Astronomy › Section F · Astronomical instruments › Topic F.3
Visible-light detectors and instruments
Collecting light is only half the job: it must also be sorted and recorded. Electronic detectors that catch most of the photons reaching them, cooled infrared cameras and spectrometers turn a telescope's light into precise, permanent data.
Beyond the human eye
The topic Telescopes described the third part of every observing system: a detector that senses light and keeps a record. The first detector was the human eye, and it has two big weaknesses. Its record lives in human memory and drawings, which are unreliable and hard to measure. It also has a very short integration time: it adds up light for only about a tenth of a second before sending a signal to the brain. A detector that can keep collecting light for minutes or hours, a long exposure, can reveal objects far too faint for any eye, even at the same telescope.
Filters and spectrometers
Before reaching the detector, light usually passes through an instrument that sorts it by wavelength. The simplest is a filter, a piece of coloured glass or coated material that lets through only a chosen band of wavelengths, just as red plastic passes red light and blocks the rest. Images taken through several filters let astronomers measure an object's brightness and colour, and so, using Wien's law, estimate a star's temperature. For finer sorting, a spectrometer spreads the light into a full spectrum so that individual spectral lines can be measured. Either way, a detector must still record the result.
Photographic plates
Through most of the twentieth century, astronomers recorded images and spectra on photographic plates: glass coated with a light-sensitive chemical layer that darkens where light falls and gives a lasting record when developed. Observatories still keep vast archives of plates that show what the sky looked like over about a century, which is valuable for spotting changes. But photography is very inefficient. Only about 1% of the photons striking a plate cause the chemical change that forms the image; the rest are wasted. Its response is also uneven, so turning a darkened patch into an accurate brightness is difficult.
Charge-coupled devices
Today most visible-light images are recorded by charge-coupled devices (CCDs) and similar electronic sensors, relatives of the chips in phone cameras. The CCD was invented in 1969 by Willard Boyle and George Smith, who shared the 2009 Nobel Prize in Physics for it. A CCD is a grid of tiny light-sensitive picture elements, or pixels. A photon striking a pixel frees electrons, which collect in that pixel until the exposure ends and are then read out and counted. Modern astronomical cameras hold many millions of pixels; the camera of the Vera C. Rubin Observatory has about 3.2 billion.
Why CCDs won
A typical CCD records 60% to 70% of the arriving photons, and the best exceed 90%, compared with about 1% for photographic plates. That efficiency means fainter objects can be detected in the same time: tiny moons circling Uranus and Neptune, small icy worlds in the Kuiper belt and dim dwarf galaxies. A CCD's signal is also proportional to the light received, so brightness can be measured accurately. And its output is digital, a set of numbers that go straight into a computer for analysis, combining and archiving. Professional astronomers rarely look through a large telescope at all.
Worked example: plate against CCD
An old photographic plate needed a 2.0-hour exposure to record a faint galaxy. How long would a CCD that records 70% of incoming photons need to collect the same signal, if the plate recorded 1%? 1. Efficiency ratio: \(\frac{70\%}{1\%} = 70\). 2. The CCD gathers the same recorded signal 70 times faster, so divide the time: \(\frac{2.0\ \text{h}}{70}\). 3. Convert: 2.0 h is 7200 s, and \(\frac{7200}{70} = 103\) s, under 2 minutes. 4. Check the logic the other way: in a full 2-hour exposure, the CCD would collect 70 times the signal, enough to reach objects much fainter than the plate could.
The infrared problem
Infrared astronomy runs from about 1 micrometre, near the long-wavelength limit of ordinary silicon CCDs, out to about 100 micrometres and beyond, so it needs special detectors. Its biggest challenge is that everything around the telescope glows in the infrared. By Wien's law, from The electromagnetic spectrum, objects at about 300 K, including the telescope, the dome and the air itself, give off infrared radiation peaking near 10 micrometres. Detecting a faint galaxy against that glow is like trying to see a dim star in daylight from inside a room lined with lamps. Infrared detectors are therefore cooled, often to a few kelvin with liquid helium, and the telescope's own heat is screened from them.
Worked example: why cold helps
Compare the thermal glow of a detector at room temperature, 300 K, with one cooled to 4 K. 1. Peak wavelength at 300 K: \(\lambda_{\max} = \frac{2.90 \times 10^{-3}}{300} = 9.67 \times 10^{-6}\) m, about 10 micrometres, right in the infrared band being studied. 2. Peak at 4 K: \(\frac{2.90 \times 10^{-3}}{4} = 7.25 \times 10^{-4}\) m, about 0.7 mm, far beyond it. 3. Power per square metre by the Stefan-Boltzmann law: the ratio is \(\left(\frac{300}{4}\right)^4 = 3.16 \times 10^7\). 4. Conclusion: cooling makes the detector's own glow about 30 million times weaker and shifts it out of the band, so faint cosmic infrared stands out.
Inside a spectrometer
More than half the observing time on most large telescopes goes to spectroscopy. In a simple spectrometer, the telescope's image falls on a narrow slit. A lens makes the light from the slit into a parallel beam, a step called collimating. A prism, or more often a grating, a surface ruled with thousands of fine grooves, then sends each wavelength off in a slightly different direction. A second lens focuses these images of the slit side by side onto a CCD, forming the spectrum. There is a cost: spreading light into many narrow wavelength bins leaves fewer photons in each, so spectra need bigger telescopes, longer exposures or both.
Common mistakes to avoid
First, thinking modern astronomers look through big telescopes; detectors record the light. Second, assuming photography was nearly as efficient as CCDs; plates used about 1% of the light, CCDs 60% to over 90%. Third, believing infrared detectors are cooled to protect them from damage; they are cooled so their own heat glow does not swamp faint sources. Fourth, expecting a spectrum to need no more light than an image; spreading the light thins it out. Fifth, confusing a filter, which passes one band, with a spectrometer, which spreads out every wavelength.
Key takeaways
The eye integrates light for only about a tenth of a second and keeps no reliable record. Photographic plates gave permanent records but used only about 1% of the light. CCDs, invented in 1969, count photons in millions of pixels with 60% to over 90% efficiency and give accurate digital data. Filters pass chosen bands for brightness and colour; spectrometers use a slit, a collimator and a prism or grating to spread light into a spectrum. Infrared detectors must be cooled, often to a few kelvin, because everything near 300 K glows at about 10 micrometres.
Flashcards
Flashcard
A device that senses radiation and makes a permanent record
Test your understanding
Question 1 of 5
What is the main advantage of a CCD over a photographic plate?
Question 2 of 5
A detector that records 2% of incoming photons needs a 4.0-hour exposure for a faint nebula. How long would a detector recording 80% need for the same signal?
Question 3 of 5
Why are infrared detectors cooled to just a few kelvin?
Question 4 of 5
In a spectrometer, what does the grating do?
Question 5 of 5
Why does recording a star's spectrum usually take longer than taking an image of it?
