TriCCS is open in shared risk in the 2021B-2024A semesters.
Images after the flat-field correction may have sky non-uniformity of ~10%. It degrades the limiting magnitudes or produces a false detection of diffuse sources.
--> The improved filters (g2, r2, i2) were installed on 8 August 2022.
z-band filter was not tested well. There may be some problems we don't recognize.
An example of (upper left) g-, (upper right) r-, and (lower left) i-band images after dark subtraction and flat correction.
Sky non-uniformity in g- and r-band images and stray light in an i-band image are seen (before 8 August 2022).
Some stray light was detected in grism-r. This is due to the sensor inside TriCCS. Be careful when you observe faint objects.
This is a 2D image after subtracting the dark (The exposure time is 20 seconds). The position is x=1200--1500, y=0--400.
# Limitations
The available filter sets are "g2, r2, i2-bands" and "g2, r2, z-bands". Observers can switch the filter set with remote access.
The maximum data rate is 98 fps in the 2024A. Only available in full-image readout.
# Recommendations
Observers should take flat frames using twilight in sunset or sunrise to collect the non-uniformity of throughput in optics and detector pixels.
To subtract the dark current pattern in CMOS sensors, observers should take dark frames with the shutter closed. The dark pattern varies depending on the exposure time and the CMOS gain. We recommend obtaining about the same total exposure time of dark frames as object frames at one field.
We advise observers to take not a single long exposure frame but multiple 10- to 30-second exposure frames for obtaining deep images. TriCCS frames saturate in a short exposure time by sky background due to the small output depth, 14-bit. Observers can reject frames affected by cosmic rays or other unexpected noises in data reduction.
TriCCS produces enormous data in the fast mode. The data amount of a 10-fps 8-hour observation is 5.2 TB, and it takes about 7 hours to copy it to a hard disk drive.
The absolute time in the fits header is occasionally inaccurate. Before using the absolute time, make sure it is correct by comparing it to the time of the previous or next frame.
