Detector Systems

Jupsat Pro Astronomy Software

Secrets of the Deep Sky

Get Instant Access

Given the previous top-level summary of instrumentation sampled in this survey, we will now shift to the detector systems used in these instruments. This includes both the detectors and advanced array controllers needed to readout signals recorded by contemporary detectors.

3.5.1 Array Controllers

Given the common need to readout array detectors with a controller, the range of controllers built in astronomy over the past few decades is remarkable. Manufacturers historically do not market controllers with the detectors they supply, leaving it to the customer to find or develop in-house their own controllers. This has led over the past few decades to a myriad of controllers in astronomy, based upon a wide range of evolving technologies. The previous decade saw controllers based on transputer or DSP technology that required fairly unique skills to program and maintain, occupying considerable amounts of rack space to generate a 16 or 32 channel system. More recently controllers based upon fast PCI processors and broadband fiber links are becoming available, which are cheaper (since they are based on PC technology), require much less space, dissipate far less heat, and are easier to program. The future points to further miniaturization as ASIC technology slowly but surely emerges with the promise of reducing controllers to a minimum of small components.

As shown in Figure 6, no single vendor has attempted to meet all the needs of astronomy. The readout requirements of such different detector architectures as mid-infrared IBCs and CCDs are obviously different (speed, clocking patterns, noise performance, etc.). However having a small/unified set of array controllers that are affordable, broadly supported, and readily available would no doubt be embraced by many observatories strapped with the complications of maintaining nearly as many types of array controllers as they have instruments. That said the bulk of the market share is currently occupied by controllers made by San Diego State University (Bob Leach). The next most common controllers in use today are those made by ESO (FIERA for the optical and IRACE for the infrared detectors). The ~30% share in the "Others" category testifies the large number of custom made systems, each of them present in quantities too small to plot separately.

Figure 6. Among the "brand names", the San Diego State University by Bob Leach is the most popular controllers.

3.5.2 Number of Pixels per Focal Plane

Some of the most interesting results of the survey are encapsulated in Figures 7 and 8. These histograms show the number of pixels per focal plane at optical and infrared (near plus mid-infrared) wavelengths for both current and future instruments. Adding the pixels in Figure 7 leads to a total of ~2.2 Giga-pixels in use today in astronomy. As shown in Figure 7, most optical and infrared systems use focal planes in the 1-10 Mpixel range and the vast majority of current NIR focal planes are well below 10 Mpixel in size. Focal planes well beyond that size are only populated by optical sensors (CCDs), again no doubt due to cost and complexity constraints.

Contemporary distribution of optical and infrared pixels across all of the instruments sampled.

Figure 7.

Contemporary distribution of optical and infrared pixels across all of the instruments sampled.

Looking forward, Figure 8 shows the future growth of large CCD mosaics in astronomy. The science drivers behind this are explained in Sec.

4. Of particular note is the peak in the "more" category, which reflects the upsurge in development of Giga-pixel class CCD focal planes intended for ultra-wide field surveys in the near future. In the same vein, notably absent are comparably large infrared focal planes which regrettably remain beyond the reach of astronomers. Amazingly, the future instruments sampled here include a staggering ~7*109 pixels in planned instrumentation, ~90% of which is in the form of CCDs in the "more" category. This means a nearly four-fold increase in detector "real estate" will be realized in the next decade compared to what is now available to astronomers using ground based instrumentation.

Number of pixels in Focal Plane [106]

Figure 8. Distributions of optical and infrared pixels for future/planned instruments.

3.5.3 Cost per Pixel

The data provided in this area is relatively scant given the number of instruments sampled. Nonetheless, results are consistent with what has been anecdotally recognized for a long time - the cost per infrared pixel is nearly an order of magnitude higher than the cost per CCD pixel (typically ~1-30/pixel). The situation is even worse when comparing CCDs with mid-infrared detectors, which have far fewer pixels yet retain a comparable cost to near-infrared arrays. While this is no doubt attributable to the much higher complexity associated with infrared arrays, the cost "challenge" of infrared arrays is an inhibiting factor in astronomical research. As mentioned before, this is certainly at least partially responsible for the fewer number of infrared instruments available to astronomers today. This in turn limits the scientific research capacity of astronomers, particularly those without access to large modern facilities. Breaking the cost curve for infrared arrays remains one of the key challenges in the future development of this technology.

Without a major cost reduction, it will probably prove impossible to finance the Giga-pixel class focal planes that are being designed with CCDs today. Table 2 lists average instrument costs provided through the survey. It is notoriously difficult to make "apples-to-apples" comparisons in instrument costs, given the different methods used to account for labor and various indirect costs. With this caveat in mind, the principal conclusion from Table 2 is that next-generation optical instruments will be much more expensive due to the use of much larger CCD focal planes. While infrared instruments will become more expensive, with a cost increase of ~25%, the ~16% cost increase foreseen in optical instruments is not foreseen in the infrared. Again this is due to the large cost per pixel for infrared detectors, which effectively prohibits development of enormous infrared focal planes. Another possible effect seen here could be "market saturation", i.e., it is simply very difficult to raise large amounts of money for a single instrument.

