AY 105 Lab Experiment #2: CCD Characteristics andOperationThis week you’ll study the characteristics of Charge Coupled Device (CCD) imag-ing array detectors. These detectors are the most common type used in astronomy foroptical wavelengths. In particular, you will learn a bout CCD quantum efficiency (QE)and the solid state architecture of CCDs in Part A, and in Part B you will investigatethe electronic circuitry that operates a CCD. There are two sub-parts to Part A, andone group can work on the experimental part while the other works on t he computer.You can then join forces with the other gro up to work on the electronics. Part B canbe continued into the next lab meeting, when we will operate the readout electronicsof a CCD detector.A. Insights into CCD Quantum EfficiencyCCD quantum efficiency is usually measured with a setup including a QTH lamp,interference filters or a monochromator, lenses or an int egrating sphere, and a calibratedphotodiode reference detector. Based on last week’s experience with the AY105, youcould probably create such a test setup and use it to measure QE as a function o fwavelength. After several hours spent making CCD exposures and photodiode currentmeasurements, you might derive a graph like one of those shown in Figure 1.Figure 1: Q E vs. λ fo r Palomar CCDs.1Ay 105 Spring 2008 Experiment 2 2In this lab, however, you will be investigating a slightly different and more inter-esting question: why do CCDs have the quantum efficiency variations with wavelengthshown in Figure 1? In particular, take special no te of three characteristics of F ig ure 1(before and during lab, as well as in the discussion section of your report). (1) The peakresponse at wavelengths 5500˚A∼<λ∼<6000˚A. (2) The decrease in QE for wavelengthsλ > 6500˚A. (3) The decrease in QE (and the wide variation from device to device) atwavelengths λ < 5000˚A.Configuration. The instructor will give you a Texas Instruments 800 × 800,three-phase thinned CCD glued between a pair of fused quartz plates. Take a momentto examine the CCD (both front and back), visually and through the low-power micro-scope. Describe some of what you see (and find most interesting) in your lab notebook.Then install the CCD sandwich in t he filter holder, and place it on a rotating stage.Arrange the green HeNe (λ = 5348˚A) laser to shine on the CCD array.The experiment al objective here will be to measure the fractions of incident laserlight that are transmitted and reflected upon striking the CCD array. As mentionedduring lecture, “thinned” CCDs are only about 8 − 10µm thick; the Si back surfacealso appears shiny, so it is not unreasonable to suppose that some photons reflect offof the surface and some photons go right through without being absorbed. Of course,photons in both of these categories stand no chance of being detected. The remainingfraction is thus an upper limit on the quantum efficiency of the CCD:QE(λ) ≤ 1 − R(λ) − T (λ) ,and so yo ur first task will be to determine how closely the actual CCDs in Figure 1approach the ideal case of detecting all photons which are not reflected or transmitted.Method. You’ll make an optical path to collect both the light transmittedstraight t hro ugh the array, plus another optical lens assembly off-axis to reimage thereflected light onto the photodiode. Start with the laser on the left side of the t able,illuminating the CCD in the middle of the bench; you will want to use the iris betweenthe laser and CCD, stopped down, to block extraneous laser light from striking theCCD. Place an optical assembly to collect reflected light back from the CCD at about135 degrees from the laser path. You’ll need to rotate the CCD slightly away from thelaser beam’s normal incidence in order to collect the reflected beam; use the 63mmdiameter, 356 mm EFL achromatic lens to collect and collimate the reflected light oversome non-zero solid angle. (Notice that the 10 µm thick CCD “membrane” is not aprecisely flat surface, so the reflected beam will in general not be very well collimated).Then use the 80mm f/ 2.8 triplet lens to focus the collimated light, forming an imageof the point on the CCD form which the light was reflected. Locating this image andplacing your photodiode at this point will enable you to measure the amount of laserpower reflected by the CCD. D etermine whether or not you are likely to collect all ofthe reflected light by calculating the ratio of the image size (on the photodiode) to theobject size (the reflecting spot on the CCD), then estimate the latter and derive anestimate of the former for comparison with the 1 mm2photodiode area. Record theAy 105 Spring 2008 Experiment 2 3photodiode current generated by the laser light reflecting from the CCD. (Note thatthe laser intensity varies, so monitor the photodiode current for a while and estimatea likely uncertainty in all of your measurements.)Next, slide the CCD sandwich within the filter holder so that the laser strikesonly the clear fused quartz plates, and measure the power reflected by the plates alo ne.Finally, substitute the diagonal mirror’s post for the CCD’s filter holder post and alignthe mirror so that the light strikes the photodiode, and measure the power reflected bythe aluminum mirror. Bare Aluminum has a reflectivity of 89% (could you determinethis yourself, and if so, how accurately?), so from this you can determine the powerin the laser beam incident on the CCD sandwich. (Why is using the mirror a betterapproach than moving the photodiode so that it intercepts the incident laser beamdirectly?) Subtracting from this the fraction reflected by the cover glass a lo ne willenable you to determine the fraction that wa s reflected by the CCD itself. This is thefraction R(λ) in equation (1) at the wavelength of the green laser. Now substitute thered laser for the green laser, and repeat these measurements for R(λ) at the wavelengthof the red laser.Next, use either the 50 mm or 180mm EFL Nikon lens o n-axis behind the CCDto focus the transmitted light, and place the photodiode at the position of the imageof the light spot on the CCD. Do you notice anything out of the ordinary in thetransmitted laser light, using a paper screen between the CCD and the Nikon lens?Can you explain t he origins of what you see? (The effect seen here is not impor t antfor the purposes of this lab, but if it intrigues you, feel free to remove the Nikon lensand measure the separation