Pixel Operation Plots 2018

Contact: cms-tracker-runcoord@cernNOSPAMPLEASE.ch

Leakage Current Distribution in BPIX Thermal Mock-Up Measurements

Link to results

Show Leakage Current Distribution in BPIX Thermal Mock-Up Measurements Show Leakage Current Distribution in BPIX Thermal Mock-Up Measurements

This page shows a characterization of the high-voltage leakage currents and the thermal behavior in BPIX using a thermal mockup. In the two-phase CO2 cooling system, a temperature drop in the order of 4K occurs along the cooling line (expected behavior). Sensor leakage current is temperature dependent and thus varies along one cooling loop. This effect is seen in the real detector and was confirmed with measurements in a BPIX thermal mockup. Leakage current distributions are visualized using 2D-maps in the z-phi plane. The coordinate system chosen here is not identical to the CMS coordinate system but is chosen here motivated by the numbering of the power sectors.

Measurements with a thermal mock-up were performed. The mechanics of the mockup are identical to one BPIX Layer 2 half-shell. A 2-phase CO2 cooling with the LUKASZ plant in the BPIX clean room was used.

Cooling lines enter the BPIX detector either from the +z-end or from the -z-end. One cooling line cools up to eight ladders (eight modules per ladder) and leave the detector at the same end where it came from.

