Our new observations were obtained with three different
automatic photoelectric telescopes (APT) in the years 1991-96: the 0.75-m
Fairborn APT on Mt. Hopkins in Arizona, U.S.A. equipped with 
filters matching the Johnson-Cousins system, the 0.25-m
Phoenix APT also on Mt. Hopkins but equipped with Johnson UBV,
and the 0.8-m Catania APT on Mt. Etna in Sicily, Italy
with UBV filters for the Johnson system.
All measurements were made differentially with respect to a nearby comparison
star. Table 2 (click here) identifies the comparison stars and the check stars and
also gives the total number of obtained mean differential V magnitudes per
star and per telescope (
for the Fairborn APT, 
for
the Phoenix-10, and 
for the Catania APT). Both APTs on Mt.
Hopkins observed each program star differentially with respect to a comparison
star and a check star in the following sequence: 
,
where N is a
bright navigation star, CK is the check star, C the comparison star, S the
sky background usually between the comparison and
the variable star, and V the variable itself. One entry for n in
Table 2 (click here) thus compromises at least three individual 
readings.
In order to eliminate the datapoints grossly in error we applied a
statistical procedure that eliminated all data with an internal
standard deviation greater than 
as well as data that deviated
from the rest by at least 
(see, e.g., Hall et al.
1986). While the 0.02-mag filtering excluded between 
of the
whole data for a particular star the number of mean magnitudes excluded by
the 3-
procedure was usually just a few datapoints. We note
that, before eliminating the bad points, the whole data set
was phased with the most likely photometric period
and then examined for possible eclipses and flares.
The relative telescope zeropoints in the V bandpass have been determined from the seasonal averages of the check-minus-comp magnitudes of several groups that were on all three APTs simultaneously and agree within their formal uncertainties of around 0.01 mag.
The 0.75-m T7 APT was put into routine operation on JD 2 449 022
in 1993. During the first year of operation it had alignment
problems with the optics that led to a reduced data precision.
Altogether four problems (A-D) occured and their duration and influence on
the data is identified in Table 3 (click here). Three of the four problems were caused by
a filter-wheel malfunction that produced VVV photometry instead of
.
| Time | Influenc e | Cause | |
| 2449+ | on data | ||
| A | 144-164 | No R and I data | Filter wheel stuck at V |
| B | 235-246 | No R and I data | Filter wheel stuck at V |
| C | 312-322 | No R and I data | Filter wheel stuck at V |
| D | 600-850 |  increased | Telescope out of focus |
After the obviously deviant R and I data had been eliminated we computed
external uncertainties for all check-minus-comparison
magnitudes. Such
uncertainties allow a quick look at the data quality expected for the
variable-minus-comp data from the T7 APT. In 1993, its mean external
standard deviation of a "nightly mean'' from a yearly mean was
, , and 
mag in V, R, and I, respectively.
In the second year the telescope was continuously out of focus and the
annual mean external standard deviation was still 
, 
,
and 
mag for the three bandpasses. This has been fixed in early
1995 and the nightly mean external standard deviations decreased then to
, , and mag in V, R, and I, respectively.
By early 1996 the telescope has been continuously watched and the external
uncertainties dropped to 0.006 in V and below 0.010 in R and I.
Integration time was usually set to 10 s except for V410 Tau where 20 s
were used.
The Phoenix-10 APT is already in routine operation since 1983 and is managed
by Mike Seeds as a multi-user telescope (see Phoenix-10 Newsletter and Seeds
1995). Strassmeier & Hall (1988a) examined the data quality
of the Phoenix-10 APT from its first four years of operation and found
external uncertainties of 
, 
, and 
mag in
V, B, and U, respectively. Integration time was set to 10 s for all
targets. Recently, Henry (1995) compared the long-term external
precision of the Phoenix-10 APT with APTs of larger aperture (the
Vanderbilt/Tennessee State 0.4 m and the Tennessee State 0.8 m) and
verified the telescope's long-term stability.
First Catania-APT data came from the fourth quarter in 1992.
Its standard group observing sequence was set to
, with the same meaning for the symbols
as above. The sky background is measured at a fixed position near each star.
Each magnitude on the variable star consists of six readings, compared
to four with the other APTs. Integration time in U, B, and V was set
to 15, 10, and 10 s, respectively. The typical standard deviations of the
averaged and 
magnitudes for stars brighter than 
mag
are of the order of 0.015, 0.010, and 0.007
mag for U, B, and V, respectively. The accuracy of the standard V
magnitude is 0.01 mag, for U-B about 0.02 mag, and for B-V about 0.01
mag.