PHOTOMETRY

Since most known $\delta $ Scuti stars are stars brighter than $V \sim$ 8, telescopes of the 0.6 to 1m class equipped with photometers using photomultiplier tubes (PMT, photon-counting devices) are still used extensively.

At the millimag level of precision, it is not sufficient to observe the variable star continuously. The transparency variations need also be determined. Multichannel photometers, regrettably, offer only a partial solution. They are excellent in correcting the transparency changes, but introduce drifts of their own, since at the millimag level the channels drift apart. This can introduce severe errors.

For over three decades the Three-star observing technique used with single-channel photoelectric photometers has been the primary method to study short-period variable stars with periods longer than about 30 minutes (f $<$ 50 c/d or f $<$ 0.5 mHz) up to a few days. Although a precision of 2 mmag per single measurement had already been achieved more than 25 years ago (e.g., Breger 1966), the improvement since then has been minor. The present limit to the photometric precision is still just slightly under 1 mmag with good photoelectric photometers under excellent weather conditions. In fact, considerable observational care has to be taken in reducing the observational errors sufficiently in order to obtain the `old' limit near 1 mmag precision per single measurement.

Three-Star Photometry

The figure shows a light curve of the variable star FG Vir and two comparisons taken with the 3-star technique. The solid line is a calculated fit of 24 frequencies detected after the 1995 FG Vir global observing campaign.

This observing technique described by Breger (1990) should be aplied to variable stars with periods of about 30 minutes to a few hours.
The variable star, two comparison stars and the appropriate sky background are measured successively with the same photomultiplier tube (V -C1 -C2 -Sky -V - ...). In case of a large sky brightness gradient (moon or dusk/dawn) the sky has to be measured for each star separately. An integration time of 40 seconds per star has turned out to be a good compromise between high precision and short cycle time. In this case the cycle time is about 6 minutes.

The advantage of this observing technique lies in its long term stability. The sensitvity drift of the photometer and the transparency variations of the atmosphere can be corrected for. Pulsation periods of a few hours can be determined with high precision. The disadvantage is obvious: pulsations with periods shorter than 12 minutes (two times the cycle time) cannot be detected. For shorter periods another observing technique has to be applied, viz.;

 

CCD-Photometry

CCD frame of 44 Tau
Most well-studied $\delta $ Scuti stars are brighter than $V \sim$ 8, for which it is not really of advantage to use CCD detectors. The low duty cycle of $\sim $ 25% of the three-star technique is not necessarily improved by CCD detectors due to the relatively long readout time as compared to the exposure time for bright stars. Another disadvantage of CCD detectors concerns the small field of view (such as 2 arc mins), in which it is very generally impossible to find comparison stars of similar brightness.

However, for faint $\delta $ Scuti stars the high sensitivity of CCD detectors, the presence of many comparison stars on the same frame, and the ability to observe through non-photometric (so-called spectroscopic) weather conditions make CCD detectors invaluable.

To achieve millimag-precision a good guiding-system is need, bias frames and flat fields (10 - 20) have to be taken each night before and after the run and linearity checks have to be performed. For a good review on how to observe (open clusters) with a CCD please refer to the STACC Open Cluster Target List (Frandsen and Arentoft 1998, The Journal of Astronomical Data 4, 6).