CERN Accelerating science

Architecture of L1Calo Trigger in Run~2.

Architecture of L1Calo Trigger in Run~2.

Normalised pulse shapes for TTs from the EM barrel and FCAL as well as the HAD barrel calorimeter regions, recorded with an oscilloscope during $pp$ collisions at the Receiver output.

Normalised pulse shapes for TTs from the EM barrel and FCAL as well as the HAD barrel calorimeter regions, recorded with an oscilloscope during $pp$ collisions at the Receiver output.

Photograph of a PPM, highlighting the different types of components as described in Section~\ref{sec:preprocessor}. This PPM is already equipped with new Multichip Modules, which are covered by heat sinks, labelled with the unique serial number of the chip (see Section~\ref{sec:nmcm_hw}).

Photograph of a PPM, highlighting the different types of components as described in Section~\ref{sec:preprocessor}. This PPM is already equipped with new Multichip Modules, which are covered by heat sinks, labelled with the unique serial number of the chip (see Section~\ref{sec:nmcm_hw}).

: Photographs of the original (top) and new (bottom) Multichip Module. The highlighted components are: (A) the PHOS4 chip, (B) the FADCs, (C) the PPrASIC and (D) the LVDS transmitters on the MCM; (1) the signal generator circuit, (2) the operational amplifiers, (3) the FADCs, (4) the CALIPPR FPGA (Xilinx Spartan-6) and (5) the power converters on the nMCM. The EEPROM device is mounted on the far side of the nMCM. (S) indicate the screw holes for mounting on the PPM. The nMCM is slightly longer than the MCM.

Photographs of the original (top) and new (bottom) Multichip Module. The highlighted components are: (A) the PHOS4 chip, (B) the FADCs, (C) the PPrASIC and (D) the LVDS transmitters on the MCM; (1) the signal generator circuit, (2) the operational amplifiers, (3) the FADCs, (4) the CALIPPR FPGA (Xilinx Spartan-6) and (5) the power converters on the nMCM. The EEPROM device is mounted on the far side of the nMCM. (S) indicate the screw holes for mounting on the PPM. The nMCM is slightly longer than the MCM.

Photographs of the original (top) and new (bottom) Multichip Module. The highlighted components are: (A) the PHOS4 chip, (B) the FADCs, (C) the PPrASIC and (D) the LVDS transmitters on the MCM; (1) the signal generator circuit, (2) the operational amplifiers, (3) the FADCs, (4) the CALIPPR FPGA (Xilinx Spartan-6) and (5) the power converters on the nMCM. The EEPROM device is mounted on the far side of the nMCM. (S) indicate the screw holes for mounting on the PPM. The nMCM is slightly longer than the MCM.

: Temperatures of all nMCMs in a PPr crate. This measurement was performed in a test crate, operating all nMCMs in playback mode (cf.\ Section~\ref{sec:nmcm_fw}) to run stress patterns. Each rectangle represents one nMCM, each column a fully loaded PPM. Circular fans at the bottom of the crate force a high-flux air stream upwards and are centred around slots 6 and 16, covering slots 3--9 and 13--19, respectively.

Temperatures of all nMCMs in a PPr crate. This measurement was performed in a test crate, operating all nMCMs in playback mode (cf.\ Section~\ref{sec:nmcm_fw}) to run stress patterns. Each rectangle represents one nMCM, each column a fully loaded PPM. Circular fans at the bottom of the crate force a high-flux air stream upwards and are centred around slots 6 and 16, covering slots 3--9 and 13--19, respectively.

Temperatures of all nMCMs in a PPr crate. This measurement was performed in a test crate, operating all nMCMs in playback mode (cf.\ Section~\ref{sec:nmcm_fw}) to run stress patterns. Each rectangle represents one nMCM, each column a fully loaded PPM. Circular fans at the bottom of the crate force a high-flux air stream upwards and are centred around slots 6 and 16, covering slots 3--9 and 13--19, respectively.

Overview of the CALIPPR firmware design.

Overview of the CALIPPR firmware design.

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

(a) The average noise for all channels in a PPM with no input cables connected, for the two cases of a PPM fully equipped with MCMs or nMCMs. For the MCM, two of the four channels on each module are noisier, leading to two separate curves. (b) The noise averaged over all channels in a PPM measured with different fine timing settings, for both the MCM and the nMCM. It should be noted that the nMCM implementation uses one less fine-timing step than the MCM (see Section~\ref{sec:performance_timing}).

