DSI Pro Mods


	Astronomical Photometry with a DSI Pro Jeff Hopkins Hopkins Phoenix Observatory 7812 West Clayton Drive Phoenix, AZ 85033 phxjeff@hposoft.com 5 September 2006
	Abstract The color Deep Sky Imager (DSI) cannot be used for astronomical filter photometry, The DSI Pro can. To do serious work requires some simple modifications to the camera. These modifications include replacing the filter slide with a filter wheel and adding a simple thermoelectric cooler (TEC). With these modifications the DSI Pro (and DSI Pro II) provide a means to do professional astronomical photometry with a modest telescope even in a light polluted suburban location. This paper describes these modifications and what is needed to do BVRI CCD photometry with the camera. Figure 1 Modified DSI Pro on Meade 12" LX200 GPS Telescope While CCD cameras for astronomical use have been around for more than a decade, it has only been recently that affordable and easy to use CCD devices have been made available to the amateur market. Some of the first of these were web cams. Some folks figured them out and made modifications that produced surprisingly high quality results. Web cams work best with brighter objects such as the planets, Moon and solar imaging. For deep sky imaging the web cams need to be modified. These modifications are beyond many amateurs. When Meade introduced the color Deep Sky Imager (DSI), that changed. Now one could do deep sky imaging without needing to make modifications to the camera. In addition, the AutoStar software that comes with the DSI series is excellent. It's worth the price of the camera alone. Learning the software requires some dedication as the quality of the software far exceeds the documentation. Two problems prevent the color DSI use for filter photometry. Because the color DSI CCD chip has built-in RGB filters, it would be hard or impossible to find proper filters to match the camera to the standard photometric system. The other problem, while mainly a problem during the warmer months, is that the camera is uncooled and operates at ambient temperature. Shortly after the color DSI was released, Meade released a monochrome version, the DSI Pro. The DSI Pro provides a good starting point for a serious CCD photometry. Because the camera is monochrome a 4-filter slide for RGB filters is included. The filters are extra. The advantages of using the DSI Pro with external filters are the monochrome camera is more sensitive and provides better resolution. While the included filter slide can be used to hold photometric filters, it quickly becomes apparent that is not a good idea. The photometric filters are expensive and the filter slide is open exposing the filters to dust, dew, fingerprints and scratches. A filter wheel is a much better way to hold and protect the filters. The first step is to replace the filter slide holder with a low profile adapter plate similar to that used on the color DSI, see figure 2. This gets rid of the large openings where the slide goes and reduce the distance to the CCD chip. These adapters can be purchased from ScopeStuff at http://www.scopestuff.com/ for $37. Note: The original 1.25" nose piece can be used with the new adapter, but it's better to mount the filter wheel closer to the camera than allowed by the nose piece. Replacing the original slide holder adapter with the low profile adapter plate is very simple. There are 4 small Phillips-head screws that hold the adapter in place. Those screws and the original adapter are removed and replaced with the low profile adapter and new shorter screws. Note: There are 4 hex-head screws at the corners of the camera. The hex-head screws hold the camera together. Do not remove those screws for this modification. Figure 2 Original DSI Pro with filter slide and low profile/original adapters With the low profile adapter installed on the DSI Pro the next step is to add a filter wheel. There are several on the market that range for well under $100 to close to $1,000. An excellent compromise is the ATIK Manual Filter Wheel. This can also be purchased from ScopeStuff for $199. The filter wheel is well-made and can hold up to 5 - 1.25" screw-in filters. Figure 3 shows the camera with the ATIK filter wheel and F/3.3 focal reducer. The reason for the focal reducer is to provide a larger field of view for the imager. This makes finding the stars and getting comparison stars in the same view much easier. Figure 3 Modified DSI Pro with filter wheel and F3.3 Focal Reducer For most photometry special filters must be used. The RGB filters for the DSI Pro and DSI Pro II cannot be used as photometric filters. For CCD photometry 5 photometric filters are available. While there are other filters, these are the most common used. They are ultraviolet, blue, visual, red and infrared (UBVRI). Unless you are planning on using a meter-sized telescope, the U filter will not be of much value. as the DSI Pro CCD is not sensitive in that band. Plan on using just the BVRI filters. There are several places you can buy filters. Astrodon has the least expensive yet good quality filters at $309 for a set of BVRI filters. See Schuler UBVRI Photometric Filters from Astrodon at http://www.astrodon.com/products/product.cfm?CatID=4 The DSI series cameras use ambient cooling. With cool