INFORMATION FOR ASTRONOMERS USING THE TEK #2, 1024 SQUARE CCD ************************************************************* Updated 24/8/92 J.R.BARTON tek2cautions.txt Updated as follows 24/8/92 Commissioning nights imaging had a mavimum of 2% p-p fringing rather than 3% INTRODUCTORY ADVICE ******************* Since the TEK QE falls rapidly with reducing CCD temperature two operating temperatures are offered. The usual 170K operation provides low dark current but has poorer QE compared with operation at the alternative temperature of 200K. 200K provides a QE increase of 20% in the UV, about 16% in B, 10% in V and R and from 10% to more than 100% at lengthening wavelengths through the I band. However, this comes at the expense of an increased dark current of about 0.1e/pix/sec. A low readout noise of 2.3 e-rms is provided by an XTRASLOW readout rate which may be useful for smaller or binned windows. Electronics gain has been adjusted so that 65535 ADUs in FAST and 35,000 ADUs in NONASTRO handles most of the available well size. NONASTRO with the Tektronix CCD is not compromised by poor horizontal charge transfer (HCTE) and does not the need several hundred e/pix as the Thomson did to overcome this deficiency. NONASTRO has some of the electronic settling times reduced in order to produce the fastest possible readout and uses a minimum format (minimum XTRP, XTRL, PREX and PREY). This may cause more rollover and ramping on bias frames but should not significantly affect high light level useage. User's should take bias frames or ensure that there is sufficient bias accuracy in the overscan to take out these errors if the observations are critical in this regard. The NONASTRO windows that are provided (TEK1K_BIASROWS for 5 bias rows ahead of the image, and TEK1K_BIASCOLS for 22 columns of overscan) are the maximum possible size allowed with the CCD format. Smaller windows and binning options may be used but the format allows only the 22 overscan columns and the 5 prescan rows to be binned. XTRASLOW provides the lowest readout noise by lengthening the DCS integration time to 2 X 160us per pixel (or bin). By using small windows or binning, the readout times are quite practical. If windows are placed clear of the columns with hot pixels these and their trails they produce will be avoided. XTRASLOW requires the preamplifier offset adjustment to be accurately set - check with the electronics staff that this adjustment is satisfactory. BINNING does not increase the RON as with the Thomson. Binning by large factors may make evident an excess readout induced dark current. User's of large binning factors should take bias frames and ensure that the overscan truly represents the bias level in the image area. This is more critical when using the XTRASLOW speed with its increased sensitivity. By doing dark frames equal to the maximum intended integration, users should ensure that there is no stray light falling on the CCD. Dome and Cassegrain cage lights have been seen to affect bias and dark frames. Operation at coude should also be checked to ensure that cableway shutters are fully effective. TEKTRONIX FORMAT **************** The CCD comprises 1028 columns and 1024 rows of 24 micron square pixels. There appears to be no aluminised covered columns or rows so that there is a sharp edge to the image area permitting the bias level to be estimated from either the H and V overscan regions once a sufficient margin is allowed for the effects of HCTE or VCTE (vertical change transfer efficiency). The first and last columns have about a 60% increased light sensitivity. The first row may have 5% extra "sensitivity". It is probably advisable to avoid the first and last columns in any critical observations. Also, they should not be included with others in windows binned along the row direction if they are not to affect the bins into which they fall. The readout register has 48 overhanging pixels between the first column and the readout amplifier. The readout register is cosmic ray sensitive so that it is possible for single row cosmic events to occur. READOUT NOISE, GAIN, LINEARITY, SATURATION AND READOUT TIMES ************************************************************ The following are the "final" best estimates of RON, Gain and Alpha. READOUT GAIN ALPHA SAT'N APPROX NOISE (e/ADU) *E-6 LEVEL RO TIME (e rms) ADU Ke/pix (secs) XTRASLOW 2.3 0.34 negl 65K 22 394 SLOW 3.6 1.36 negl 65K 88 120 NORMAL 4.8 2.74 -0.03 65K 180 75 FAST 7.2 5.5 -0.07 65K 370 52 NONASTRO 11 11 -0.14 35K 400 33 The readout times are calculated for a 1050 X 1024 window. These figures apply to both 170 and 200K operating temperatures. There is no degradation in readout noise with