Thermal Considerations

If someone were to tell you that having an electric fan on a telescope was, perhaps, the most important decision you would make in your observing career, you would probably brand them as insane! But, as the years have gone by (and I have been observing planets for over 30 years) I have become increasingly convinced that the ability of a telescope to cool to the night air is absolutely crucial to getting good results. Back in the 1970s, the French optician Jean Texereau claimed that a temperature difference of 1/7th degree Celsius in a telescope tube could perceptibly alter the performance of the instrument. This would not surprise me. In terms of optics cooling to the night air there are two prime considerations. Firstly, the turbulence created inside the telescope tube, especially near to the mirror; secondly, the distortion of the optical components, as they cool down. Prior to the mid-20th century, most amateur telescope mirrors were made from plate glass with considerable temperature expansion coefficients. It was common for telescope-makers in those days to undercorrect their parabolic Newtonian mirrors slightly (correcting in this context means correcting from a spherical to a parabolic surface) such that the cooling of the plate glass at night would cause contraction to "snap" the mirror into the perfect shape early in the observing session. However, since the introduction of Pyrex mirror blanks, the contraction of telescope mirrors has not been a major problem, unless you are still using a historic instrument with a plate glass mirror. In recent years the (admittedly expensive) option of the glass called Zerodur has meant that amateur telescope mirrors with a zero coefficient of expansion are available. So, when we are talking about the detrimental thermal effects of a warm telescope mirror in the 21st century, we are talking solely about the turbulent air problem and not the shape of the mirror.

In 2004 I acquired an excellent 250mm f/6.3 Newtonian reflector from the U.K. manufacturer Orion Optics. It featured a relatively thin (by historic standards), 8:1 ratio primary mirror mounted in a lightweight metal tube. It also featured a cooling fan built into the primary mirror cell (Figure 3.10). The whole optical tube assembly only weighed 11 kilograms. I was astounded (and still am) at the performance of that telescope. I had previously owned 36- and 49-cm reflectors with good optics as well as 30- and 35-cm SCTs and yet this relatively modest aperture Newtonian gave me sharper planetary views than any of those larger telescopes. As far as I can judge, there are three reasons for the lunar and planetary performance of this telescope. Firstly, the optical quality; the primary mirror has a 1/45th wave RMS surface, verified by a Zygo interferometer (I have more to say on this subject in a few pages' time). It also has a Strehl ratio of 0.981. (A Strehl of 1.0 defines a perfect unobstructed primary where 84% of the light from a star goes into the Airy

Figure 3.10. The cooling fan on the rear of the author's 250-mm f/6.3 Orion Optics Newtonian. The mirror is just over three centimeters thick and cools to within a fraction of a degree of the night air in under one hour of fan cooling. Image: Martin Mobberley.

disc and 16% into the rings; 0.981 is very good.) Secondly, the aperture is ideally suited to exploit the best resolution achievable from the U.K. I never feel I need to stop the aperture down and I rarely get multiple images, even in poor seeing. These were common in the 36- and 49-cm reflectors. Thirdly, the telescope has a low thermal mass. I feel that this third point is crucial and, possibly, as important as the other two, because the seeing has always seemed pretty good through this telescope. This indicates that a lot of my seeing problems with larger telescopes were instrumental, not atmospheric.

Telescopes that are larger than about 25 cm in aperture tend to suffer from serious cool-down problems, especially in the evening, when nighttime air temperatures can drop rapidly. A mirror that is more than about 30 mm thick will have serious difficulty adapting to the temperature of the night air, which can drop by several degrees Celsius per hour after sunset. The mass of a telescope mirror increases with the cube of its diameter, assuming its thickness is maintained at the same value as aperture increases. However, the surface area of such mirrors, vital for radiating the heat away, only increases with the square of the diameter. Thus, unless the thickness of a telescope mirror is fixed at, say, 30 mm (necessitating a sophisticated mirror support system for large, thin mirrors, to stop them bending), serious thermal problems will arise in mirrors of 30-cm aperture and above. This problem was investigated in detail in the 1960s by the British mirror-maker Jim Hysom. He found that the mirror core temperature for 30-, 45-, and 76-mm-thick mirrors cooled at typical rates of 3.3°, 1.6°, and 0.9°C per hour after the onset of night. Bearing in mind the air temperature can fall at more than 3°C per hour in the early evening, it can be seen that mirrors more than 30 mm thick cannot even match the falling air temperature with both front and rear faces exposed, unless fan cooling is employed. One solution to this problem is to use conical telescope mirrors, such as used in SCTs. In such mirrors, weighing about 60% of a conventional mirror, the mirror is made from a honeycombed cantilevered structure with the mirror edges thinner than the center. Mirrors up to 40 cm or so in aperture can be supported at the center only with this design and cooling is typically twice as rapid as in a conventional mirror. The renowned U.S. optician William Royce offers mirrors of this type in a Newtonian format.

