '10' for number of frames to capture. Ensure that you have a good dark frame to start with. The process will take 10 x 40 seconds = 7 minutes to complete so go and have a cup of tea. You will see the images automatically downloaded and added to the image in memory. At the end hit 'A' to store the average of all the frames. This file is saved to disk by the computer and a copy is left in memory. Using the same statistics routine write down the noise here: | % |
| By taking 10 images, mathematics tells us that the noise should have gone down by the square root of 10. Write down the actual and predicted noise of the 10 frame dark field: | Actual: ______%
Predicted:______% |
| Go back and load your original image file, subtract the averaged dark frame and look at the result. You may not be able to see the difference but there will be only half the noise in the resulting image. The main noise is now purely in the image file. To remove noise further we must take multiple images of the object and combine them. | |
| Read the section from the CCD Software user Guide entitled Taking Better Images. | |
| Use the continuous capture facility to capture 10 images of your object calling them test1, test2, test3, ... , test10. | |
| Use the program 'offset' as described in the same document to add the ten images into one image using your averaged dark frames. |