X binning factor (tested for binning factors up to 10 in the row direction). NOTE: The accurate setting of the preamp offset bias is essential when using the XTRASLOW readout speed. Readouts at the XTRASLOW, SLOW, NORMAL and FAST rates are all ADC limited and at 65K ADUs, except for FAST, the CCD is nowhere near saturation. CCD saturation occurs at about 450,000e/pixel. Readouts using NONASTRO are limited to a maximum of about 35K ADU because of the saturation of the electronics amplifiers ahead of the DCS stage and not that of the CCD. QUANTUM EFFICIENCY ****************** Measured in the dewar and include the losses of the dewar window. Care was taken to ensure no change in the measuring setup for the two measurements, ie, the same location on the CCD was illuminated at each temperature. Note that a spot illumination of about 5mm diameter was used, which sampled only a small area on the CCD. The CCD was operated in a diode mode for these tests. Quantum Efficiencies (%) at spot wavelengths (nm) 300 320 340 360 380 400 420 440 460 480 500 550 600 700 800 900 1000 1100 At 200K deg, 14/7/92 6.9 15 26 39 57 64 66 67 67 67 67 69 72 72 60 32 8.9 0.4 At 170K deg, 14/7/92 5.7 12 21 33 48 54 56 57 58 58 59 62 65 66 54 28 6.8 0.2 % improvement in QE at 200K over that at 170K 21 20 20 20 20 19 18 16 15 14 13 12 10 9 11 16 31 120 HORIZONTAL AND VERTICAL CHARGE TRANSFER EFFICIENCY (HCTE and VCTE) ****************************************************************** These are very good with no need for _BESTCTE options as with Thomson CCD. In addition, the NONASTRO speed is not compromised with poor HCTE as with the Thomson. Note that some compromises have been made to the CTE with lower clock levels being applied to reduce a clocking induced excess dark current generated only during the readout process and also to keep the readout times short. TRAPPING SITES ************** No trapping sites have been detected either as dark portions of columns or as trailing charge extending into the vertical overscan of low light level exposures made at either 170K or 200K. FRINGES AT THE TELESCOPE ************************ On the second commissioning night, 6/7/92, imaging at PF and using the I filter, night sky emmission lines generated fringes at a maximum level of 2% peak-to-peak. With the TEK fitted to the RGO(25) and using a 1200 line grating a set of tungsten lamp flat-fields was taken on 30/7/92 with a 100 micron slit at the maximum length slit (imaging 200 columns at the centre of the CCD). From these frames 10 column wide YSTRACTS indicated that - 7100A to 7900A showed about 1% p-p fringing at about 7300A and 2.5% from 7600 to 7900A with some fringes reaching 4% p-p. - frames at 7500 to 8300A and 7900 to 8700A showed worsening fringes. - between 8700 and 9500A fringing was generally from 10 - 16% p-p with a few 20% p-p fringes noted. - the greatest fringing was from 9100 to 9900A with 11 to 23% p-p being recorded. - some fringes were very closely spaced with separations of only 12 pixels noted. On 2/8/92, flat-fields were taken on the RGO spectrograph with a 270 line grating fitted and using a 300um slit with the widest dekker (#50). Preliminary results from a 10pixel YSTRACT indicate less than 1% fringing at 7500A or less, 2.7% maximum fringing from about 7700 to 9000A wherafter it rises rapidly to 9% at 9300A, 17% at 9600 and peaking at about 22% from 9800A to 10000A where the CCD response is rapidly falling off. The fringing is dependant on the position of a particular wavelength on the chip. For example, the fringing around 7900A is 2.7% on the 7300 central wavelength frame and only about 1.5% on the 9300A central wavelength frame. Fringes have been seen in the laboratory as described below under High Light Level Flat Field Characterisitics. HIGH LIGHT LEVEL FLAT FIELD CHARACTERISTICS ******************************************* High light level characteristics and defects are as follows (these are referenced to (1,1) as the first pixel of a 1028 by 1024 image read out of the array using amplifier C): RED LED (660nm) FLAT FIELD. These frames were found to be very flat. For example, 27000 ADU average level had a standard deviation of 400 ADUs, excluding the overscan pixels and the first and last column. A histogram of such a flat-field within the 1028 by 1024 image area had the following profile: - 40% to 80% of average response 20 pixels - 80% to 90% 139 - 90% to 95% 1294 - 95% to 105% 1048645 - 105% to 110% 526 - 110% to 137% 0 - 137% to 150% 2048 - the first and last columns Visible characteristics and defects, which appear to flat-field well, are as follows: 1) There are a total of 24 horizontal lines running across the full width of the image area and these are evenly spaced every 42 or 43 rows in the vertical direction. The response in these lines is down by 0.25% to 0.5% and only one or two rows are involved in each line. The predominant rows with this effect are 21, 64, 107, 149, 192, 235, 277, 320, 363, 406, 448, 491, 534, 572, 620, 662, 705, 747, 790, 833, 875, 917, 960 and 1003. These lines appear to be the result of some manufacturing error. 