2. Matters arising from the minutes:
P. Borgeaud enquired about the status of plans to organize a common workshop on the radiation-tolerance of power supplies and other power systemsí components located in the ATLAS and CMS caverns. G. Stefanini replied that a common informal working group, convened by R. Richter, was addressing the issue. The working groupís minutes can be found on the ATLAS Front-end Electronics page at:
3. RD29 - Further Oral Report on the Status of DMILL Stabilization at MHS
by M. Dentan, Saclay)
Further to his report at the previous meeting of the Board, M. Dentan gave an oral update on progress towards stabilization of the radiation-hard DMILL technology at the integrated circuit manufacturer MHS and the formal acceptance by the DMILL Consortium of the technology transfer from CEA-LETI to MHSís foundry in Nantes F).
The process parameters were being optimized to ensure that the gain of bipolar transistors always remained greater than the value 70 after a total dose of 10 Mrad and a neutron fluence of 1014 cm-2. In addition, the batch-to-batch spread of the high-value resistor RSRHV) was being improved by reducing the ion-implant doping. For high-dose applications a new high-value resistor device REXTRINS), with improved radiation hardness had been introduced. Several batches incorporating these changes were being processed and would be measured in order to check that the conditions for final acceptance of the technology transfer have been met.
M. Dentan concluded by summarizing the various acquisitions that had recently resulted in MHS passing from joint ownership by the Daimler-Benz and Lagardere groups, via Vishay Intertechnology Inc., to full ownership by the US semiconductor manufacturer Atmel.
Pending the submission of the final written
RD29 status report expected after the formal acceptance of the technology
transfer) the Board congratulated the DMILL Consortium on its thoroughly
professional work. The Board decided to seek reports at its future meetings
from those development groups implementing designs in the DMILL technology.
4. Status of Frame Contract for Procurement of Rad-hard ASICs
Ph. Farthouat, G. Hall, and G. Stefanini reported on the status of negotiations to define a formal framework for placing contracts for procurement of rad-hard ASICs. Meetings have taken place between representatives of ATLAS, CMS, CERN SPL division and each of the potential suppliers MHS DMILL) and Honeywell. An invitation to Harris to conduct a similar discussion remains open.
Following tentative requirements put forward by the ATLAS and CMS collaborations,
MHS have now submitted draft proposals addressing guarantees on yield and
acceptable ranges for specified process parameters. However, the foundry
can only guarantee yield if it is first assured of the quality and robustness
of the designs themselves with respect to variations in the process parameters.
MHS therefore can offer yield guarantees only after each design has been
submitted to a series of "corner runs". These corner runs would deliberately
push critical process parameters to the limits of their allowed range in
order to evaluate the robustness of the design. Adoption of this procedure
would introduce additional processing costs, manpower requirements and
delays. Further negotiation will be needed with MHS in order to converge
to a mutually acceptable set of financial and technical conditions and
procedures. Honeywell have not yet given an answer on the question of yield
guarantee, and more discussion will be needed.
5. Oral Report on the Status of the new LHC Integrated Circuit Test Lab by J. Christiansen, CERN)
J. Christiansen reported on the status and initial experience with the mixed-signal integrated circuit test facilities installed in the CERN microelectronics development group in December 1997. This infrastructure can be made available to LHC ASIC design teams under well-defined conditions. It is optimized for design verification testing and full characterization of mixed analog-digital ASICs in order to check design robustness and quality prior to launching volume production. Testing can be carried out on-wafer, with packaged chips or hybrids. A limited amount of time can be made available for test of small production runs. Access is on an assisted, self-use basis only. The LHC electronics coordinators, in agreement with the CERN microelectronics group, decide the access schedules and priorities.
After completion of the commissioning and training phase, two ASICs have been fully characterized a 32-channel TDC and a multi-range ADC) and production testing has been carried out for a batch of 2400 TDCs. A further seven ASICs for ALICE, ATLAS and CMS are being prepared for test and the schedule is now fully booked until the autumn. It is estimated that about 1 engineer-year of debug effort could have been saved on a previous TDC ASIC development had the facility been available earlier.
The Board congratulated all concerned in the selection, installation and commissioning of the IC testing infrastructure and the achievement of the successful first results.
6.a Deep sub-micron technology evaluation for rad-tolerant applications:
P. Jarron and A. Marchioro summarized the current status of RD49ís work on the evaluation of high-volume, commercial submicron CMOS technologies for radiation tolerant LHC applications. Test devices from two different 0.25mm technologies have been measured with respect to their intrinsic radiation tolerance, which is a result of their use of very thin gate oxides <7 nm) and advanced isolation techniques. One of these technologies has been shown to be rad-tolerant up to total doses gammas) of 200 krad. Devices in these technologies can be further hardened to total doses in excess of 10 Mrad by applying special layout techniques enclosed NMOS device geometry and guard-rings). Neutron irradiation up to fluences of 1014 cm-2 showed only 8% degradation in the deviceís transconductance.