Table 2. The current and future median instrument costs derived from the survey are listed.

Optical

Infrared

Current

$400,000

$3,750,000

Future

$6,600,000

$5,000,000

3.5.4 Typical Mosaic Building Blocks

The unit detector or building block used to assemble large scale mosaics varies radically from optical to mid-infrared detectors. As shown in Figure 9, the most currently popular CCD format has dimensions of 2048 x 4096, and these are typically buttable on 2 or 3 sides.

Optical Detector Format

Figure 9. The most popular current and future CCD building blocks used in astronomy.

Optical Detector Format

Figure 9. The most popular current and future CCD building blocks used in astronomy.

In some cases optical instruments have incorporated 4096 x 4096 monolithic detectors. In contrast, only one of the various near-infrared instruments sampled incorporated a 256 x 256 detector. The majority of near-infrared instruments in use today rely upon 1024 x 1024 arrays, with a significant fraction using the more recent generation 2048 x 2048 buttable devices just now becoming available. Finally, the handful of mid-infrared instruments in the survey principally used 240 x 320 format devices, with only 1 using an older 128 x 128 detector. None of these devices are readily buttable.

From the instrument builder's perspective, while the availability of large format single detectors has and will continue to be important, what is arguably just as important is the possibility of making these devices buttable. What has long been possible with CCDs has only recently become possible in the near-infrared. This single advancement in technology makes it possible to build truly enormous focal planes without relying upon the development of ever larger monolithic structures. From a pure marketing perspective, this area is one in which infrared vendors have been slow to recognize the sales potential of providing not just bigger detectors, but detectors that the customer can combine to make instruments previously impossible to build. From a science perspective, this has enormous implications, as it facilitates the development of instruments such as: wide field imagers, cross-dispersed spectrometers covering large wavelength ranges, adaptive optics imagers which rely on large numbers of small pixels to sample an exquisitely fine focal plane, or ELTs that simply need enormous focal planes that scale with the rest of the telescope.

Though the natural tendency may be to develop ever larger infrared arrays, it has already been demonstrated that the ~20482 building blocks offered by CCDs can be used effectively to build enormous mosaics.

Detector Forrrot

Figure 10. The most popular current and future infrared arrays used in astronomy.

Detector Forrrot

Figure 10. The most popular current and future infrared arrays used in astronomy.

While near-infrared instrument builders will no doubt use larger monolithic arrays (as demonstrated Figure 10), the real breakthrough in infrared detector technology from the astronomer's perspective has been the advent of buttable infrared arrays, from which larger mosaics can be built. Nonetheless, for future ELT applications, where the focal plane is intrinsically large and the detector package must expand correspondingly, detector costs must fall radically. If not only a small portion of an ELT's focal plane that cost ~$1B to produce will be adequately sampled with a detector system. Infrared array manufacturers are quick to accentuate that costs will fall if the number of devices built increases. While no doubt true, this trade space remains to be explored and will be the subject of considerable dialog as the ambitions of astronomers drive the development of ever larger infrared sensors.

3.5.5 Manufacturers

A unique aspect of the survey is that it allows a fairly comprehensive assessment of market shares for the various manufacturers of astronomical detectors. The situation is fairly simple for the current crop of mid-infrared instruments, where Raytheon dominates the market, though it remains comparatively small and specialized. This field remains challenging to support from an observatory perspective, though scientifically it remains one of the most "target rich" in astronomy. In the near-infrared most instruments operate from 1-2.5 ^m and use Rockwell detectors. In fact over 1/2 of all infrared arrays in use are made by Rockwell. This should also remain true in the near future. The instruments in the 1-5 ^m range typically use Raytheon detectors. With only a few exceptions, these two manufacturers dominate the infrared astronomy market today - a situation that has existed for the better part of 2 decades (see Figure 11).

Figure 11. The market share for various infrared array manufactures, both for current and future instruments.

In the optical, there are more manufacturers today providing high quality CCDs to astronomy. That said, the bulk instruments sampled use e2v detectors, with SITe/Tektronix coming second, closely followed by MIT/LL. Loral, and a handful of other manufacturers follow. e2v is still dominating the future market share according to our survey, though a large fraction of the surveyed future market, as testified by the larger percent of N/A entries for the future, remains uncommitted. This is principally due to the detectors, (over 200), foreseen for the LSST. The vendor has not yet been selected. It is also important to note that this survey only included science detectors, some of which are used for wavefront sensing and none for technical purposes. A large number of these small format high speed CCDs are used in astronomy but for the purposes of this survey were not included. Figure 12 shows the market share for various CCD manufactures, both for current and future instruments.

70.0

70.0

Figure 12. Market share for various CCD manufactures is plotted, both for current and future instruments.

Was this article helpful?

0 0
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