2) There are six obvious spot defects, one being a prominant region of about 20 pixels diameter, the remainder being much smaller regions. They are as follows: - centred at (871, 167) a bell shaped profile about 20 pixels diameter with the central pixel down 11%. - centred at (424, 213) a group of about 20 pixels in a diagonal slash, the worst pixel down about 30%, 3 more down 20%, another 3 at 17% and the remainder verging into the background. - centred at (337, 665) a group of 6 pixels the worst being about 13% down - centred at (124, 985) a group of about 17 pixels, the worst being 14% down, with 6 adjoining pixels about 11% down. - centred at (688, 677) a circle about 9 pixels diameter, the central one being 16% down. - centred at (249, 996) a cluster of 6 pixels, 2 of which are 50% down and four surrounding these are from 20-35% down. 3) Most pixels contained in the lower tail of the above responsivity histogram profile are seen as many hundreds of mostly single pixels or small clusters of double (or more) pixels that are from 2 to 15% down on their surroundings. These are fairly evenly scattered over the image area. 4) Some large scale structure is evident in the red LED flat-fields and appears as shading or patches of brighter areas with some regions of broad striping. These appear at the +/-2.5% level. 5) The small-scale pixel-to-pixel variation is generally about 0.7%. IR LED (880mn?) FLAT FIELD This is dominated by a large scale variation in responsivity from the most sensitive corner at (1, 1024) diagonally across the image area to the least sensitive corner at (1028, 1). The variation is about 13% p-p. In the more sensitive regions the flat-field takes on a mottled, blotchy appearance on the scale of 40 to 100 pixels. The horizontal lines and obvious defects listed for the red LED illumination can be seen. FLAT-FIELD ILLUMINATION USING A Hg+? UV CURING LAMP Unfiltered, this lamp (a Demetron Research Corp "Ultra Novar" lamp, Model # UV100, Se # 01990 UE, used for UV setting epoxies) produced an image full of fringes at a level just below 2% peak-to-peak. Interposing narrow-band filters and the standard B, V, R, I filters between the light source and the CCD showed that the fringes were generated only in the I-band. Narrow-band filters with central wavelengths of 418, 449, 502 and 578nm, to separate the Hg lines, showed no fringes, the response across the CCD appeared flat to better than +/-4%, and the horizontal lines and obvious defects listed above for the red LED illumination could be seen. The B, V and R filters behaved similarly. The I filter with the Hg+? lamp produced a maze of fringes with a p-p amplitude of up to 10%. As well as the Hg lines, this lamp had, starting at 6910A, a series of strong lines extending up into the IR to 11310A. DARK FRAMES AND DARK CURRENTS ***************************** 200K DARK FRAMES At 200K the dark current is about 0.1e/pix/sec and there is evidence that the dark current generation rate is a function of dark frame (DF) integration time (see below). 200K Dark Frames show brightened top and bottom rows. The top and the bottom rows of the CCD are affected almost equally with the first and last row of the CCD being the brightest. At longer exposure times (> 100sec) the first and last rows were found to be about about 60% brighter than the general image area dark current. The intensity rapidly falls to the general dark current level about 50 rows into the image area. Note that at 200K the dark current generation rate is sufficiently high that during the time taken to read out a frame the dark current is seen to ramp upwards from the start to end of the frame. 