Vol. 7 No. 2 April 1, 2011 Journal of Double Star Observations Page 64 Double and Multiple Star Measurements at the Southern Sky with a 50cm-Cassegrain and a Fast CCD Camera in 2008 Rainer Anton Altenholz/Kiel, Germany e-mail: rainer.anton”at”ki.comcity.de Abstract: Using a 50cm Cassegrain in Namibia, recordings of double and multiple stars were made with a fast CCD camera and a notebook computer. From superpositions of “lucky images”, measurements of 149 systems were obtained and compared with literature data. B/W and color images of some remarkable systems are also presented. 50 cm and focal ratio of f/9. It is located on a guest Introduction farm in Namibia and owned by the Internationale As has already been demonstrated in earlier pa- Amateur-Sternwarte (IAS) [1]. The setup for re- pers in this journal, the accuracy of double star cording was the same as I used earlier at home as measurements can be significantly improved by the well as at the southern sky [2-5]. In most recordings, technique of “lucky imaging”. Using short exposure I used a 2x Barlow lens (Televue) to extend the effec- times, only the best frames out of some thousands tive focal length to about 9 m (f/18). With my b/w- are registered and stacked. Thus, seeing effects are CCD camera (DMK21AF04, The Imaging Source) effectively reduced, and the resolution of a telescope with pixel size 5.6 µm square, this results in a reso- can be pushed to its limits, even under non-optimum lution of about 0.13 arcsec/pix. An exact calibration average seeing conditions. The accuracy of position was obtained by an iterative method by measuring measurements can even be better than this by about well documented double stars (see below). Generally, one order of magnitude. In total, 179 measurements I used a red filter to reduce chromatic aberration of of 149 double and multiple systems, mostly brighter the Barlow lens, as well as the atmospheric spec- than 10 mag, have been done. While in the majority trum. For systems with pronounced color contrast, I of cases, not too many data are found in the litera- made additional recordings with green and blue fil- ture to compare with, often with large scatter, 34 ters in order to produce composite images. To pairs with sufficiently well known separations could roughly compensate the variation of sensitivity with be used as reference for calibration. About 36 pairs wavelength, exposure in blue was typically set to two are binaries with more or less well known orbits. In times that in red and green. Exposure times varied some cases, deviations from ephemeris data were between 0.5 msec and 100 msec, depending on the found, in accordance with measurements from other star brightness. Under good seeing conditions, some authors. systems were also recorded with exposures up to 0.5 Instrumental sec, in order to image faint companions. The proce- dure of image processing was essentially the same as The telescope is of Cassegrain type with aperture reported earlier. The yield of “lucky” frames ranged Vol. 7 No. 2 April 1, 2011 Journal of Double Star Observations Page 65 Double and Multiple Star Measurements at the Southern Sky with a 50cm-Cassegrain ... from several tens to more than a hundred frames, de- constant becomes important at greater separations. A pending on the seeing. These were then re-sampled, similar analysis was made for recordings without registered (often manually), and finally automatically Barlow, resulting in a calibration constant of 0.257 stacked. This process resulted in smooth intensity arcsec/pix. In Figures 1 and 2, the final residuals of profiles, and in position measurements with sub-pixel the separations and position angles are plotted for all accuracy. pairs, for which extrapolation of literature data or even direct comparison is reasonable. It should be Calibration noted that the remaining scatter is due to possible The image scale was calibrated by measuring a errors of both literature data, and of own measure- number of double stars with sufficiently well known ments. Because of the above limitation, the scatter separations. These are mainly selected from the WDS appears relatively small, but in fact, it is much larger [6], and the so-called speckle catalog [7]. Data avail- for other systems, which are not included here be- able up to fall 2010 are taken into account, as of writ- cause of too few or less trustworthy literature data. ing this article. Criterion is that literature data can Position angles were measured by referring to unambiguously be extrapolated to the actual date. star trails in east-west direction, which were obtained This may even include binaries. In total, measure- by superposing frames with short exposures, while ments of 34 such systems were obtained, which are the system