I have been studying a variable star – RV CVn – since April of 2022 – in relation to a post I made that showed a partial light curve.  My goal was to fill out the light curve with additional observing runs and have now gotten enough data to be worthy of posting at this point.  This binary star is composed of two stars of very similar mass and brightness that orbit one another in a tight circular orbit; the orbit is so tight that the outer atmospheres are actually in contact.  Owing to the close orbital proximity, the orbital period is short, currently stated to be 0.269566 days.  Since the two component stars orbit one another along our line of sight, they take turns eclipsing one another.  During an eclipse, the total light visible to us diminishes and when the stars are apart in our view, the light visible is at its maximum.  Accordingly, the magnitude undergoes periodic dimming in a peak-like fashion when the eclipses happen.  Since the stars are nearly alike, the eclipses produce very similar total dimming and owing to the circular orbit, the eclipses are separated nearly equal in time.

RV CVn is not extensively studied.  However, a paper by P. Zasche, M. Wolf, R. Uhlar, and H. Kucakova (Astronomical Institute, Charles University in Prague, Faculty of Mathematics and Physics) included observations of RV CVn.  The authors note that the orbital period of the component stars appears to be cyclically variable, and they hypothesize that this may be due to a third star in the system that is too dim to be located photographically, but is perturbing the orbit of the main components.  In addition, there is some uncertainty in the published magnitude range of RV CVn as found by comparing data given in various catalogs. This is the light curve I obtained (based on 4 sets of observations on 4/8/2022, 2/26/2023, 2/27/2023 and 3/18/2023):


Of important note is that all of the data sets were analyzed with new software; instead of relying on Muniwin, I switched to ASTAP, which (1) manipulates fit images without damaging the header files, (2) stacks and aligns fit images, (3) does photometric analysis and (4) uses an extensive database of catalogued stars to determine magnitudes of both comparison and variable stars!  To ensure that the comparison stars are truly not variables themselves, the program outputs their magnitudes over time, giving this result for one of the data sets as an example:


A comparison of my light curve to one published by Zasche et. al. shows reasonably good similarities, but there are some differences.  Given that RV CVn is possibly subject to changes in period and perhaps magnitudes, the differences may be real and, of course, they could be all or partly experimental error on my part.  As a check, I tracked down the magnitude of the comparison star for the blue data points above in an independent catalog, obtaining 13.9 versus a mean of 14.0 from ASTAP, suggesting that the ASTAP magnitudes for RV CVn are accurate to within about 0.1 mag. I hope to get a bit more data and then submit my results to the AAVSO.  Here is a rough graphical comparison of the Zasche results and my light curve (my data in blue):


Since RV CVn lies close to M3 (but is probably not part of the cluster), I put all of the data together to create the image of M3 posted herewith.  The tallying of the integration time that resulted is pretty complex due to bad images that got tossed and other things, but I estimate it to be about 4 hours.  Most of the images in the data set were 30 seconds in duration.
I apologize for the long post due to the inclusion of the diagrams.  I may be able to handle variable stars, but I am far less able to manage showing illustrations as revisions in a post!