170K DARK FRAMES Dark frames are featureless apart from the readout induced hot top rows and the hot pixels that are described below. Dark current is about 0.55e/pix/2000sec provided the CCD is well cleaned out and not recovering from powering-on, saturation or high light level residuals. The hot top rows (at worst, the first two or three rows that are read out) may have hot spots that peak to several e/pix. The intensity of these is not a function of dark frame integration time and they appear in bias frames. Frames binned in X show them more clearly. Their intensities have been reduced in the process of reducing an excess dark current induced by the clocks during readout (described below). Before this reduction, several broad regions of brightening in the column direction were noticed corresponding to the hot spots of the hot top rows. The brightened columns have been almost eliminated but may show in XTRASLOW bias frames when pixels are summed in the column direction. The brightening is estimated to be less than 0.15 e/pix in XTRASLOW bias frames. An excess dark current is generated by both the vertical and horizontal clocks during the readout process. It has been reduced to less than 0.03e/pix allowing binned images to be made with up to 36 pixels/bin before the excess dark current exceeds 1 e/bin. DARK CURRENT GENERATION RATE At 200K the dark current generation rate itself seemed to increase with dark frame time as shown in the following sequence: DF Dark Rate EXPTM Current e/pix/sec (secs) e/pix 30 1.7 0.057 100 7.7 0.077 300 36 0.102 1000 120 0.118 3000 368 0.123 A repeat measurement of this effect, made with the final commisioning operating parameters for the CCD, produced the following result: DF EXPTM Norm Dark Current TEKBIN EXPTM (secs) (e/pix/2000sec) NORMAL RO + half 170K 200K 200K dk curr. RO time (e/pix) (secs) 1 - 47 0.34 15 3 - 52 0.42 17 10 - 73 0.72 34 30 0.523 106 1.94 44 100 0.532 138 7.9 114 300 0.584 192 30.3 314 1000 0.583 235 119.3 1014 2000 0.575 250 251.9 2014 3000 0.588 - 4000 0.566 - 6000 0.561 - There is no obvious dependence of dark current on EXPTM in the 170K data, compared with that found in the 200K data. HOT PIXELS ********** At 200K six hot pixels have been found. Their locations are referenced to (1,1) as first pixel out of array amplifier C in a 1028 X 1024 image. Pixel (343, 161) 5000e- in bias frames with a 100e-/pix trail behind them and possibly a 2e-/pix trail ahead of them in NORMAL readout. Generation rate is about 1900 e-/sec in DFs (344, 161) 2000e- in bias frames with a 40e-/pix trail behind them in NORMAL readout. Generation rate about 680 e-/sec in DFs (631, 606) 1000e- in bias frames with a 30e-/pix trail behind them in NORMAL readout. Generation rate about 410 e-/sec in DFs (900, 963) 600e- in bias frames, 20e-/pix trail behind them in NORMAL readout. Generation rate about 250 e/sec in DFs (632, 635) not seen in bias frames. Generation rate about 24 e-/sec in DFs (482, 247) not seen in bias frames. Generation rate about 1.5 e-/sec in DFs At 170K only the four brighter hot pixels in the above list are seen and at a very much reduced magnitude. On NORMAL readout there are no visible trails (343, 161) 100e- in bias frame, generation rate = 27 e/sec (344, 161) 17e- in bias frame, generation rate = 5.9 e/sec (631, 606) not seen in bias frames. Generation rate 4.1 e-/sec (900, 963) not seen in bias frames. Generation rate 1.9 e-/sec In bias frames made with the slower readout speeds there is evidence of trails from these hot pixels. These trails can seen clearly when the frame is summed in the column direction and may be visible in image displays especially of XTRASLOW frames. READOUT TRAIL INTENSITY (e/pix) at SPEED (343,161) (344,161) (631,606) NONASTRO none none none FAST 1.3 " " NORMAL 1.3 " " SLOW 2.7 0.6 " XTRASLOW 8.7 2.8 0.6 RESIDUAL IMAGES *************** Residual images from gross overexposure are very weak. At 200K, residual images from a 10X saturation intensity resulted in only about 4e/pix in a following 100sec dark frame (DF) and they had completely disappeared in a second 100 sec DF started 5 mins after the saturating readout. Residuals at 170K are barely visible. Residual images from a 10X saturation exposure resulted in only about 2e/pix in a following 500sec DF started immediately (3secs) after the completion of the readout of the saturating image. In a second 500sec DF, started 9 minutes after the first, the residuals had decreased to less than 1e/pix/500secDF. Vertical smearing above the saturated image areas due to shifting the saturated charge towards the readout amplifier produced weak residuals of less than 0.3e/pix/500sec in the first DF. At 170K, bright non-saturating illumination also produces very weak residuals. Folowing a 340Ke/pix exposure a 1000sec DF started 2 minutes later picked up only 1.9 e/pix. This fell to 1.4 e/pix in a following 1000sec DF. POWER-OFF-ON CLEANOUT ********************* The Tektronix CCD recovers far faster than the Thomson from powering off and than on again while cold. At 200K, powering the electronics off, then on again, causes dark currents that after five minutes settled to within about 30% of their ultimate value. At 170K, dark current (in e/pix/2000sec) falls to 10 after 5-10mins and to less than 2.5 after an