night time temperatures these units perform well. Dark frames must still be created and subtracted from each image to get rid of hot pixels, however. During the summer months in Phoenix, Arizona observatory temperatures can approach 100 degrees Fahrenheit at midnight. With these temperatures, even subtracting dark frames does not work well. Adding cooling to the DSI Pro started as an experiment to see what could be done at a minimum cost and amount of modifications. Note: This cooling modification can be used on all the DSI series cameras. In addition to reducing dark current cooling increases the camera's sensitivity. Perhaps the easiest way to cool the CCD is using a Thermoelectric Cooler (TEC). TECs are fascinating devices. TECs consist of a sandwich of semiconductors between two metal plates. By applying a DC voltage across the plates, one plate gets hot while the other one is cooled. It can actually be surprising the amount of heat and cooling produced. Care must be taken when experimenting so that the unit doesn't overheat. A heat sink is very important on the hot side. Thermal grease is needed on both plate surfaces to provide effective transfer of the heat and cooling. Some will note that there is no temperature regulation or sensor. These could be added, but add to the complexity. The lack of temperature regulation is not a problem. If the camera is allowed to run with power on and cooling for 15 to 30 minutes, the temperature will stabilize. The exact temperature is not important. Once the unit has stabilized new dark frames for that evening should be taken. This procedure works well. For parts I recommend the following: 40 mm TEC, part number PJT-7 @ $14.75 12 VDC 3.5 A power supply part number PS-1231 @ $15.85 Heat sink with 12 VDC fan, part number CF-215 @ $7.50 A tube of thermal grease, part number TG-20 @ $4.25. You will need a power connector too. You can get one at Radio Shack for a couple of dollars. Total cost is under $50. Parts can be obtained from All Electronics at http://www.allelectronics.com/ Modifications to the DSI Pro require opening the camera and drilling two holes in the back of the case for 6-32 screws (for added stability 4 screws could be used) to hold the TEC/heat sink/fan assembly to the back of the DSI case. Note: This will void the camera's warranty. Use a 2.5 mm Allen wrench and open the camera by removing the 4 corner hex-screws. Be careful of the gasket (see Figure 4). Place the center of the TEC so it is directly opposite the cold finger. Drill and tap 2 (or 4) matching holes in the heat sink. Make sure the hole spacing is sufficient to clear the TEC. Use two nylon screws to hold the heat sink/TEC to the back of the camera. The nylon screws will prevent heat transfer from the hot heat sink back to the case. Figure 4 Inside the DSI Pro with Cold Finger shown Add some foam for insulation around the camera to keep it cool. See Figure 5 for a view of the modified camera with foam insulation added. Figure 5 Modified DSI Pro with TEC/Heat Sink/Fan/Filter Wheel, Focal Reducer and Foam The polarity of the TEC is a bit ambiguous. After the doing the wiring check and see which side gets hot and place that side on the heat sink. Don't forget ample thermal grease. Figure 6 shows a drawing of the modification and an electrical schematic. Figure 6 Modification and Electrical Drawing The details of photometry are beyond the scope of this article. More information on photometry can be found at http://www.hposoft.com/Astro/PEP.html and http://www.aavso.org/observing/programs/ccd/. The following discussion is meant to give those who are not familiar with astronomical photometry an overview of it. Astronomical photometry is the precise measurement of the brightness of an astronomical object. While the objects can be most anything, stars are the most common object measured. Astronomical photometry is an area where amateur astronomers with modest equipment can make significant contributions from their backyard in a light polluted suburban setting. The magnitude of an object is simply Raw Instrumental Magnitude = -2.5 log10 (Intensity or flux) The intensity or flux is some measure of the brightness of the object. This can be an analog voltage, current level, a digital readout or even a line on an old strip chart recorder. This magnitude is called the raw instrumental magnitude. It does not represent the true magnitude until some corrections are made. Consider a given star observed with 8" and 14" telescopes. Since the 14" telescope gathers considerably more light than the 8", the intensity and thus the raw instrumental magnitude will be much greater than that produced by the 8" telescope. To account for this, a zero point factor must be used for each system to normalize the magnitude. Another problem has to do with the spectrum of the object. Consider a given star observed with an 8" telescope through 5 different filters (UBVRI). The results will usually be 5 different magnitudes for the same star. If no filter is used, even another magnitude will result. To be useful these and other factors must be taken into account when calculating the standard magnitude of an object. Figure 7 show a comparison of stars through B and I filters. Both images were taken near the same time using a 30 second exposure