Figure in png format	other formats	Description
	.pdf	BPIX power sectors layer 1: The maps show the BPIX power sectors for layer 1 in the phi-z-plane. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Dashed lines indicate individual modules. Arrows indicate the inlet and outlet of the 2-phase CO 2 cooling lines.
	.pdf	BPIX power sectors layer 2: The maps show the BPIX power sectors for layer 2 in the phi-z-plane. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Full lines indicate the power sectors, dashed lines indicate individual modules. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines.
	.pdf	BPIX power sectors layer 3: The maps show the BPIX power sectors for layer 3 in the phi-z-plane. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Full lines indicate the power sectors, dashed lines indicate individual modules. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines.
	.pdf	BPIX power sectors layer 4: The maps show the BPIX power sectors for layer 4 in the phi-z-plane. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Full lines indicate the power sectors, dashed lines indicate individual modules. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines.
	.pdf	BPIX leakage currents layer 1: The maps show the BPIX power sectors. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Normal lines indicate the power sectors. Dashed lines indicate individual modules. The values in the map are: sector leakage current divided by the number of connected modules. For sectors (or modules) which are not powered, the corresponding bin is left blank. Modules in Layer 1 marked with a white asterisk have been exchanged during YETS 2017/18 and have less integrated luminosity. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines. The plots show that the leakage current decreases along one cooling loop: a temperature drop in CO2 cooling causes lower a leakage current towards the outlet.
	.pdf	BPIX leakage currents layer 2: The maps show the BPIX power sectors. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Normal lines indicate the power sectors. Dashed lines indicate individual modules. The values in the map are: sector leakage current divided by the number of connected modules. For sectors (or modules) which are not powered, the corresponding bin is left blank. Modules in Layer 1 marked with a white asterisk have been exchanged during YETS 2017/18 and have less integrated luminosity. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines. The plots show that the leakage current decreases along one cooling loop: a temperature drop in CO2 cooling causes lower a leakage current towards the outlet.
	.pdf	BPIX leakage currents layer 3: The maps show the BPIX power sectors. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Normal lines indicate the power sectors. Dashed lines indicate individual modules. The values in the map are: sector leakage current divided by the number of connected modules. For sectors (or modules) which are not powered, the corresponding bin is left blank. Modules in Layer 1 marked with a white asterisk have been exchanged during YETS 2017/18 and have less integrated luminosity. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines. The plots show that the leakage current decreases along one cooling loop: a temperature drop in CO2 cooling causes lower a leakage current towards the outlet.
	.pdf	BPIX leakage currents layer 4: The maps show the BPIX power sectors. Thick black lines indicate the four quadrants of BPIX (top-left: BmI, top-right: BpI, bottom-left: BmO, bottom-right: BpO). Normal lines indicate the power sectors. Dashed lines indicate individual modules. The values in the map are: sector leakage current divided by the number of connected modules. For sectors (or modules) which are not powered, the corresponding bin is left blank. Modules in Layer 1 marked with a white asterisk have been exchanged during YETS 2017/18 and have less integrated luminosity. Arrows indicate the inlet and outlet of the 2-phase CO2 cooling lines. The plots show that the leakage current decreases along one cooling loop: a temperature drop in CO2 cooling causes lower a leakage current towards the outlet.
	.pdf	BPix leakage currents vs phi in layer 1: same data as above. Red markers: positive z-end-sectors. Blue markers: negative z-end- sectors. Error bars in x indicate the phi-coverage of the sector. Dashed lines indicate the inlet or outlet of a cooling line. Gray arrows indicate the CO2 flow direction.
	.pdf	BPix leakage currents vs phi in layer 2: same data as above. Red markers: positive z-end-sectors. Blue markers: negative z-end- sectors. Error bars in x indicate the phi-coverage of the sector. Dashed lines indicate the inlet or outlet of a cooling line. Gray arrows indicate the CO2 flow direction.
	.pdf	BPix leakage currents vs phi in layer 3: same data as above. Red markers: positive z-end-sectors. Blue markers: negative z-end- sectors. Error bars in x indicate the phi-coverage of the sector. Dashed lines indicate the inlet or outlet of a cooling line. Gray arrows indicate the CO2 flow direction.
	.pdf	BPix leakage currents vs phi in layer 4: same data as above. Red markers: positive z-end-sectors. Blue markers: negative z-end- sectors. Error bars in x indicate the phi-coverage of the sector. Dashed lines indicate the inlet or outlet of a cooling line. Gray arrows indicate the CO2 flow direction.
	.pdf	BPix thermal mockup temperatures: The plot shows the temperature values measured with the BPIX thermal mock-up. Measurement conditions: module power 3.1W, nominal CO temperature -22.5°C, ambient temperature -15°C, preheating power 30W. Arrows indicate the inlet and outlet of the 2-phase CO 2 cooling lines. Module temperatures decrease along one cooling loop causing a temperature drop in the 2-phase CO2 cooling. Some irregularities appear which have been cross-checked many times. It is still not clear what is causing them. The white bin is a broken dummy module. Conclusion of the plot: The modules have lower temperatures towards the return of the cooling loop. The module temperature in the detector is not uniformly distributed, but determined by the CO2 flow and its 2-phase behavior.
	.pdf	BPix thermal mockup temperatures: The mean module temperature is plotted versus the average phi-position of the modules for different mass flows of 2-phase CO2 cooling. Measurement conditions: module power 3.1W, nominal CO temperature -22.5°C, ambient temperature -15°C, preheating power 30W. The plot shows that the overall temperature difference in the mock-up half-shell is about 1.5K smaller if the CO2 mass flow is lowered from 2.5 g/s to 1.5 g/s. Conclusion of the plot: The effect observed is due to the properties of the CO2 in two-phase state, e.g. the velocity of mass flow, friction, boiling behavior. While the module temperature at the return point (phi ~= 90°) stays nearly constant, it is lower for modules which are closer to the inlet of the cooling loop (phi ~= 0°). Reducing the CO 2 mass flow might thus reduce the module temperatures and leakage currents (and their spreads with respect to the whole BPIX).
	.pdf	BPix thermal mockup temperatures: The plot shows the temperature values measured with the BPIX Thermal Mock-up for different module powers. It is estimated that the BPIX Layer 2 modules currently have a power of ~ 3W. A module power of 4.8 W can be considered as an upper limit for the end-of-life-time power. The mean module temperature is plotted versus the average phi-position of the modules for different module powers at a CO2 mass flow of 2.5g/s. Conclusion of the plot: A significant higher module power also affects the module temperatures (6K higher module temperatures at 2W higher power. The overall temperature difference in the mock-up half-shell is increased from 4K for a module power of 2.7W to 6K for a module power of 4.8W. The temperature differences in the detector do also depend on the full heat load.
	.pdf	BPix thermal mockup temperatures: The modules are arranged into groups of four modules per ladder and z-end. The factor of HV leakage current per module is normalized to the lowest module temperature T0 that is measured in the mock-up. Then the z-axis is normalized to the module group with the lowest leakage current. Conclusion of the plot: The plot shows that a factor of 1.9 between leakage currents at different ladders is expected from the measurements with the thermal mock-up. These leakage current factors which are estimated from the mock-up measurements, were compared to the detector currents where the location of the cooling was taken into account. The results are in good agreement with the actually measured leakage currents in the BPIX detector.