$\eta$--$\phi$-maps of the total noise in all EM PPr channels in ADC counts, for the system equipped with (a) MCMs and (b) nMCMs.

$\eta$--$\phi$-maps of the total noise in all EM PPr channels in ADC counts, for the system equipped with (a) MCMs and (b) nMCMs.

$\eta$--$\phi$-maps of the total noise in all EM PPr channels in ADC counts, for the system equipped with (a) MCMs and (b) nMCMs.

$\eta$--$\phi$-maps of the total noise in all EM PPr channels in ADC counts, for the system equipped with (a) MCMs and (b) nMCMs.

Linear correlation coefficients as measured from noise for all possible pairs of channels on a PPM: (a) the result for a PPM equipped with MCMs, (b) the result for the same PPM equipped with nMCMs.

:

Linear correlation coefficients as measured from noise for all possible pairs of channels on a PPM: (a) the result for a PPM equipped with MCMs, (b) the result for the same PPM equipped with nMCMs.

Linear correlation coefficients as measured from noise for all possible pairs of channels on a PPM: (a) the result for a PPM equipped with MCMs, (b) the result for the same PPM equipped with nMCMs.

: Linear correlation coefficients as measured from noise for all possible pairs of channels on a PPM: (a) the result for a PPM equipped with MCMs, (b) the result for the same PPM equipped with nMCMs.

Linear correlation coefficients as measured from noise for all possible pairs of channels on a PPM: (a) the result for a PPM equipped with MCMs, (b) the result for the same PPM equipped with nMCMs.

(a) The scale factors for the width of the rising edge of the pulse, $\sigma_\textrm{rise}$, as function of $|\eta|$. (b) The different timing delay distributions obtained from fits to EM TT pulses in the very central $\eta$ region. Both plots compare the results from the Run~1 and Run~2 methods.

(a) The scale factors for the width of the rising edge of the pulse, $\sigma_\textrm{rise}$, as function of $|\eta|$. (b) The different timing delay distributions obtained from fits to EM TT pulses in the very central $\eta$ region. Both plots compare the results from the Run~1 and Run~2 methods.

(a) The scale factors for the width of the rising edge of the pulse, $\sigma_\textrm{rise}$, as function of $|\eta|$. (b) The different timing delay distributions obtained from fits to EM TT pulses in the very central $\eta$ region. Both plots compare the results from the Run~1 and Run~2 methods.

(a) The scale factors for the width of the rising edge of the pulse, $\sigma_\textrm{rise}$, as function of $|\eta|$. (b) The different timing delay distributions obtained from fits to EM TT pulses in the very central $\eta$ region. Both plots compare the results from the Run~1 and Run~2 methods.

Filter coefficient values for the EM calorimeter, normalised to the central coefficient, as a function of $|\eta|$: (a) the matched filter, (b) the AC filter scheme.

Filter coefficient values for the EM calorimeter, normalised to the central coefficient, as a function of $|\eta|$: (a) the matched filter, (b) the AC filter scheme.

Filter coefficient values for the EM calorimeter, normalised to the central coefficient, as a function of $|\eta|$: (a) the matched filter, (b) the AC filter scheme.

Filter coefficient values for the EM calorimeter, normalised to the central coefficient, as a function of $|\eta|$: (a) the matched filter, (b) the AC filter scheme.

The efficiency of the Peak-Finder algorithm identifying the correct BC for both matched and AC filters in Run~2 as a function of transverse energy, for (a) the EM barrel and (b) the EM FCAL. The BCID efficiency is defined with respect to the transverse energy, $E_{\textrm{T}}^{\textrm{calo}}$, provided by the calorimeters for the respective BC.

The efficiency of the Peak-Finder algorithm identifying the correct BC for both matched and AC filters in Run~2 as a function of transverse energy, for (a) the EM barrel and (b) the EM FCAL. The BCID efficiency is defined with respect to the transverse energy, $E_{\textrm{T}}^{\textrm{calo}}$, provided by the calorimeters for the respective BC.

The efficiency of the Peak-Finder algorithm identifying the correct BC for both matched and AC filters in Run~2 as a function of transverse energy, for (a) the EM barrel and (b) the EM FCAL. The BCID efficiency is defined with respect to the transverse energy, $E_{\textrm{T}}^{\textrm{calo}}$, provided by the calorimeters for the respective BC.