Two major questions that remain to be answered by RD49 before these technologies can be confidently applied in rad-tolerant applications are the sensitivity to single event upset and single event latchup. The feasibility of the rad-tolerant layout approach in 0.25mm technologies also needs to be demonstrated on the scale of a large circuit. In order to answer these questions a test chip has been designed in the most promising 0.25mm technology and is about to be submitted to fabrication. The test chip also contains circuits to study the analog characteristics of the technology matching and noise), to implement a number of digital standard cells in rad-tolerant layout, and to evaluate a circuit for upset-hardened digital storage elements. The test chip also contains a full pixel cell designed to the specifications of ALICE using the rad-tolerant layout techniques.
Pending the release of the full vendor-supported design kit in the autumn of 1998 a minimal design kit for the Cadence tool environment has been developed. Commercially available standard cell libraries should be suitable for applications up to about 100 krad total dose and, for higher doses, a standard cell library using the rad-tolerant layout technique is under development.
The Board fully supported RD49ís technology evaluation work-plan as described above and agreed that the approach is highly promising for rad-tolerant applications and as a "fall-forward" solution for development of rad-hard digital circuits and low- to medium-voltage swing analog circuits. Large dynamic-range analog circuits, such as those required for calorimetry, are expected to be more difficult to implement in low-voltage submicron technologies.
The Board discussed a proposal for a coordinated HEP-wide mechanism
for eventual access to such high-volume submicron technologies, which are
not normally offered to customers with the low-volumes on the industrial
scale) of the LHC projects. The companies involved insist on all commercial
and technical issues being coordinated through a single point of contact,
which they prefer to be CERN. Independent non-disclosure agreements must
be signed by each end-user organization currently planned for RAL, Saclay
and INFN). The HEP community will have to take responsibility for monitoring
the radiation tolerance of the technologies. RAL has volunteered to organize
a Multi-Project Wafer MPW) service for the HEP community.
6.b Issues concerning use of Commercial Off The Shelf COTS) Components at the LHC:
P. Jarron presented the status of RD49ís work on COTS components and the contacts established with specialized agencies and consultants Sandia National Labs, CEA-DEIN, ESA-ESTEC, CNES and IC Radiation Technologies). These agencies and consultants have extensive experience in the use of COTS components for Space and Nuclear applications. P. Jarron identified four major issues concerning the use of COTS components at the LHC
The Board fully supported the proposal of P. Jarron to recommend the collaborations to make a start by jointly awarding a consultancy contract to radiation experts from Sandia National Labs. A draft offer received from Sandia covers access to Sandiaís COTS database, advice on how to mitigate Single Event Effects, and advice on how to adapt Space COTS qualification protocols to meet the requirements of LHC. The Board recommends that the Sandia draft consultancy contract should be reformulated as a review of the risks of operating COTS in the LHC radiation environments. It should include training/mentoring of key electronics engineers and project leaders not members of RD49), and it should address the four critical points identified by P. Jarron.
In view of the risks associated with COTS, the extra procedures and the expense involved in COTS qualification and procurement, the Board strongly encourages the collaborations to take steps to:
ii) Report to the LHCC on the strategic COTS issues and stress need for concerted action by management of the collaborations, starting with a review by the experts from Sandia.
iii) Discuss the form of the consultancy contract with Sandia.
The deadline for receiving abstracts for
the Rome LEB workshop was the 30 April the date of this meeting). E. Petrolo
reported that 35 abstracts and 10 requests for extension of the deadline
had been received. Abstracts will be available on the workshopís home page
) by end May, and paper selection will be performed at the next LEB meeting
19 June). The workshop will be organized with plenary sessions in which
invited speakers introduce topics that are then addressed in detail in
parallel sessions. The Board reviewed the proposed list of themes and the
corresponding invited speakers.
P. Sharp summarized his report to the LHCC
at its meeting of 3-4 March 1998. This covered the status of DMILL,
prospects for rad-tolerant electronics in 0.25mm
technologies, the main issues concerning rad-tolerant COTS components,
and the important questions of risk assessment and project management.
9. Strategic Objectives of the LEB for 1998
The Board revisited the question of what
are the important topics to be tackled in the coming months. It was agreed
that the electronics for slow controls, calorimeters and muon detectors
should be examined. Other topics that the Board considered had been underestimated
or that should be reviewed were: organizing the test and integration of
the LHC electronics, risk management strategies and the QA procedures adopted
by the collaborations. The issues of crate standards, power supplies and
cabling etc., should be revisited.
10. Any Other Business
The Board decided to seek clarification of the role of the newly created Advisory Committee for Electronics Support ACES) and its relation to the LEB.
G. Hall pointed out that, within LEMIC and the LHCC, CMS had raised the need for test beams with a 25ns structure by early 1999. He was surprised that ATLAS had not strongly supported the request, which he viewed as essential to the proper test and evaluation of prototype detectors and their associated front-end electronics.
The IEEE Nuclear Science Symposium and Medical Imaging Conference has been scheduled to take place in Europe in October 2000, either in Brussels or at a location to be chosen in France. It was suggested that the LEB workshop of 2000 could be in some way combined with this event.
The date of the next LEB meeting, which will be mainly dedicated to paper selection and the planning of the programme of the Rome workshop, was changed to 19 June 1998.