was drifting across the field with the tele- contained in the table below, and are marked with scope drive shut off. This was done for every recording shaded lines. The calibration constant was adjusted of any system in order to avoid errors, which may oc- such that the sum of the residuals as well as the stan- cur by slight misalignment of the mounting. The error dard deviation assumed minimum values. For re- margin was ±0.1 degree. The residuals of the position cordings with Barlow, statistical analysis resulted in angles, delta PA, are plotted versus the separation a value of 0.132 arcsec/pix, with an error margin of +/- rho in Figure 2. Deviations increase with decreasing 0.5%. The standard deviation of the residuals (of the separation, mainly because the lateral resolution is pairs used for calibration) is ±0.038 arcsec, with range fixed by the pixel size, but also due to significant scat- between maximum and minimum of ±0.09 arcsec. The ter of literature data. When taking into account only absolute total error margin is the sum of both contri- systems used for calibration of the image scale, statis- butions. While at small separations, the statistical tics result in a standard deviation of ±0.33 degrees error dominates, the possible error of the calibration with range between maximum and minimum of +0.7 Figure 1: Plot of the residuals of rho versus rho. Semi- Figure 2: Plot of the residuals of the position angle versus logarithmic scale. Full and open circles indicate systems rho. Semi-logarithmic scale. The increase of scatter to- recorded with and without Barlow, respectively. Circles with wards small separations is caused by scatter of literature crosses represent systems, which have been used for cali- data, as well as by the fixed image resolution. Some pairs bration of the image scale. The curves mark the total error are indicated with numbers of corresponding notes below. limits. The reason for the peculiar deviation of the pair beta STF 147 is not resolved, and the P.A. is only a rough esti- Leporis, as indicated (the number refers to the notes be- mate. low), is unclear, as both extrapolation of literature data seems to be fairly unambiguous, and the own measure- ment rather trustworthy, but differing. See text. Vol. 7 No. 2 April 1, 2011 Journal of Double Star Observations Page 66 Double and Multiple Star Measurements at the Southern Sky with a 50cm-Cassegrain ... and -0.9 degrees, respectively. response of the CCD camera to large differences of brightness, combined with the strong variation with Results wavelength. In all images, north is down, and east is All measurements are listed in the following ta- right. When indicated at the bottom, position angles ble, which is followed by individual notes. Asterisks and separations are as measured from the respective denote systems, for which images are shown below. image. These are chosen for different reasons. In general, A number of the systems listed here have also they depict the image quality, which can be obtained been recorded during the same stay in Namibia with with lucky imaging, which is often apparent by the a 40 cm Cassegrain telescope. The results have been diffraction rings. Images of close binaries are shown published in a previous paper in this journal [4]. Gen- in Figure 3. Figures 4 and 5 illustrate the case of a erally, agreement was found within the error limits. binary (STF 2744 Aqr), of which the calculated orbit In doubt, the present data may be more accurate, be- is still somewhat uncertain. It also compares the scat- cause of the larger telescope. ter of data from lucky imaging with that of speckle It may be noted that in several cases, data from measurements. Other images show large delta m lucky imaging had a posteriori been confirmed by pairs (Figure 6), or interesting multiple systems speckle data of about the same epoch, within the error (Figure 7). RGB composites are presented of pairs limits. with marked color contrast (Figure 8). These should not be taken too seriously, because of the non-linear (Continued on page 76) Table 1: List of all measurements. Systems used for calibration of the image scale are marked by shaded lines. Names, positions and magnitudes are taken from the WDS. Position angles (P.A.) are in degrees, separations (rho) in arcseconds. N is the number of different recordings. Residuals delta are given, when extrapolation of literature data appears reasonable. P.A. rho delta delta PAIR RA + DEC MAGS DATE N NOTES meas. meas. P.A. rho BU 391AB 00 09.4 -27 59 6.13 6.24 259.3 1.38 2008.725 1 ~0 -0.02 1 I 701 00 20.2 -35 54 8.33 8.76 58.0 0.63 2008.749 1 +0.6 -0.02 2 LCL 119AB 4.28 4.51 168.5 27.1 2008.725 1 +0.3 ~0 00 31.5 -62 57 3 I 260CD 4.60 6.54 250.1 0.32 2008.725 1 -0.6 -0.01 STF 39AB-C 00 34.5 -04 33