Skip to content

Detector specific

Detector Specific troubleshooting

AGIPD

No corrected data even though there are available trains.

It has been reported more than once through the instruments and DOC that there offline correction showed didn't produce any data. And after the DOC checked the reports there is warning for no trains available to correct for all modules, therefore no processed data available.

This can happen in case the offline calibration is configured to correct illuminated data based on the LitFrameFinder device. This device points to the calibration pipeline which frames are illuminated to process. In case the data didn't have any no data correction will take place and no processed files will be produced.

No LitFrames

One of the reason for this can be that there were no beam at the time when the data was acquired. This is something the instrument scientists should be aware of and it is usually possible to check the shutter configuration switches for this particular run. As the shutter switches names can be changed. It wont be included here to avoid giving outdated dataset key names.

Below is a reason a table on the shutter switches state for a run acquired with no beam.

No beam

Gotthard2

GH2-25um: Correction fails No trains to correct error in PDF report.

GH2 25um No trains to correct

Check the corresponding sequence file using similar to what is mentioned in the general troubleshooting section

For 25um it is expected to find data for two modules (two separate receivers). Below is an example when a run had missing trains for one of the two modules. This is the out of h5glance

module load exfel exfel-python
h5glance /gpfs/exfel/exp/SPB/202401/p005476/raw/r0020/RAW-R0020-GH200-S00000.h5 INSTRUMENT/SA1_XTD9_HIREX/DET

/gpfs/exfel/exp/SPB/202401/p005476/raw/r0020/RAW-R0020-GH200-S00000.h5/INSTRUMENT/SA1_XTD9_HIREX/DET
├GOTTHARD2_MASTER:daqOutput
│ └data (3 attributes)
│   ├adc    [uint16: 100 × 2720 × 1280] (4 attributes)
│   ├bunchId    [uint64: 100 × 2720] (17 attributes)
│   ├frameNumber    [uint64: 100 × 2720] (17 attributes)
│   ├gain   [uint8: 100 × 2720 × 1280] (4 attributes)
│   ├memoryCell [uint8: 100 × 2720] (17 attributes)
│   ├timestamp  [float64: 100 × 2720] (17 attributes)
│   └trainId    [uint64: 100] (1 attributes)
└GOTTHARD2_SLAVE:daqOutput
  └data (3 attributes)
    ├adc    [uint16: 0 × 2720 × 1280] (4 attributes)
    ├bunchId    [uint64: 0 × 2720] (17 attributes)
    ├frameNumber    [uint64: 0 × 2720] (17 attributes)
    ├gain   [uint8: 0 × 2720 × 1280] (4 attributes)
    ├memoryCell [uint8: 0 × 2720] (17 attributes)
    ├timestamp  [float64: 0 × 2720] (17 attributes)
    └trainId    [uint64: 0] (1 attributes)

Jungfrau

No constants retrieved for burst mode data

Burst mode operates solely at fixed gain. This is realized either by fixed medium gain, or adaptive gain under the assumption of low illumination, thus being equivalent to fixed high gain with an additional safety margin. Therefore, we decided to only incorporate only fixed gain constants for burst mode.

The middleware tasked with executing dark runs has been updated to reflect this operational adjustment. Nonetheless, there have been occurrences where dark runs were manually acquired in instruments, creating a potential for error by injecting burst mode constants in adaptive mode. This misstep could result in either the failure to retrieve constants, as the anticipated fixed gain constants were not injected, or the retrieval of outdated constants, leading to incorrect corrections.