The efficiency of the Peak-Finder algorithm identifying the correct BC for both matched and AC filters in Run~2 as a function of transverse energy, for (a) the EM barrel and (b) the EM FCAL. The BCID efficiency is defined with respect to the transverse energy, $E_{\textrm{T}}^{\textrm{calo}}$, provided by the calorimeters for the respective BC.

:

Filter coefficient values, normalised to the central coefficient, for (a) TTs in the central $|\eta|$ region of the EMB and (b) TTs in the most forward $|\eta|$ region of FCAL1. Shown are the five coefficients derived from six different data samples with varying values of $\langle\mu\rangle$, one of which was recorded during LHC operation using the 8b4e filling scheme.

Filter coefficient values, normalised to the central coefficient, for (a) TTs in the central $|\eta|$ region of the EMB and (b) TTs in the most forward $|\eta|$ region of FCAL1. Shown are the five coefficients derived from six different data samples with varying values of $\langle\mu\rangle$, one of which was recorded during LHC operation using the 8b4e filling scheme.

: Filter coefficient values, normalised to the central coefficient, for (a) TTs in the central $|\eta|$ region of the EMB and (b) TTs in the most forward $|\eta|$ region of FCAL1. Shown are the five coefficients derived from six different data samples with varying values of $\langle\mu\rangle$, one of which was recorded during LHC operation using the 8b4e filling scheme.

Filter coefficient values, normalised to the central coefficient, for (a) TTs in the central $|\eta|$ region of the EMB and (b) TTs in the most forward $|\eta|$ region of FCAL1. Shown are the five coefficients derived from six different data samples with varying values of $\langle\mu\rangle$, one of which was recorded during LHC operation using the 8b4e filling scheme.

Filter coefficient values, normalised to the central coefficient, for (a) TTs in the central $|\eta|$ region of the EMB and (b) TTs in the most forward $|\eta|$ region of FCAL1. Shown are the five coefficients derived from six different data samples with varying values of $\langle\mu\rangle$, one of which was recorded during LHC operation using the 8b4e filling scheme.

The mean shift of the pedestal in the EM FCAL relative to its value in the absence of beams as a function of the BC Number in the LHC orbit. Results are shown for different $\langle\mu\rangle$ values, corresponding to different luminosity points in the same run. The shaded areas indicate BCs with colliding bunches in representative parts of the full LHC orbit of 3564 BCs. Results are shown for (a) an LHC filling scheme containing long bunch-trains and (b) an LHC fill using the 8b4e filling scheme.

The mean shift of the pedestal in the EM FCAL relative to its value in the absence of beams as a function of the BC Number in the LHC orbit. Results are shown for different $\langle\mu\rangle$ values, corresponding to different luminosity points in the same run. The shaded areas indicate BCs with colliding bunches in representative parts of the full LHC orbit of 3564 BCs. Results are shown for (a) an LHC filling scheme containing long bunch-trains and (b) an LHC fill using the 8b4e filling scheme.

The mean shift of the pedestal in the EM FCAL relative to its value in the absence of beams as a function of the BC Number in the LHC orbit. Results are shown for different $\langle\mu\rangle$ values, corresponding to different luminosity points in the same run. The shaded areas indicate BCs with colliding bunches in representative parts of the full LHC orbit of 3564 BCs. Results are shown for (a) an LHC filling scheme containing long bunch-trains and (b) an LHC fill using the 8b4e filling scheme.

The mean shift of the pedestal in the EM FCAL relative to its value in the absence of beams as a function of the BC Number in the LHC orbit. Results are shown for different $\langle\mu\rangle$ values, corresponding to different luminosity points in the same run. The shaded areas indicate BCs with colliding bunches in representative parts of the full LHC orbit of 3564 BCs. Results are shown for (a) an LHC filling scheme containing long bunch-trains and (b) an LHC fill using the 8b4e filling scheme.

The rate of a L1 $E_{\textrm{T}}^{\textrm{miss}}$ trigger with a threshold of \SI{35}{\giga\electronvolt} as a function of instantaneous luminosity for two different runs at the beginning of 2015. One of the runs was taken without the pedestal correction, while the second run has the pedestal correction enabled, resulting in an almost linear behaviour of the rate. The efficiency of the $E_{\textrm{T}}^{\textrm{miss}}$ trigger is unaffected by this change. The figure is taken from Ref.~\cite{TRIG-2016-01}.