The most common solution to the mirror and tube current thermal problem is to use lightweight vibrationless fans. In my own 250-mm Newtonian, this fan sits behind the primary mirror. However, a slightly better solution is to employ additional filtered fans in the side of the telescope tube, blowing air across and sucking air out of the region just in front of the primary mirror, where the worst turbulence occurs. Newtonian telescopes are renowned for having tube currents. But if the telescope tube and optics are all at the same temperature these currents tend to disappear. There are two schools of thought on what to do with the problems arising from the telescope tube itself cooling down. One approach (as with my 250-mm instrument) is to have a very thin-walled metal tube (1 mm in my case) that cools to the night temperature rapidly. Another approach is to have the tube manufactured from materials with insulating properties so that they only radiate heat slowly. Historically, mahogany tubes lined internally with cork have been advocated and self-adhesive cork tiles are available from good DIY stores. Regardless of this, a good approach is to have the telescope tube of much larger diameter than the primary mirror and to have a gap in the center of a long tube, to let air in. Completely open tubes are excellent thermally, but terrible from the perspective of dew, dust, and insect ingress. They also allow the observer's body heat to drift across the light path. My ultrathin-walled metal tube seems to be a good compromise and is light, too. I have thermocouples on the tube of my telescope and on the top edge of the primary mirror (Figure 3.11). After some 30 minutes of fan cooling the telescope is at equilibrium and the lunar and planetary views can be exquisite. It took me some 30 years to appreciate how important telescope cool-down was and why mirrors larger than 250 mm have a real thermal problem. Fortunately, the Earth's atmosphere rarely allows resolutions finer than 0.5 arc-seconds to be achieved anyway, so a 250-mm telescope is pretty much all you need in the webcam era. In passing, I would like to add that the renowned Florida planetary observer Maurizio Di Sciullo recently revealed to me that he has to use a 20-cm fan on his 250-mm (40-mm-thick) Newtonian mirror to cool it to ambient temperature. He has concluded that the only way he can ever cool his new 360-mm (60-mm-thick) planetary telescope's mirror to ambient temperature, within the duration of a night, is by water-cooling the primary; fan-cooling is not sufficient! Mirror cooling is a very serious issue.

Different telescopes have different thermal properties. Newtonian telescopes can behave very badly if the tube is sealed at the mirror end. An open mirror cell design, with the glass mirror back open to the air is essential. On my Newtonian, capping the mirror end while examining a star at high power, and then uncapping

Figure 3.11. A cheap indoor/outdoor thermometer attached to the telescope can be used to measure the mirror temperature (via a lead) as well as the tube or air temperature. These devices are so cheap and light that several can be employed all over the telescope. Image: Martin Mobberley.

it, is dramatic! As soon as a free air flow through the tube end is available the star's diffraction pattern becomes near-perfect on good nights. With the mirror end capped, stars distort into a tear-shaped drop, due to air currents seeking a way to escape. Large Schmidt-Cassegrain and Maksutov telescopes (above 250 mm in aperture) can store considerable amounts of heat in their glass components and in the trapped air inside the instrument. To alleviate this issue, one manufacturer has even designed a probe that can be inserted into the eyepiece barrel of compound instruments to replace the trapped air with cooler air from the outside world. However, having studied the webcam AVI videos of Damian Peach and Dave Tyler, obtained with 235- and 280-mm Celestron Schmidt-Cassegrains, it does appear that the sealed tube properties of SCTs do offer refractor-like closed tube advantages, with conventional Newtonian-type tube currents being virtually eliminated (as they are in a Newtonian at equilibrium). Maksutovs can be a different kettle of fish, though, and instruments over 20 cm aperture can have chronic cool-down problems as the curved corrector lens in a 25-cm instrument can weigh as much as the primary and the whole instrument can become a giant thermos flask. Maksutovs over 20 cm aperture without serious fan-cooling are best avoided. In 2005 Damian Peach acquired such an instrument (at great cost) and concluded it was totally unsuitable for planetary work. The optics were good, but the instrument stored heat like a bread oven. A Celestron 9.25 SCT, at an eighth of the price, proved to be a far superior planetary instrument as it cooled down much more rapidly!