-- JorineMirjamSonneveld - 2019-01-29

Development of the CMS Phase-1 Pixel Online Monitoring System and the Evolution of Pixel Leakage Current

Link to results

Show Development of the CMS Phase-1 Pixel Online Monitoring System and the Evolution of Pixel Leakage Current Show Development of the CMS Phase-1 Pixel Online Monitoring System and the Evolution of Pixel Leakage Current

This page explains the development of Phase-1 pixel online monitoring system and its representative features. It also illustrates the distribution of temperature within different layers/disks of the pixel detector. It contains module leakage current evolution of both pixel barrel and endcap through the whole phase-1 operation period (2017 ~ 2018). The picture below is the panel of the monitoring interface, which is a webpage for visualizing and monitoring the following parameters online and offline. It consists of many drop-down menus for users to access and plot different detector related parameters. There are also links to other relevant websites or tools. The function of this system is to centralize and correlate detector information to have a good overview on the detector performance, and also to provide users with an easier access.

The summary of the categories of variables that this tool provides are listed below:

Environment variables	Detector parameters	CMS Run variable property
Dew point	Power supply voltage	Luminosity
Air pressure	Power group current	Detector run status
Air temperature	Module temperature	Data acquisition status
Himidity	Cooling flow status	Data quality monitoring
...	...	...

