A round tuit arrived today so I uploaded the two nights of THL data to the CalTech periodogram engine. The results are (to me at least) are interesting. Several methods of period searching were used. All gave peaks at close to 0.0281 days and twice that (0.0562 days). The former is 40.3 minutes, which seems rather short. Twice that, 80.6 minutes, is absolutely typical of CV binaries. Perhaps a little on the short side but not exceptionally so — that of WZ Sge is 0.05671 days. The 40.3m light curve shows only a single dip. The other, of course, shows two which if real suggests eclipses of two stars of fairly similar magnitudes.
Another period which comes up strongly is at 0.08 days — 2 hours or so — or perhaps twice that. This is doubtless the 0.1 variation mentioned in the earlier post.
Only two nights and 359 measurements were analysed. Time for me to download more of the BAA-VSS database and see whether the patterns hold up.