Skip to topic | Skip to bottom
ILC.GeometriaR-phiEnElVELODeLHCbr1.2 - 04 Feb 2009 - 09:06 - Main.iglesiastopic end

Start of topic | Skip to actions

The use of cylindrical geometry (rf coordinates), rather than a simpler rectilinear scheme, was chosen in order to enable fast reconstruction of tracks and vertices in the LHCb trigger. Indeed, simulations showed that 2D (rz) tracking allows a fast reconstruction in the HLT with sufficient impact parameter resolution to efficiently select events with b-hadrons. For this reason, an rf geometry was selected for the design.

Each VELO module was designed to provide the necessary 3D spatial information to reconstruct the tracks and vertices. One of the two sensors of the module, called the f-measuring sensor, or f-sensor, provides information on the azimuthal coordinate around the beam. The other sensor, called the r-measuring sensor, or R-sensor, provides information on the radial distance from the beam axis. The third coordinate is provided by knowledge of the position of each sensor plane within the experiment. The rz tracking requirement imposes the additional constraint that the VELO circular strips should be centered as perfectly as possible around the beam axis.



For both the R and f-sensors the minimum pitch is designed to be at the inner radius to optimize the vertex resolution. The conceptual layout of the strips on the sensors is illustrated in figure 5.4. For the R-sensor the diode implants are concentric semi-circles with their centre at the nominal LHC bem position. In order to minimize the occupancy each strip is subdivided into four 45_ regions. This al so has the beneficial effect of reducing the strip capacitance. The minimum pitch at the innermost radius is 38 mm, increasing linearly to 101.6 mm at the outer radius of 41.9 mm. This ensures that measurements along the track contribute to the impact parameter precision with roughly equal weight.


The f-sensor is designed to readout the orthogonal coordinate to the R-sensor. In the simplest possible design these strips would run radially from the inner to the outer radius and point at the nominal LHC beam position with the pitch increasing linearly with radius starting with a pitch of 35.5 mm. However, this would result in unacceptably high strip occupancies and too large a strip pitch at the outer edge of the sensor. Hence, the f-sensor is subdivided into two regions, inner and outer. The outer region starts at a radius of 17.25mm and its pitch is set to be roughly half (39.3 mm) that of the inner region (78.3 mm), which ends at the same radius.

The design of the strips in the f-sensor is complicated by the introduction of a skew to improve pattern recognition. At 8mm from the beam the inner strips have an angle of approximately 20_ to the radial whereas the outer strips make an angle of approximately 10_ to the radial at 17 mm. The skew of inner and outer sections is reversed giving the strips a distinctive dog-leg design. The modules are placed so that adjacent f-sensors have the opposite skew with respect to the each other. This ensures that adjacent stations are able to distinguish ghost hits from true hits through the use of a traditional stereo view. The principal characteristics of the VELO sensors are summarized in table 5.1.

The technology utilized in both the R- and f-sensors is otherwise identical. Both sets of sensors are 300 mm thick. Readout of both R- and f-sensors is at the outer radius and requires the use of a second layer of metal (a routing layer or double metal) isolated from the AC-coupled diode strips by approximately 3 mm of chemically vapour deposited (CVD) SiO2?. The second metal layer is connected to the first metal layer by wet etched vias. The strips are biased using polysilicon 1MW resistors and both detectors are protected by an implanted guard ring structure.

The pitch as a function of the radius r in mm increases linearly and is given by the following expressions:

The sensors were developed for high radiation tolerance. Early prototype detectors used p-stop isolation. This was later replaced by p-spray isolated detectors which showed much higher resistance to micro-discharges. The n+n design was compared with an almost geometrically identical p+n design and was shown to have much better radiation characteristics as measured by charge collection as a function of voltage


Algunas preguntas

porque dice que es mucho mas rápida la reconstrucción usando las coordenadas cilíndricas r-phi que las rectilíneas?

Porque con coordenadas cilíndricas podes facer unha reconstrucción 2D empregando unicamente un único tipo de sensores (máis a información da z de cada sensor, que xa a tes). De feito o que se fai é primeiro facer un fit so coas r e logo incorporarlle a información das phi. É dicir, o fit faise 2D e logo incorpórase a dimensión extra. Isto non se podería facer así empregando coordenadas cartesianas. Tes máis información sobre isto aquí:

a que se refiere con el diseño "dog-leg"?

Refírese a que os sensores en phi non son exactamente liñas rectas, senón que forman un certo ángulo (unha especie de "cóbado") nun certo valor de r. Se te fixas na figura 5.4 vese claramente.

como se definen exactamente R y phi dentro de cada sensor? Son simplemente as coordenadas cilíndricas de toda a vida. Consideras o punto (0,0) de (x,y) como a tua orixe. r sería o radio da circunferencia que pasa polo teu hit e está centrada na orixe. phi é o angulo axial: phi=atan(y/x), ou x=r*cos(phi),y=r*sin(phi)

Veo que aparacen R y phi en función de r, que son los demás valores de la formula?

Non, aí se les ben fálache do "pitch" en R e Phi. Esa é "finura" do detector, é dicir, á distancia entre cada strip, tanto en R como en phi. Se te fixas na figura 5.4 vese esa distancia é indícanse varios valores do pitch. No caso das R vese moi claramente como vai aumentando coa distancia ao centro (coordenada r)

como se definen R y phi en función de x, y z?

Isto é o de antes. O z é igual ao cartesiano.

-- Main.iglesias - 02 Feb 2009

to top

I Attachment sort Action Size Date Who Comment
r-phiasafunctionoftheradiusr.jpg manage 12.7 K 02 Feb 2009 - 16:00 Main.iglesias  
CrosssectioninxyoftheVELO.jpg manage 27.4 K 04 Feb 2009 - 09:01 Main.iglesias  
r-phygeometryoftheVELOsensors.jpg manage 31.7 K 04 Feb 2009 - 09:01 Main.iglesias  
InsideoftheRF_foil.jpg manage 22.7 K 04 Feb 2009 - 09:02 Main.iglesias  
CharacteristicsofVELOsensors.jpg manage 59.4 K 04 Feb 2009 - 09:02 Main.iglesias  

You are here: ILC > GeometriaR-phiEnElVELODeLHCb

to top

Copyright © 1999-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding this material Send feedback