Figure in png format	other formats	Description
	null	The trend of instantaneous luminosity & HV current of sector 1 in layer 3 of Pixel Barrel (BpI), during a normal LHC fill 7320 (CMS Run 324968, CMS Run 324970) (2018.10.19 14:49 — 2018.11.20 05:58). The HV current (leakage current) was dropping through the decreasing instantaneous luminosity. Emittance scan took place after the stable beam of p-p collision established , which leaded to some fluctuations of luminosity and leakage current. At the end of the fill, pixel HV went off (STANDBY mode).
	null	The digital occupancy of pixel layer 4 during CMS Run Number 322013 (2018.08.31). There are four half cylinders, each cylinder has 32 ladders, and each ladder has 4 modules, and each module has 8 readout chips. In the plot, one bin corresponds to one readout chip (ROC). Every red marked rectangle represent a region recorded with entries in database of known problems (keep track).
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 1) during cosmic rays. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 320448 (Cosmic run on Jul.28) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 1 (totally 4 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 1) during pp collisions. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 322625 (stable beam run on Sep.10) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 1 (totally 4 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 2) during cosmic rays. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 320448 (Cosmic run on Jul.28) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 2 (totally 4 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 2) during pp collisions. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 322625 (stable beam run on Sep.10) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 2 (totally 4 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 3) during cosmic rays. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 320448 (Cosmic run on Jul.28) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 3 (totally 8 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 3) during pp collisions. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 322625 (stable beam run on Sep.10) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 3 (totally 8 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet Empty units: no valid reading
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 4) during cosmic rays. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 320448 (Cosmic run on Jul.28) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 4 (totally 8 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet
	pdf	Pixel barrel temperature gradient along each cooling loop (layer 4) during pp collisions. Each cooling loop has three temperature probes, which are located respectively at the beginning (inlet), middle, end (outlet) positions These groups of temperature were obtained during CMS Run Number 322625 (stable beam run on Sep.10) It shows the gradient of the temperature along each cooling loop of Pixel Barrel layer 4 (totally 8 cooling loops) As expected for CO2 cooling, the temperature at the outlet is lower than at the inlet
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 1) Temperature measured with cosmic rays The azimuthal coordinate yields the CMS coordinates Layer 1 has 4 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings -- excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 1) Temperature measured with pp beams The azimuthal coordinate yields the CMS coordinates Layer 1 has 4 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings -- excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 2) Temperature measured with cosmic rays The azimuthal coordinate yields the CMS coordinates Layer 2 has 4 cooling loops (black arrows pointing to the directions of cooling flows), each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings — excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 2) Temperature measured with pp beams The azimuthal coordinate yields the CMS coordinates Layer 2 has 4 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings — excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 3) Temperature measured with cosmic rays The azimuthal coordinate yields the CMS coordinates Layer 3 has 8 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings — excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 3) Temperature measured with pp beams The azimuthal coordinate yields the CMS coordinates Layer 3 has 8 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes (few of them give invalid readings — excluded from the plots) Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 4) Temperature measured with cosmic rays The azimuthal coordinate yields the CMS coordinates Layer 4 has 8 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel barrel temperature w.r.t azimuthal coordinate (layer 4) Temperature measured with pp beams The azimuthal coordinate yields the CMS coordinates Layer 4 has 8 cooling loops, each of which covers approximately one quadrant in azimuthal plane Each cooling loop has three temperature probes Assume each temperature probe occupies the one third of the azimuthal plane coverage of each cooling loop As a result of the 2-phase state of CO2 cooling flow, decreased CO2 flow leads to its absorbing heat more sufficiently, resulting in more efficient cooling, lower temperature, less temperature spread
	pdf	Pixel endcap leakage current distribution Pixel endcap detector consists of two endcaps or cylinders Each half cylinder is a quadrant with 3 disks HV currents were measured at 10 minutes after stable beam declared during LHC nominal fill 7144 (Sep.9th) Currents were normalized by the number of connected readout chips (ROC) for each power group Each cylinder consists of 2 rings, and modules in RING1 are closer to the beam than those in RING2, so higher leakage current is observed in RING1 than in RING2 The module leakage current distribution in the same ring is roughly uniform Note: a power group in disk 1 has significant high current that has been seen since 2017, to be investigated during the long shut down 2 (LS2)
	pdf	Pixel endcap leakage current distribution Pixel endcap detector consists of two endcaps or cylinders Each half cylinder is a quadrant with 3 disks HV currents were measured at 10 minutes after stable beam declared during LHC nominal fill 7144 (Sep.9th) Currents were normalized by the number of connected readout chips (ROC) for each power group Each cylinder consists of 2 rings, and modules in RING1 are closer to the beam than those in RING2, so higher leakage current is observed in RING1 than in RING2 The module leakage current distribution in the same ring is roughly uniform Note: a power group in disk 1 has significant high current that has been seen since 2017, to be investigated during the long shut down 2 (LS2)
	pdf	Pixel endcap leakage current distribution Pixel endcap detector consists of two endcaps or cylinders Each half cylinder is a quadrant with 3 disks HV currents were measured at 10 minutes after stable beam declared during LHC nominal fill 7144 (Sep.9th) Currents were normalized by the number of connected readout chips (ROC) for each power group Each cylinder consists of 2 rings, and modules in RING1 are closer to the beam than those in RING2, so higher leakage current is observed in RING1 than in RING2 The module leakage current distribution in the same ring is roughly uniform Note: a power group in disk 1 has significant high current that has been seen since 2017, to be investigated during the long shut down 2 (LS2)
	pdf	Pixel barrel module leakage current evolution LHC fills from beginning of 2017 until end of October in 2018 data-taking are employed (proton-proton collisions) Currents measured within 20 minutes from Stable Beam declaration Average current per pixel module measured from power groups (no temperature correction) Leakage current increased gradually due to accumulated radiation dose through the year Closer to beam spot -> more accumulated radiation dose -> higher leakage current (layer 1 > layer 2 > layer 3 > layer 4) There are some drops of leakage current from the global trend because of: Annealing during Machine development or technical stop period Power supply replacement HV setting change
	pdf	Pixel endcap module leakage current evolution LHC fills from beginning of 2017 until end of October in 2018 data-taking are employed (proton-proton collisions) Currents measured within 20 minutes from Stable Beam declaration Average current per pixel module measured from power groups (no temperature correction) Note: The 4th power group giving much higher current in disk 1 (seen in slide 25) is removed from the average Leakage current increased gradually due to accumulated radiation dose through the year Closer to beam spot -> more accumulated radiation dose -> higher leakage current (ring 1 > ring 2) There are some drops of leakage current from the global trend because of: Annealing during Machine development or technical stop period Power supply replacement HV setting change