Thermal problems are not just restricted to the telescope either. Thermally unfriendly observatories exist in abundance. Typically, these are brick-walled observatory domes with a massive concrete base, a narrow dome slit, and no ventilation system. Such structures can make observing the planets at high resolution virtually impossible after a sunny day. Planetary telescopes are far better if they are in the open air, so a run-off shed, run-off roof, or tarpaulin cover shelter are far preferable to a huge domed building. Finally, are you contributing to the thermal problems? Human beings give off a lot of body heat, so, when acquiring those webcam frames, stay well away from the telescope aperture.

Although the rapid download speed of webcams makes planetary focusing much easier than with single-shot cameras, it is still a battle to focus a planetary image, especially when the planet is rippling and distorting at the whim of the atmosphere. In typical seeing conditions the planet looks like it is constantly being focused and de-focused. So how do you decide where the true focus point is? The best answer to this question will not sound very helpful; it is simply "try your hardest"! There are, admittedly, techniques used for focusing star fields that might be considered if you are desperate. The two most popular methods are the so-called Hartmann mask and the diffraction spike technique. In the former method, the telescope aperture is stopped down by two or more smaller apertures. For example, a 30-cm telescope might be covered by a 30-cm cardboard mask with two 7-cm diameter holes at the aperture edges. This will produce, in effect, two 7-cm telescope images. When the telescope is badly out of focus, two slightly overlapping images of the planet will appear. When perfect focus is achieved, the two images will merge. Unfortunately, while this method works fairly well for stars in low-resolution work, for planets we want the highest resolution we can get. The telescope resolution will be seriously hampered by the Hartmann mask, as, even when focused, the mask will cause horrendous diffraction effects. Worse still, planets at the sort of f-ratios used for imaging, especially faint planets like Saturn, will appear very ghostly when imaged through such a mask. The diffraction spike technique also only works well with bright pinpoint stars. Using this method, the secondary mirror support vanes of a Newtonian are used as an indicator of how sharp the focus is: a very bright star is imaged and the sharpness of the star's diffraction spikes are used to assess focus. (Artificial vanes can be placed over the aperture for non-Newtonians.) Unfortunately, this technique really does only work for long exposures with deep sky objects.

So, is there a solution to planetary focusing? Well, as a priority, a webcam planetary imager must have a motorized focuser. Even with the sturdiest telescope mountings the slightest touch of the astronomers hand on the focuser will shake the telescope. The best motorized focusers for amateur astronomy are made by JMI (Jim's Mobile Industries) and every top planetary imager I know uses one (Figure 3.12). Once a motorized focuser is in place, you can sit in a comfortable position at your PC screen, confident that the planet's oscillations are due to the atmosphere (and the telescope drive) and not your hand shaking the focuser. It goes without saying that the smoother the telescope drive, the better.

Figure 3.12. One of JMI's excellent motorized focusers. This one is an NGF DX designed for Newtonians. Focusing "shake" is eliminated with this design. Image: Martin Mobberley.

