S
ERIAL
I
NTERFACE
D
EVELOPERS
G
UIDE
(O) Host will monitor the status code data member until it is set by the embedded application to
CAL_CALC_COEFFICIENTS or CAL_DELTA_TOO_SMALL. If the CAL_DELTA_TOO_SMALL
code is received then the embedded application has aborted the calibration.
(P) The host/user may now remove the hot reference if desired.
(Q) Host will monitor (periodically read) the status code data member until it is set by the embedded
application to CAL_COMPLETED.
(R) Read the “CameraConfig.camStats.HostData” member to retrieve number of defects if desired.
(S) Calibration is complete.
7.7 Defective Pixel Detection
Lumitron has removed the automatic detection feature from the camera. Defective pixels are now
solely identified through calibrations and external operations defined by the user.
7.8 User Defined Defective Pixel Map
This process can be followed to supply a user defined defective pixel map to the active NUC table. It
assumes the host already has created a defective pixel map of the entire array. The pixels will marked
as defective in the active NUC table, which will be erased in the process. A 2-point calibration is
required after this process is complete.
(A) Disable the automatic NUC refresh operation by setting BIT-15 of the camera configuration
bitFieldIndex member.
(B) Using the ‘McbWriteDataMem’ command - write a portion of the map to the scratch pad buffer.
Remember it is sized at 160 words.
(C) Send the CMD_WRITE_UTILITY_MEMORY command with the proper address offset (base is
MAR_IMAGE_GRAB_B) and data size (see paragraph
X
5.3.43
X
) to move the data from the DSP
scratch pad to the Xilinx utility memory.
(D) Repeat steps (B) and (C) until the entire map is loaded to utility memory.
(A) Send CMD_ADV_DETECT_BAD_PIXELS command (no arguments).
(B) Embedded application will set the status code to BEGIN_PROCESS, and then begin execution of
the routine.
(C) Host will monitor (periodically read) the status code data member until it is set by the embedded
application to DEFECT_ERASE_FLASH.
(D) If DEFECT_ERASE_FLASH status is detected then the embedded application needs to know
whether the host wants the NUC flash erased or not.
(E) Host needs to set the “CameraConfig.camStats.HostData” member to 0xFFFF for erase flash or to
0x0000 to skip the erase process and exit.
(F) Host application sets the status code to DEFECT_ERASE_ACK.
(G) Host will monitor the status code data member until it is set by the embedded application to
DEFECT_COMPLETED.
(H) Re-enable automatic NUC refresh bit.
(I) The user defective map now exists in the active NUC flash and is ready for a two point calibration.
7.9 Upload NUC Table from Host
This process can be followed to supply a host created set of NUC coefficients to the desired table
index. It assumes the host already has created the offsets, gains, and replacement index (if needed)
41
