This comment is related with file ATL-DAQ-PROC-2016-015, version 2
I received a request for minor revisions from the JINST referees. I have implemented all of them and submitted the new version to CDS and JINST.
The comments from the JINST referees and the modifications on the text are described below:
Referee 1
Line 79. factor of 3 ->a factor of 2, or just give the target value, since the 2016 peak luminosity reached 1.5E34.
--> done as suggested.
Lines 87-88 provide the numerical granularity since the factors are difficult to follow.
--> added "ten super-cells per trigger-tower".
Line 180: Provide length of the electrical cables.
--> added "11m".
Lines 222, 230, 234, 246, 276: Provide references for the Xilinx and Samtec parts (could just be one reference each for Samtex and Xilinx).
--> added references for Xilinx Artix-7 and Kintex UltraScale FPGAs and for the Samtec FireFly optical transmitters.
Referee 2 (mostly rewording requests, implemented all of them as suggested)
Line 65-68 add the Both processor systems produce results in the form of Trigger Objects (TOBs), which describe the location, energy and type of identified physics objects.
Line 70-71 move also which also receives object multiplicities from the CP and JEP subsystems.
Lines 71-74 Add preposition of and reword slightly: "Based on all of this information the CTP takes a final L1 trigger decision for each BC. If positive, then the CTP issues a Level-1 Accept (L1A) signal to all of the ATLAS sub-detectors and readout electronics. Upon receipt of the L1A, all of the L1Calo subsystems transmit event related data to the ATLAS Data Acquisition (DAQ) system."
Line 75-77 The CP and JEP simultaneously produce Region-of-Interest (RoI) information that describes the coordinates of the detector regions where the interesting physics signatures have been identified and send it to the HLT.
Lines 88-91 Add article the, move also and reword: "Also, the digital input from the LAr Calorimeter will be of finer granularity than the similar analogue input, to help L1Calo improve the event selection. The Tile Calorimeter will continue to send analogue trigger-tower signals to the PreProcessor until the Phase-II upgrade (2023-25) is implemented. In order to also provide the FEX processors with Tile digitised results."
Line 125 - Define acronym: "These were upgraded to new MCMs (nMCMs) during"
Line 127-128 Reword: "(double the LHC bunch-crossing frequency)"
Line 135-136 Define acronym: "(Finite-Impulse-Response (FIR) Filter and PeakFinder for pulses in linear range,..."
Line 150-153 Reword: "Upon the receipt of an L1A from the CTP, the event data are first collected and formatted by an on-board Readout Manager (ReM) FPGA, then they are presented to a Rear G-Link Transmitter Module (RGTM), mounted on the back side of the crate, and finally they are sent to the DAQ via an external Readout Driver (ROD) module."
Line 156 "The configuration and control of the PPM are realised"
Line 165 Move also: "In order to also supply the FEXes with digital trigger data from the Tile Calorimeter,"
Lines 172-175 Split run-on sentence: "The PPM real-time data path will be extended to the TREX to facilitate the extraction of signal copies for the FEX processors. The cable driver functionality, currently implemented on the LCD card, will be re-located to the TREX, to provide continuity in the trigger operation."
Lines 185-187 Reword: "The data will then be serialised and sent to the FEXes, using up to 48 high-speed optical links. The baseline rate for these links has been established by the L1Calo and LAr groups to 11.2 Gbps"
Line 198-191 Reword: "Once the performance of the FEX processors has been validated during the commissioning phase, and the CP and JEP subsystems have been decommissioned, the PPM can be configured to double the output data rate on the real-time path, i.e. to 960 Mbit/s. This operational mode will permit either the BC-multiplexing of the 0.1 × 0.1 ET results to be removed or the additional information to TREX to be sent, if needed."
Lines 194-195 Reword: The current LCD card, which was designed more than a decade ago, is based on old generation FPGAs that cannot cope with the higher rate."
Lines 195-197 Reword: "The new LCD design will be simplified to include only passive components, and the functionality of the card will be reduced to route only the LVDS signals to the PPM output connector."
Lines 198-199 Reword: "The TREX also will have the task of collecting, formatting and transfering the PPM event data to the DAQ system."
Lines 213-214 Reword: "For testing and diagnostic purposes, the TREX also will have the capability to receive and analyse data from up to 12 high-speed optical links."
Line 252-254 Move also: "The prototype design also includes a 10-channel PLL to generate low-jitter clock signals for the multi-Gb/s transmission."
Line 258 Move also: "The PLL chip also will have the task"
Line 271 Reword: "to power on and off the module at the same time as the front PPM."
