CALPRODUCT-VS-025: VIS Trap Pumping

Parent requirement

R-CAL-B-VS-4351

Scheduling

PV phase: Dedicated observations

Routine phase: Taken as part of the ROS

Time-critical

No

Objective

Characterise and monitor the defects in the detectors, which are responsible for charge transfer inefficiency. The number of these defects will increase over time due to radiation damage, and existing ones may change or disappear due to annealing effects.

Trap pumping (TP) is suitable to monitor these defects. This is achieved by clocking the charge back and forth in the detector. Depending on the shift time used, the trap will produce a dipole and by using a range of shift times, a dipole curve can be produced. This curve can be fitted to extract the emission time constant and the pumping efficiency of the trap.

Both serial and parallel trap pumping is required to constrain the traps within a pixel.

Observations

Observations are taken during PV phase, as well as part of the ROS.

Full parameter set for parallel trap pumping

Full parameter set for serial trap pumping

Full parameter set for trap pumping in the ROS

The Charge injection is followed by charge shuffling sequence.

CI parameters common for parallel and serial trap pumping:

  • TOI_CHINJ = 143 us

  • ID_DELAY = 191 us

  • DAC_IG2 = 6 V

  • 2 CI levels (IG1=4V and IG1=4.75V)

Parallel trap pumping: full parameters set

Parallel / vertical trap pumping needs to be done at a range of CI levels, VMODES, and TOI values. A total of 90 vertical trap pumping sequences have been defined (VTP01 to VTP90).

  • All 4 VMODES need to be run sequentially (234, 123, 341, 412)

  • Each VMODE needs to be run at the 2 CI levels defined above (IG1=4V and IG1=4.75V)

  • Each VMODE and CI level needs to be run at 11 TOI values (\(\mu{\rm s}\)):

    10, 20, 50, 100, 200, 500, 1000, 2000, 4000, 8000, 16000

  • SHUFFLES:

    • 4000 for CI level IG1=4V (Except for TOI_TP=16000 where 3000 SHUFFLES should be used to stay below 150s)

    • 2000 for CI level IG1=4.75V

    • Additionally, SHUFFLE=1 corresponds to a “no pump,” to get CI levels, which serves as a reference.

Within the ROS during the wide survey, The 90 VTP sequences will be run sequentially with the reference VTP sequences (VTP01 and VTP13, each with SHUFFLE=1) interspersed as necessary.

Serial trap pumping: full parameter set

Serial trap pumping is done where a single command is a multimode scan where 8 separate parameter sets can be run in a single sequence. A range of CI levels, SMODES, and TOI values need to be run. A total of 12 serial trap pumping sequences has been defined (STP01 to STP12).

  • 2 SMODES need to be run sequentially (23, 31)

  • Each SMODE needs to be run at the 2 CI levels (IG1=4V and IG1=4.75V)

  • There are 3 different Multi-STP configurations of [SER_DWELL, SHUFFLES, NROWS] (see below).

  • A reference STP (with SHUFFLE=1) must be placed at the top of the 8 STP sequence, and should contain extra number rows compared to the rest of the STP types, to mitigate against the initial charge injection non-uniformity.

  • The last STP in the 8 STP sequence shall also contain a reference STP (SHUFFLE=1), to be able to adjust for any ramp in the CI level (that was seen in the VTP data), and monitor any gain/bias level offsets that was present in the initial MSTP sequences.

  • Due to hardware limitations, the “no pumping” reference (to get the CI level) is approximated by a SHUFFLE=1 run.

configuration 1:

  • SER_DWELL (\(\mu{\rm s}\)): [0, 0, 4.75, 19, 57, 114, 342, 0]

  • SHUFFLES: [1, 2000, 2000, 2000, 2000, 2000, 2000, 1]

  • NROWS: [220, 150, 150, 150, 150, 150, 95, 100]

configuration 2:

  • SER_DWELL (\(\mu{\rm s}\)): [0, 1.175, 9.5, 28.5, 71.25, 142.5, 285, 0]

  • SHUFFLES: [1, 2000, 2000, 2000, 2000, 2000, 2000, 1]

  • NROWS: [220, 150, 150, 150, 150, 150, 110, 100]

configuration 3:

  • SER_DWELL (\(\mu{\rm s}\)): [0, 2.375, 14.25 , 42.75 , 85.5 , 171, 228 ,0]

  • SHUFFLES: [1, 2000, 2000, 2000, 2000, 2000, 2000, 1]

  • NROWS: [220, 150, 150, 150, 150, 150, 135, 100]

In the wide survey, STP will not be executed within the ROS, but separately as a monthly calibration. Due to hardware limitations, the Multi-STP sequences must be followed by 3x “trash” biases, to restore the BIAS levels to the original values. (Residual bias after Multi-STP appear as CCD top-bottom asymmetry in the BIAS level.)

Fulfilled by

CALBLOCK-PV-027: VIS Trap Pumping

CALBLOCK-F-006: Wide Survey

CALBLOCK-F-015: VIS Serial Trap Pumping

Data volume

Parallel: Full frame, but CI data, so compression should be above average.

Serial: 200-300 lines in total, but data format and compression not known.

Processing input

Ground Processing Input Data: Parallel For each VMODE and CI level, a set of 11 TOI_TP values to be used

Serial For each VMODE and CI level, a set of configuration 1 and 2 frames.

Data Description: Parallel 90 full frames in 8 different VMODE and CI level configurations

Serial 12 frames with either 1165, 1180, or 1205 lines

Processing method

VIS_CTI_TPCALIBRATE

Processing output

Position, emission time constant and pumping efficiency estimate of each detected trap.

Required accuracy, constraints

TBW

Processing tools

VIS_CTI_TPCALIBRATE

Validation

TBW

Recovery

ROE timeout error is generated for any serial trap pumping (STP) that lasts longer than 78 seconds. This particular error message does not indicate a failure in STP data acquisition. However this feature is retained to signal failure in all other normal readouts (STP is a special type of readout). There is no recovery required for this error.

The relevant limitation is as follows:

VIS-L-010: The readout phase in trap pumping sequences can last longer than the maximum readout time of 78s, which is hardcoded in CPDU FPGA, depending on the settings applied (number of lines, number of shuffles, dwell time, …). When this happens, the CDPU will report a ROE timeout error.

Quality control parameters

DataModel: https://gitlab.euclid-sgs.uk/

Contact person

VIS instrument scientist