Phase-I Pixel DAQ Configuration Times in 2018

Link to results

Show Phase-I Pixel DAQ Configuration Times in 2018 Show Phase-I Pixel DAQ Configuration Times in 2018

Pixel Barrel Radiation Damage Phase-1 - Leakage Currents and Depletion Voltages

Link to results

Show Phase-I Pixel Barrel Radiation Damage Phase-1 - Leakage Currents and Depletion Voltages Show Phase-I Pixel Barrel Radiation Damage Phase-1 - Leakage Currents and Depletion Voltages

General Introducion

Illustration of the principle:

The Hamburg model ( Hamburg Model and Plots: M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999. Remake plots taken from C. Barth, Physikalische Analyse des Ansprechverhaltens des CMS Siliziumdetektors beim Betrieb am LHC, Karlsruher Institut für Technologie, 2013)

is used to calculate the depletion voltage. It is time-/temperature-dependent and fluence-dependent. An empirical model for leakage current is used:

The expected leakage current in each of the pixel barrel layers is calculated based on the full temperature- and irradiation history using the empirical radiation damage model I (Φ, t, T) = I0 + α( t), T (Φ, V). Radiation-induced increase of leakage current depends on fluence Φ, time t, temperature T, volume V. The α-parameter set for radiation damage model used comes from M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999 https://mmoll.web.cern.ch/mmoll/thesis/. For the simulation, a FLUKA fluence simulation** was used with high granular resolution and detector geometry as input.

The data for comparison is from the CMS pixel detector with an available granularity per sector: there are several modules per sector, the data is not resolved in z.

Current Version

Figure in png format	other formats	Description

Superseded Versions

Show version of plots from 10/2018 Show version of plots from 10/2018

The pixel cooling set point is at -22 °C. For the silicon module temperature estimation, measurements from carbon fibre (near cooling loop) are used taking the average of all sensors and average over a day. Temperature is around -11.5 °C when low voltage is on. Actual silicon temperature is higher than measurements near cooling loop when module power (low voltage) is on. Estimation from mock-up layer 2 in laboratory: + 2-3 K offset to carbon fibre measurements. Temperatures used are:

Figure in png format	other formats	Description
	.pdf	The expected leakage current in each of the pixel barrel layers is calculated based on the full temperature- and irradiation history using the empirical radiation damage model I (Φ, t, T) = I0 + α( t), T (Φ, V). Radiation-induced increase of leakage current depends on fluence Φ, time t, temperature T, volume V. The α-parameter set for radiation damage model used comes from M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999 https://mmoll.web.cern.ch/mmoll/thesis/. For the simulation, a FLUKA fluence simulation** was used with high granular resolution and detector geometry as input. Actual temperature history is taken from the database** adding an offset** (L1 Silicon at -8.5 ± 2 °C). Whenever the low voltage is on, the silicon temperature is greater than the measurement near cooling loop. * these parameters introduce the main uncertainties. The leakage current simulation for layer 1 is corrected by factor 1.3. The data for comparison is from the CMS pixel detector with an available granularity per sector: there are several modules per sector, the data is not resolved in z.
	.pdf	The expected leakage current in each of the pixel barrel layers is calculated based on the full temperature- and irradiation history using the empirical radiation damage model I (Φ, t, T) = I0 + α( t), T (Φ, V). Radiation-induced increase of leakage current depends on fluence Φ, time t, temperature T, volume V. The α-parameter set for radiation damage model used comes from M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999 https://mmoll.web.cern.ch/mmoll/thesis/. For the simulation, a FLUKA fluence simulation* was used with high granular resolution and detector geometry as input. Actual temperature history is taken from the database** adding an offset** (L2 Silicon at -8.5 ± 2 °C). Whenever the low voltage is on, the silicon temperature is greater than the measurement near cooling loop. * these parameters introduce the main uncertainties. The leakage current simulation for layer 2 is corrected by factor 2.4. The data for comparison is from the CMS pixel detector with an available granularity per sector: there are several modules per sector, the data is not resolved in z.
	.pdf	The expected leakage current in each of the pixel barrel layers is calculated based on the full temperature- and irradiation history using the empirical radiation damage model I (Φ, t, T) = I0 + α( t), T (Φ, V). Radiation-induced increase of leakage current depends on fluence Φ, time t, temperature T, volume V. The α-parameter set for radiation damage model used comes from M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999 https://mmoll.web.cern.ch/mmoll/thesis/. For the simulation, a FLUKA fluence simulation* was used with high granular resolution and detector geometry as input. Actual temperature history is taken from the database** adding an offset** (L3 Silicon at -8.5 ± 2 °C). Whenever the low voltage is on, the silicon temperature is greater than the measurement near cooling loop. * these parameters introduce the main uncertainties. The leakage current simulation for layer 3 is corrected by factor 1.8. The data for comparison is from the CMS pixel detector with an available granularity per sector: there are several modules per sector, the data is not resolved in z.
	.pdf	The expected leakage current in each of the pixel barrel layers is calculated based on the full temperature- and irradiation history using the empirical radiation damage model I (Φ, t, T) = I0 + α( t), T (Φ, V). Radiation-induced increase of leakage current depends on fluence Φ, time t, temperature T, volume V. The α-parameter set for radiation damage model used comes from M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999 https://mmoll.web.cern.ch/mmoll/thesis/. For the simulation, a FLUKA fluence simulation* was used with high granular resolution and detector geometry as input. Actual temperature history is taken from the database** adding an offset** (L4 Silicon at -7.5 ± 2 °C). Whenever the low voltage is on, the silicon temperature is greater than the measurement near cooling loop. ** these parameters introduce the main uncertainties. The leakage current simulation for layer 4 is corrected by factor 1.8. The data for comparison is from the CMS pixel detector with an available granularity per sector: there are several modules per sector, the data is not resolved in z.
	.pdf	Based on the full temperature- and irradiation history the expected full depletion voltages of the pixel tracker layers are simulated using the Hamburg model (M. Moll, Radiation Damage in Silicon Particle Detectors, Universität Hamburg, DESY-THESIS-1999-040, 1999) for radiation damage. Warm periods during various technical stops lead to a change of depletion voltage due to annealing. Simulation input: FLUKA fluence simulation** with high granular resolution and detector geometry, and for the material, the different impact of charged and neutral particles on oxygenated silicon are taken into account. In the Hamburg model, the Hamburg parameter set** for oxygenated Si (DOFZ) was used. The actual temperature history is taken from a database** where whenever the low voltage is on, the silicon temperature is greater than the measurement near cooling loop. Data points ** are taken from HV bias scans. Considering the high sensitivity to input data, the simulation matches the data well. **all of these parameters introduce significant uncertainties.