I received a request for minor revisions from the JINST referees. I have implemented all of them and submitted the new version to CDS and JINST.
The comments from the JINST referees and the modifications on the text are described below:
Referee 1
Line 79. factor of 3 ->a factor of 2, or just give the target value, since the 2016 peak luminosity reached 1.5E34.
--> done as suggested.
Lines 87-88 provide the numerical granularity since the factors are difficult to follow.
--> added "ten super-cells per trigger-tower".
Line 180: Provide length of the electrical cables.
--> added "11m".
Lines 222, 230, 234, 246, 276: Provide references for the Xilinx and Samtec parts (could just be one reference each for Samtex and Xilinx).
--> added references for Xilinx Artix-7 and Kintex UltraScale FPGAs and for the Samtec FireFly optical transmitters.
Referee 2 (mostly rewording requests, implemented all of them as suggested)
Line 65-68 add the Both processor systems produce results in the form of Trigger Objects (TOBs), which describe the location, energy and type of identified physics objects.
Line 70-71 move also which also receives object multiplicities from the CP and JEP subsystems.
Lines 71-74 Add preposition of and reword slightly: "Based on all of this information the CTP takes a final L1 trigger decision for each BC. If positive, then the CTP issues a Level-1 Accept (L1A) signal to all of the ATLAS sub-detectors and readout electronics. Upon receipt of the L1A, all of the L1Calo subsystems transmit event related data to the ATLAS Data Acquisition (DAQ) system."
Line 75-77 The CP and JEP simultaneously produce Region-of-Interest (RoI) information that describes the coordinates of the detector regions where the interesting physics signatures have been identified and send it to the HLT.
Lines 88-91 Add article the, move also and reword: "Also, the digital input from the LAr Calorimeter will be of finer granularity than the similar analogue input, to help L1Calo improve the event selection. The Tile Calorimeter will continue to send analogue trigger-tower signals to the PreProcessor until the Phase-II upgrade (2023-25) is implemented. In order to also provide the FEX processors with Tile digitised results."
Line 125 - Define acronym: "These were upgraded to new MCMs (nMCMs) during"
Line 127-128 Reword: "(double the LHC bunch-crossing frequency)"
Line 135-136 Define acronym: "(Finite-Impulse-Response (FIR) Filter and PeakFinder for pulses in linear range,..."
Line 150-153 Reword: "Upon the receipt of an L1A from the CTP, the event data are first collected and formatted by an on-board Readout Manager (ReM) FPGA, then they are presented to a Rear G-Link Transmitter Module (RGTM), mounted on the back side of the crate, and finally they are sent to the DAQ via an external Readout Driver (ROD) module."
Line 156 "The configuration and control of the PPM are realised"
Line 165 Move also: "In order to also supply the FEXes with digital trigger data from the Tile Calorimeter,"
Lines 172-175 Split run-on sentence: "The PPM real-time data path will be extended to the TREX to facilitate the extraction of signal copies for the FEX processors. The cable driver functionality, currently implemented on the LCD card, will be re-located to the TREX, to provide continuity in the trigger operation."
Lines 185-187 Reword: "The data will then be serialised and sent to the FEXes, using up to 48 high-speed optical links. The baseline rate for these links has been established by the L1Calo and LAr groups to 11.2 Gbps"
Line 198-191 Reword: "Once the performance of the FEX processors has been validated during the commissioning phase, and the CP and JEP subsystems have been decommissioned, the PPM can be configured to double the output data rate on the real-time path, i.e. to 960 Mbit/s. This operational mode will permit either the BC-multiplexing of the 0.1 × 0.1 ET results to be removed or the additional information to TREX to be sent, if needed."
Lines 194-195 Reword: The current LCD card, which was designed more than a decade ago, is based on old generation FPGAs that cannot cope with the higher rate."
Lines 195-197 Reword: "The new LCD design will be simplified to include only passive components, and the functionality of the card will be reduced to route only the LVDS signals to the PPM output connector."
Lines 198-199 Reword: "The TREX also will have the task of collecting, formatting and transfering the PPM event data to the DAQ system."
Lines 213-214 Reword: "For testing and diagnostic purposes, the TREX also will have the capability to receive and analyse data from up to 12 high-speed optical links."
Line 252-254 Move also: "The prototype design also includes a 10-channel PLL to generate low-jitter clock signals for the multi-Gb/s transmission."
Line 258 Move also: "The PLL chip also will have the task"
Line 271 Reword: "to power on and off the module at the same time as the front PPM."
Victor.