The globes of most planets appear as depressingly featureless fuzzy balls in the raw video stream from a webcam. Even Jupiter only shows two obvious features, namely the north and south equatorial belts. This can be quite demoralizing to the beginner, but do not worry, because raw images always look like this. However, with such ghostly features, what on Earth can you focus on? Mars and the Moon are really the only planetary bodies with sharp enough features to easily focus on. The Moon is a doddle, especially near the day/night terminator where contrast is high. With Jupiter, I use the Jovian moons as a focus reference. Yes, I know these are tiny discs and not point source objects, however, they are far better than anything else. On a typical night, with the webcam gain set high enough, Jupiter's moons will be easily recorded at 10 or 15 frames per second with a 25-cm reflector. However, the moons, even when focused, will rarely appear as small discs. They will appear as eggshapes, blobs, spiders, and even stubby lines as the atmosphere wreaks havoc on the incoming light. The best one can do is to spend a good 10 minutes, just moving in and out of the focus point and developing an educated "gut feeling" for where the best focus position is. While this may seem rather unscientific, well, to be honest, it is! However, planetary imaging is a bit of a black art. JMI moto-focusers have the option of an extra feature known as digital position readout. Typically, this tells you where the focuser is within 0.01 mm. This can be a useful aid when trying to remember when the planet looked sharpest. Getting back to the subject of Jupiter's moons: is there always one within a reasonable distance of Jupiter, I hear you cry. The answer is, invariably, yes. By creeping Jupiter east or west until it just leaves the webcam PC window you will almost always spot a Jovian Moon if the webcam gain is high enough. The closest Galilean Moon Io never strays more than 3.5 arc-minutes from Jupiter and with three other bright moons to consider you will never be short of focus targets. Of course, when seeing really is excellent you will be able to focus the Jovian moons into tiny discs. But this might only occur a few times per year. Fortunately, Saturn has one excellent feature to focus on: the rings. Specifically, the gap between the A and B rings, the Cassini division, is a focusing gift from heaven. Focusing is not something that can be rushed. After a while you develop a second sense of when something is focused "as well as it can be" and, at that point, you can start saving your webcam video to the PC hard disk.

Another crucial point here is the following one: does your motorized focuser actually allow you to increment the focus point in small enough bursts to exploit the very best seeing conditions. I have already mentioned that digital readout can record the focuser's position to within 0.01 mm, but it is often quite tricky to actually give a short enough "jab" on the focuser key pad to move it this small a distance. The increments needed for focusing a planetary image accurately are microscopic when you are working at Newtonian f-ratios. With an f/10 Schmidt-Cassegrain, a short jab on a motorized focuser fitted to the back of the telescope will just be precise enough. But with an f/5 Newtonian it will not, one jab will take you from inside the optimum focus position to well outside. A 25-cm f/5 Newtonian will, in theory, resolve about half an arc-second. This equates to 3 microns at the Newtonian focus. (A micron is 1/1000th of a millimeter.) This, in turn, equates to 15 microns of focusing tolerance at the same focus, or 0.015 mm. It makes no difference whether a Barlow lens follows the focusing point, unless you can arrange for the focuser to be after the Barlow lens. Even the best motorized focusers struggle to position the focuser to this accuracy and, even if they can, you literally need to jab the buttons for an imperceptible period. In my case, I contacted the focuser manufacturers and they told me which resistor I could change to make the focusing increments smaller. The only practical alternative would be to spend even more time jabbing back and forth across the focus point until you are happy, or switch to a telescope with a much longer nominal f-ratio. Schmidt-Cassegrains have a distinct advantage here if you add a motorized focuser to the drawtube where the light cone is a gentle f/10.

There is one other point I would like to mention in the context of focusing, and it adds an extra sledgehammer-type weapon to the planetary imager's arsenal. Essentially you need a lot of hard disk space! If you cannot determine exactly where the focus point is on the raw image, your chances of hitting it are increased if you take as many webcam videos as possible. Between each imaging run you should have another stab at focusing, because 1) you may get lucky; 2) the telescope tube may have contracted as the temperature dropped; and 3) the seeing might have improved, making focusing easier. When you can check the recorded runs indoors, at leisure, you may well conclude that one of them is much better than the rest. When images from this best run are stacked and processed, a superb image may result. However, bear in mind that you do need a large hard disk. Planetary imaging runs can be 1 or 2 Gigabytes in size. If you take a dozen during the night that is a lot of hard disk space.

Above and beyond all these high-resolution considerations are the issues of image processing wizardry. I will have much more to say about these in Chapters 8 and 9.

Before we leave the issue of high-resolution considerations it is time we indulged in the never-ending debate as to what constitutes the best high-resolution instrument.

Telescopes Mastery

Telescopes Mastery

Through this ebook, you are going to learn what you will need to know all about the telescopes that can provide a fun and rewarding hobby for you and your family!

Get My Free Ebook

Post a comment