The rate of a L1 $E_{\textrm{T}}^{\textrm{miss}}$ trigger with a threshold of \SI{35}{\giga\electronvolt} as a function of instantaneous luminosity for two different runs at the beginning of 2015. One of the runs was taken without the pedestal correction, while the second run has the pedestal correction enabled, resulting in an almost linear behaviour of the rate. The efficiency of the $E_{\textrm{T}}^{\textrm{miss}}$ trigger is unaffected by this change. The figure is taken from Ref.~\cite{TRIG-2016-01}.

The rate of a L1 $E_{\textrm{T}}^{\textrm{miss}}$ trigger with a threshold of \SI{35}{\giga\electronvolt} as a function of instantaneous luminosity for two different runs at the beginning of 2015. One of the runs was taken without the pedestal correction, while the second run has the pedestal correction enabled, resulting in an almost linear behaviour of the rate. The efficiency of the $E_{\textrm{T}}^{\textrm{miss}}$ trigger is unaffected by this change. The figure is taken from Ref.~\cite{TRIG-2016-01}.

Illustration of the four cases of the Sat80BCID algorithm. Solid lines represent samples digitised at \SI{40}{\mega\hertz}, while the dashed lines are additional samples digitised at \SI{80}{\mega\hertz}. For the \SI{80}{\mega\hertz} samples, only those used in the algorithm are shown.

Illustration of the four cases of the Sat80BCID algorithm. Solid lines represent samples digitised at \SI{40}{\mega\hertz}, while the dashed lines are additional samples digitised at \SI{80}{\mega\hertz}. For the \SI{80}{\mega\hertz} samples, only those used in the algorithm are shown.

Illustration of the four cases of the Sat80BCID algorithm. Solid lines represent samples digitised at \SI{40}{\mega\hertz}, while the dashed lines are additional samples digitised at \SI{80}{\mega\hertz}. For the \SI{80}{\mega\hertz} samples, only those used in the algorithm are shown.

Calibration of the Sat80BCID algorithm: (a) the behaviour of the ADC samples from the rising edge of pulses in an EMB TT as function of $E_{\textrm{T}}$ as measured in the calorimeters. The results of linear fits for each of the samples are indicated by lines of the respective colour. (b) The threshold values used in the EM calorimeter.

Calibration of the Sat80BCID algorithm: (a) the behaviour of the ADC samples from the rising edge of pulses in an EMB TT as function of $E_{\textrm{T}}$ as measured in the calorimeters. The results of linear fits for each of the samples are indicated by lines of the respective colour. (b) The threshold values used in the EM calorimeter.

Calibration of the Sat80BCID algorithm: (a) the behaviour of the ADC samples from the rising edge of pulses in an EMB TT as function of $E_{\textrm{T}}$ as measured in the calorimeters. The results of linear fits for each of the samples are indicated by lines of the respective colour. (b) The threshold values used in the EM calorimeter.

Calibration of the Sat80BCID algorithm: (a) the behaviour of the ADC samples from the rising edge of pulses in an EMB TT as function of $E_{\textrm{T}}$ as measured in the calorimeters. The results of linear fits for each of the samples are indicated by lines of the respective colour. (b) The threshold values used in the EM calorimeter.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

Example of an event triggered one BC early by the L1Calo BCID algorithm. The figures show $\eta$--$\phi$ maps of the LUT output for all TTs in the hadronic layer for two consecutive BCs. The axis labels indicate both the real $\eta$ and $\phi$ coordinates as well as the corresponding integer bin numbers (cf.\ Section~\ref{sec:inputpath}). In the hadronic layer, one~LUT count corresponds to approximately \SI{1}{\giga\electronvolt} (cf.\ Section~\ref{sec:performance_lut}). The plots are taken from the automated monitoring of the Saturated BCID performance.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.

JEP-LUT noise thresholds used during Run~2 for (a) the EM layer and (b) the HAD layer as function of $|\eta|$, separately for individual data-taking periods. The given $\langle\mu\rangle$ values correspond to peak average numbers of $pp$ interactions per bunch crossing. Artificial $|\eta|$ values (3.4, 3.8, 4.2, 4.65) have been assigned to the four hadronic FCAL towers, which in reality cover twice the $\eta$ range but are situated behind each other.