25–29 Sept 2006
Valencia, Spain
Europe/Zurich timezone

State of the art technologies for front-end hybrids

27 Sept 2006, 09:45
25m
Valencia, Spain

Valencia, Spain

IFIC – Instituto de Fisica Corpuscular Edificio Institutos de Investgación Apartado de Correos 22085 E-46071 València SPAIN

Speaker

Rui de Oliveira (CERN)

Description

The front-end hybrids for solid state and gas detectors will be crucial components of the next generation detectors. Requirements such as high-density and high-speed interconnects, low mass, radiation resistance and high- current and high-power dissipation capabilities are examples of the challenges to be solved concurrently. Over the past ten years we have been working on these problems for a variety of projects. The technologies for front-end hybrids developed at CERN are presented and future possibilities such as embedding active and passive circuits are described. Comments are made concerning the ability to access these technologies for large scale production by industry.

Summary

The front-end hybrids for solid state and gas detectors will
be crucial components
of the next generation detectors. There are many different
technologies that can be
used to make those circuits that interconnect the different
components.
Unfortunately low cost conventional PCB technology shows
rapidly its limitations
when high pin count components, high density connectors or
chips without packages
are used.

Other technologies such as HDI, MCM/D and MCM/C can
increase the density but also
have their weak points. Another way to increase the density
without impacting too
much the costs is to split the boards and to keep the high
density only there where
it is needed, for example with pitch adaptors.

As the front-end hybrids carry many signal lines in a small
area other problems due
to high speed signals arise such as crosstalk. Apart from
optimising the layout of
the signal lines, care should be taken to choose good
dielectrics (low loss) and
metals (skin effects).

Increasing density will often create thermal management
problems as the space
available for heat sinks decreases and the power density
increases. There are many
ways to place thermal management devices in these
modules for which each time the
three ways of power cooling (convection, conduction, and
radiation) should be
considered. The heat sink material itself is often also an
issue.

With the need for circuits with a very low mass as not to
influence particles in
their trajectory, HEP front-end electronics differs from
modules usually made by
industry. To this end CERN has developed a unique process
to fabricate multi-layer
circuits from polyimide with aluminium conductive layers
instead of copper. This
process has been used to make circuits for the ALICE pixel
detector and for low
mass aluminium cables carrying power and data.

To increase the density in commercial applications, industry
is developing methods
to embed passive components such as resistors and
capacitors directly in the
circuit board. Even several methods for embedding active
components have been
tested with good results but they need special treatment of
the dices. It is also
possible to stack active silicon dies. The way of building
these blocks is the
basis of embedding active components in PCBs.

Many of these highly advanced technologies are attractive
for front-end
electronics. However, mass production of them is not easy
as our experience with
specialised companies has shown. Technology transfer and
collaborations are crucial
to make mass production a success.

Primary author

Co-authors

Mrs A. Gris (CERN) Mr A. Teixeira (CERN) Mr D. Berthet (CERN) Mr E. van der Bij (CERN) Mr S. Ferry (CERN)

Presentation materials