with a 12" telescope. Figure 7 30 Second CCD Images of EE Cephei with B and I Filters The most accurate type of astronomical photometry is differential photometry. One or more stars of known magnitude and known not to vary are used to compare the star of interest (program star). This provides much more accurate data. Many years before the CCD camera became popular photometry was done mainly with single channel devices. This meant observing one star at a time. For detectors a photomultiplier tube (PMT) or pin diode was used. Many of these devices are still producing good results. The CCD, PMT and photodiode all work on the photoelectric principle where a photon knocks off an electron. While some people differentiate the single channel and CCD photometry as PEP and CCD, they are all PEP (photoelectric photometry). It is a matter of single verse multi-channel photometry where CCD is multi-channel. The PMT and pin diode produce a voltage or current proportional to the light intensity falling on the photocathode of the PMT or active area of the pin diode. A PMT system can also be configured in a photon counting mode where photons hitting the photocathode produce pulses which are then amplified and counted. Each detected photon produces one count. The number of counts per second is proportional to the intensity of the light. Big advantages of the PMT system are a very large dynamic range (well over several million) and sensitivity in the U band. PMTs usually work best in the UBV bands. The pin diode while not normally sensitive in the U band, extends detection to the longer wavelengths into the R and I bands and with a special pin diode, into the J an H bands. The only commercially produced photometers are those from Optec (http://www.optecinc.com/). The SSP-3 pin diode photometer covers the BVRI bands and the SSP-4 has a special TEC cooled diode which is sensitive in the J and H far infrared bands. Optec also offers a PMT based photometer (SSP-5) that allows UBV work. In fact it is not too difficult to make your own PMT based system. CCD photometry is multi-channel photometry. Each pixel of the CCD acts as a detector similar to a pin diode. Most CCD cameras available to amateurs are sensitive in the BVRI bands. The big advantage of the CCDs is the image taken can contain multiple comparison and program stars and the sky reading. It also does not require precise pin pointing the star for the reading. This is important for faint stars in a bright sky area. In the Phoenix area I can easily image stars 12th magnitude or fainter, yet cannot see those stars visually through the main optics of my 12" telescope. Single channel photometry on these stars would not be possible as I would not be able to find the stars. CCDs have another advantage over the singe channel systems (particularly the PMT versions). The Quantum Efficiency (QE) of the CCD is double that of a typical PMT. For CCDs the QE is around 60% while with the PMTs the QE is around 30%. QE is simply the percent of photons hitting the detector that get detected. If 100 photons hit a pixel of a CCD, 60 of those will knock off 60 electrons. With the PMT out of 100 photons hitting the photocathode 30 will knock off 30 electrons and produce 30 pulses or a corresponding current. While the lower QE of a PMT may not seem good, it actually works fine because the QE is very constant. Once a value for raw instrumental magnitude is determined, the procedure for either single channel or CCD photometry is the same. Corrections for the system constants, e.g., zero point and color corrections) must be made. A meaningful magnitude can then be produced. In many respects single-channel photometry is a lot easier than CCD photometry. With CCD photometry you must use a computer, dark frames must be taken and subtracted, flat frames taken and divided. Then a sum of pixels containing the object must be calculated with a sky reading subtracted. Most of this can be done automatically with software, however. Experimentation must be done to determine the correct exposures for each filter for a given program. This can get complex with multiple stars of different colors and magnitudes in one image. For fainter stars and where a comparison star can be found in the same image as the program star, CCD photometry rules. The AutoStar documentation does not explain photometry well. There are Gain and Offset slider bars. Default is for the Gain= 100 and Offset= 50. Keep these at the default values. The Maximum count for pixels is linear up to the maximum allowed for the 16 bit system, 65,535. If the Gain or Offset are changed the maximum count must be reduced to stay linear. With the addition of photometric filters and a few simple modifications the DSI Pro and DSI Pro II can be used for serious astronomical photometry. Beware, photometry can be addictive, but very rewarding. About the author He is located in Phoenix, Arizona and has been doing photometry since the early 1980's. He is currently working on the mysterious epsilon Aurigae star system in anticipation of the 2009 eclipse. He uses a C-8 with a home-built UBV photon counting system and a 12" LX200GPS for J and H band photometry with an Optec SSP-4 and BVRI CCD photometry with a modified DSI Pro.
	Return Created 25 April 2006 Modified 10 February 2007