hour. Several hours (perhaps even 24) are required before the ultimate figure of less than 1 e/pix/2000secs is reached. COOLING DEWAR PLUS POWER-ON CLEANOUT ************************************ It takes about 2 hours after filling the dewar with LN2 for the CCD temperature to fall to 170K and commence temperature regulation. Powering up the CCD as it cools provides a dark current of less than 7 e/pix/2000sec only 4 minutes after it attains 170K temperature regulation. SETTLING TIME 170K <---> 200K ***************************** The time it takes to change CCD operating temperatures is affected by the temperature of the dewar in that it influences the warming radiation falling onto the surface of the CCD. Going from 170 to 200K took the heaters 29 minutes at 20 degC top hat temperature, 40 mins at 15 degC and probably an hour at say 5 degC. The dark current at 200C settled as follows: TIME AFTER DARK CURRENT ATTAINING 200K e/pix/2000sec to start DF EXP 6 mins 212 (from 100sec DF) 11 201 " 15 192 " 21 188 " 34 180 " 44 174 " ultimate value 138 " This seems to indicate that about two or three hours should be allowed for the temperature and the dark current to settle. Going from 200 to 170K took 74 minutes at 20 degC top hat temperature, 46 mins at 15 degC and should be a liitle faster at say 5 degC. The dark current at 170K measured as follows: TIME AFTER DARK CURRENT ATTAINING 170K e/pix/2000sec to start EXP 3 mins 0.396 (from 500sec DF) 13 0.257 (from 1000sec DF) 34 0.356 " The CCD seems ready for use as soon as it has attained 170K. Re-adjustment of the pre-amp offset potentiometer by a technician is required whenever the temperature is changed. COSMIC RAY EVENT RATE ********************* Cosmic ray event rate is about 730 per 2000secs. This is about double the rate per square cm of chip area compared with the Thomson CCDs. At this rate, one event per 2.7 can be expected and even bias frames will have quite a few hits. BIAS LEVEL ESTIMATION ********************* NON-FLAT BIAS FRAMES The bias frames are reasonably flat, generally to within an ADU. The following table lists the magnitude of these defects found during a series of test frames. READOUT DEFECT (in ADUs) SPEED Y DIRECTION X DIRECTION LE ROLL RAMP LE ROLL NONASTRO -1.1 - < 0.1 FAST -0.8 - 0.1 NORMAL -0.4 - 0.1 SLOW -0.1 - < 0.1 XTRASLOW +0.8 0.4 0.3 DISCUSSION The bias level is a pure electronics offset. In itself it is noiseless - no photons can be attached to this number, it is the zero point - and this level has the readout noise of the CCD superimposed. Whether the bias level is 50 ADUs or 5000 ADUs will not affect the readout noise. In the past, variations in the bias levels with time have been noticed. Generally, the bias level fluctuated slowly plus or minus one or two ADUs but on rare ocassions the bias level had been seen to jump by several ADUs between two readouts (an extremely rare event during a readout). These problems seem to have been solved but some drifting in bias level may still be apparent. Considerable effort has been spent to remedy this but although some changes have been made and suspect components replaced, some drifting may still occur. The drifts are made worse (compared with the GEC and RCA CCDs) by the fast readout time of the Thomson and Tektronix CCDs requiring an extra gain of 2.5 to be placed after the DCS where, it appears, most of the drifts take place. The setting of a "gain balance" facility in the DCS circuit has been found to be readout rate dependant and the reason for this is not clear. The setting has been optimised for the NORMAL and the SLOW readout rates - FAST and NONASTRO are not so well compensated for general system drifts and may show a poorer bias level stability. THINGS THAT AFFECT THE BIAS LEVEL Astronomers should note that the bias level will be affected by the following: 1. Readout speed. There will be a significant change in bias level between XTRASLOW, SLOW, NORMAL, FAST, and NONASTRO so astronomers must take bias frames (and, indeed, all images for their programs) at the readout speed selected for the program. 2. Binning. All bias windows should be binned in the same way as the program image window 3. The Application. Cable lengths affect the bias level (focal reducer, PF, Cass, and on UCLES) through crosscoupling of the clocks to the video and the much reduced analog delay times instituted to speed up the readout rate. 4. Where the bias level is measured. There can be a small (usually less than 1 ADU) difference in bias level estimated from the extra rows (pre-scan rows) artificially generated at the top of the CCD and that estimated from overscanning the rows. The latter is regarded as the best technique and enables the astronomer to track more closely any variations of bias level if these are critical to the observation. 