ROC Thresholds during Commissioning 2018

The following plots show the status of the pixel detector by the beginning of the run in 2018 and its evolution as of June 2018. The measured thresholds of the readout chips (ROCs) are shown for two periods; Commissioning 2018 (beginning of the year after 50fb-1 delivered LHC luminosity) and June 2018 (during TS1 additional 25fb-1 delivered LHC luminosity). An expected, slight degradation of the thresholds due to increased irradiation is observed for pixel barrel layers 2-4 and pixel forward, as the mean of these distribution is shifted towards smaller values in June compared to the beginning of the year. Pixel barrel layer 1 shows no significant degradation, as the radiation effect has already reached a plateau here. The distributions are broadened by radiation effects, more pronounced on the lower edge of the distributions.

Figure in png format	other formats	Description
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for all four layers of the pixel barrel detector normalized to the number of ROCs in the respective layer. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The distribution of layer 1 is different, because another type of ROC is used here. Modules in layer 2 that were damaged in 2017 and could not be replaced are shown in a separate distribution. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 1 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 2 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules that were damaged in 2017 and could not be replaced are shown in yellow, while the modules showing normal behavior are shown in blue. The damaged modules show a higher threshold. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 3 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 4 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for the whole pixel forward detector normalized to the number of ROCs in the respective ring. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules in the forward pixel that were damaged in 2017 and could not be replaced are shown in a separate distribution (small effect). The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for ring 1 of the pixel forward detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules that were damaged in 2017 and could not be replaced are shown in yellow, while the modules showing normal behavior are shown in blue. The damaged modules have a higher threshold. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for ring 2 of the pixel forward detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed during the commissioning phase in 2018. The number of electrons were calculated from the measured calibration units (VCal).

ROC Thresholds during TS1 (June) 2018

Figure in png format	other formats	Description
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for all four layers of the pixel barrel detector normalized to the number of ROCs in the respective layer. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The distribution of layer 1 is different, because another type of ROC is used here. Modules in layer 2 that were damaged in 2017 and could not be replaced are shown in a separate distribution. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs. The effect cannot be seen in layer 1, as the effect has already reached a plateau here for most ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 1 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). No degradation effect due to irradiation with respect to the measurement during commissioning is observed here, as for most ROCs a plateau of this effect is already reached.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 2 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules that were damaged in 2017 and could not be replaced are shown in yellow, while the modules showing normal behavior are shown in blue. The damaged modules have a higher threshold. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 3 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for layer 4 of the pixel barrel detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for the whole pixel forward detector normalized to the number of ROCs in the respective ring. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules in the forward pixel that were damaged in 2017 and could not be replaced are shown in a separate distribution (small effect). The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for ring 1 of the pixel forward detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. Modules that were damaged in 2017 and could not be replaced are shown in yellow (small effect), while the modules showing normal behavior are shown in blue. The damaged modules have a higher threshold. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.
	.eps .pdf	The distribution of the readout chip (ROC) threshold in number of electrons for ring 2 of the pixel forward detector. The threshold is measured as the value where the turn-on of the ROC reaches an efficiency of 0.5. The shown measurement was performed in June 2018. The number of electrons were calculated from the measured calibration units (VCal). A minimal degradation of the threshold with respect to the measurement during commissioning is observed. This effect is due to the increased irradiation of the ROCs.

