- 22 Apr 2020
LMIRcam detector linearity
As almost every detecor, LMIRCam has a linear range, a non-linear range, and a saturation level. These depend on the detector settings such as bias level. Regular measurements of the linearity are important for a range of aspects such as instrument health checks, non-linearity correction of science data, and selecting the appropriate peak counts when taking data. Data are taken by using a range of integrations and measuring the detector counts on the background from low counts to saturation.
- Type of observation: Closed dome, typically during night time.
- Duration: 3 h (data acquisition, can be split in three individual sequences for various detector settings), 2 h (data analysis & publication).
- Who: Steve Ertel or delegate (data acquisition), Steve Ertel or delegate (data analysis & publication).
- Interval: Data should be acquired once per observing run or every other observing run (few times per semester). The first data set of a semester is to be analyzed and provided to our science PIs. Later data sets are to be analyzed for monitoring purpose and published as needed.
Acquiring linearity data
This data acquisition procedure assumes that the operator is able to perform basic LMIRCam imaging data acquisition.
- Move telescope to zenith in closed dome.
- Set LMIRCam up for standard, single-sided imaging. Align the pupils as much as possible, use SX or DX aperture. Use full frame.
- Set PID to 19.
- Use the linearity scripts provided in the engineering director of the script templates. One script exists for each detetor mode (fast, medium, slow). Adjust the filters and integration times in the script so that saturation is reached around 2/3 of the maximum integration time. Also, make sure that at least 20 different integration times are used in total. These depend on ambient temperature, so have to be adjusted each time a script is used. Depending on setup, execution time can go up significantly, so be diligent in selecting a sensible filter/integration time combination.
- Run the script. It will take a sequence of images at different integration times and then a sequence of corresponding darks.
- Take data for slow, medium, and fast mode. These can be taken at different nights.
Analyzing distortion data
Data are analyzed using simple python scripts (ask Steve Ertel to provide his scripts if needed). Non-linearity at the 5% level should be reached around 3000 raw counts in fast mode and around 50000 raw counts in medium and slow mode.
Correcting science data
Correcting non-linear effects on science data is up to the science PIs. The raw or analyzed linearity data can be provided to PIs depending on their preference.
- Currently, Steve is the only one analyzing these data, which is not ideal. This is also not hard, so somebody else on the team should be trained.