5. The overscan columns used to estimate the bias. Avoid columns less than 5 or 6 beyond the image area as these provide a margin for row direction CTE effects to decay. On ocassions the last column of a window has been suspect so it should first be compared with columns ahead of it before it is included in the estimate of bias level. HISTOGRAMS OF DATA VALUES ************************* Some problems of data conversion have been experienced in the past. Reduced analog delays to facilitate fast readout times have led to the possibility that the analog to digital conversion may be less than perfect. Astronomers should keep an eye open for recurring odd data values or histograms of data that show the presence of missing ADC codes. FAST CLEANOUT ************* This discussion relates mostly to the Thomson CCDs which suffer from large residual images that result from extreme saturation. With the Thomson CCD The effects described below are a serious problem with the Thomson CCDs but are not expected to be a problem for the Tektronix CCD due to its much reduced residual image characteristic. However the discussion is included so that astronomers are reminded of the possibility of the problem. In July, 1989, the CCD controllers were fitted with a fast cleanout facility. Before this feature was added, the cleanout phase that takes place just after a CCD readout, was achieved by the repetition (1000 times for small CCDs, 2000 times for the Thomson and the Tektronix) of the sequence one vertical shift followed by a shift out of the full row (plus a few more pixels) of the readout register. The cleanout phase of the Thomson would take about 120 seconds in "_bestcte" or 18 seconds otherwise. For the Tektronix CCD it takes about 30 seconds. Fast cleanout is a repetition (1000 or 2000 as before) of the sequence one vertical shift followed by only one horizontal shift of the readout register. This cleans out the Thomson CCD in just over 0.1 sec in the "_bestcte" mode of operation. The Tektronics CCD is cleaned out in just over one second. Using the same technique, fast cleanout rejects the unwanted CCD rows ahead of and following an astronomer window. This avoids the otherwise long delays that would result if these unwanted rows were fully shifted out in the customary fashion. In fact, the minimum "_bestcte" readout time of the Thomson CCD would be 72 secs for a very small window if the fast cleanout mode was not available. For the Tektronics CCD this time would be about 17 seconds. HOWEVER, it should be recognised that the fast cleanout process in effect bins all of the unwanted rows into the readout register and if these rows have a high intensity background, such as in high level flat-fields, there is a risk that the readout register will saturate (note that the readout register has a much higher charge handling capacity than an image area pixel - usually a factor of 3 or 4 times) and saturating charge will spread down the columns from the readout register. The extent of the downward spread from the unwanted charge existing above the window will not reach the astronomer window. Also, the readout register is thoroughly cleaned out by extra rows added above the astronomer window which are read out first and the data thrown out. But if there are a large number of unwanted rows below the astronomer window extending to the last row of the CCD, and these contain extended areas of high intensity, the resulting saturation when these rows are binned in the readout register may spread down from the readout register into the astronomer window area (but not into his data as that has already been read out). The result, however, will be the generation of higher dark currents eminating from the residuals of this saturated charge spread and may be seen in a subsequent dark frame or a long exposure on a faint field. NOTE that this is a problem only when windowing on a region of the CCD smaller than the "full" or "exact" windows, and then only when windowing in the vertical direction (ie full ROWS are being discarded), and then only if extended intense images are attempted such as high level flat fields with, say, 20,000e/pix. The resultant dark current effects on the next image are too difficult to quantify as it depends on the window size, position, intensity of previous bright images/flat-fields, time since these images, etc... OPTICAL ALIGNMENT ***************** The final optical alignment measurements indicated that the CCD has: a 20 micron tilt in X a 6 " " Y and is 20 " behind datum focus point ----- End Included Message -----