DCDC Characterization

In October 2017 DCDC converters in the CMS Pixel started failing. At the end of 2017 data taking 5% of the DCDCs did not provide power to the modules. In the YETS the Pixel was extracted and the DCDC converters were examined. During the YETS examination 30% of the DCDC where found to have a higher current than expected. The preliminary findings were reported at the ACES conference After this report additional findings suggest a failure mechanism that is only present in a radiated DCDC ASIC. In the disabled state one of the radiated transistors has a leakage current, which is amplified by the circuit and not drained. This stress can then either break the DCDC ASIC or make it a high current DCDC. In the enabled state the current is drained.

This material shows the results of the investigations during the YETS. First the characterization of the DCDC is shown and then the distribution of the DCDCs in the detector. The Pixel DCDC are used to provide the low voltage to the analog and the digital part of the modules, therefore the DCDCs have different output voltages 2.4-2.5V for the analog and 3.0 - 3.3V for the digital voltage.

Classification based on I-V curve measurement

Figure in png format	other formats	Description
	.eps .pdf	Two tests are performed to classify the DCDC converters as either working normally, having high-current or being broken. The first test is the disable test, where the DCDC converter is disabled and the input voltage is scanned from 0 V to 5.5 V (upper plot). The second test is the enable test, where the DCDC is enabled and a load of 1.5A is connected. The input voltage is then scanned from 0 V to 10 V. There is a difference in the behavior between a high-current DCDC and a normal working DCDC in the disabled test, while the converters show the same behavior in the enable test. A DCDC converter is considered as having high-current, when the input current at 5.5V is larger than 4mA. The behavior of a broken DCDC converter is not shown. A DCDC converter is considered broken, when the input current at 10V in the enabled test is less than 400mA.
	.eps .pdf	The input current of the DCDC converters that were extracted from the detector after the 2017 run. The current was measured at an input voltage of 5.5V while the DCDC is disabled. DCDCs with an input current less than 4mA are considered as working normally. DCDCs with an input current greater than 4mA are considered as high-current.
	.eps .pdf	The input current of the DCDC converters that were extracted from the detector after the 2017 run. The current was measured at an input voltage of 10V while the DCDC is enabled. A clear separation between the broken DCDC converters (input current less than 400 mA), the working analog converters (input current between 400 and 540 mA) and the working digital converters (input current greater 540 mA) is observed. While a DCDC converter is enabled no difference between the normal behavior and the high-current behavior is visible in this distribution.

Maps with DCDC classifications after run 2017

Pixel Barrel

Figure in png format	other formats	Description
	.pdf	Summary of the characterization of the 2017 DCDC converters for the barrel pixel detector. The number of normal behaving, high-current and broken DCDC converters is given for each half-cylinder separately and for the total barrel pixel detector. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in minus direction wrt. the beam on the inner side (BmI) of the barrel pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in minus direction wrt. the beam on the outer side (BmO) of the barrel pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in plus direction wrt. the beam on the inner side (BpI) of the barrel pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in plus direction wrt. the beam on the outer side (BpO) of the barrel pixel detector. The coordinates correspond to the position of the DCDC converter in the supply tube in the detector. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.

Pixel Forward

Figure in png format	other formats	Description
	.pdf	Summary of the characterization of the 2017 DCDC converters for the forward pixel detector. The number of normal behaving, high-current and broken DCDC converters is given for each half-cylinder separately and for the total forward pixel detector. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in minus direction wrt. the beam on the inner side (BmI) of the forward pixel detector. The converters are grouped into readout groups (ROG) and the coordinates correspond to the position in the detector. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in minus direction wrt. the beam on the outer side (BmO) of the forward pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in plus direction wrt. the beam on the inner side (BpI) of the forward pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.
	.eps .pdf	Overview of the DCDC converter classification after the 2017 run for the half-cylinder in plus direction wrt. the beam on the outer side (BpO) of the forward pixel detector. The coordinates correspond to the position of the DCDC converter in the detector during the run. Analog DCDC converter are marked with an “a”, digital DCDC converter are marked with a “d”. The defects of DCDCs are randomly distributed.