Introduction R2E Validation Levels

COTS Management COTS Identification COTS Traceability

Single Event Total Ionizing Dose Displacement Damage

Introduction

R2E/EPC Projects will be based on Commercial Off The Shelves - COTS components. Ensuring a given hardness to a design and production based on this component category (not designed for) requires a strong strategy.

R2E Validation Levels

A given machine failure rate leads to following figures driven by the quantities of units being in operation. For example, an IGBT transistor with a cross section of 1E-11 cm², used 4 times in 100 converters will lead to 4 LHC Machine failures / year at a E>20MeV fluence of 1E9 part./(cm².year), for this component only.

Testing component in a mixed field like CNRAD gives the opportunity to validate the SEE and the TID criteria, but SEE sensitivity can be under-estimated due to facility low energy spectrum.

Component Validation Levels

*   100Gy is very high compared to non-tunnel installed units, but it is a bit low for tunnel installed units: in such a case, a dose spread over the full unit family installed is obtained moving / replacing units from low dose <=> high dose locations in the full LHC machine life.
*   ELDRS impact can justify a higher TID limit, around 200 Gy to profit a safety margin of 3x to 4x compared to the LHC 50 Gy dose.
** This number is to be adapted (can be lower) versus component occurence in the final design unit, and should lead to a global MTBF only considering the radiation of 100 000 hours. A cross section of a component being used once in a converter type will give a probability of 1 failure / year for this component only, given 100 converters of same type are installed and receive 1E10 part./cm² in a year.

COTS Management

COTS components sensitivity is given by the original component design (inherent structure at the semiconductor level), but also by the foundry being used for production, and then is very hard to keep under control. A component being tested some years ago doesn't give a lot of information on the today component behaviour, since perhaps produced differently: new design or process or even in another foundry.

A Quality Plan has then to be strictly followed to minimize the risk.

R2E COTS CERN Validation Process .vsd

COTS Identification

COTS Identification is the process which determines if a given COTS is sharing the same BATCH than another one.

COTS Traceability

COTS traceability is certainly a must, even if perhaps not mandatory in our cases.

R2E Component Single Event Validation

All components are not showing the same sensitivity to radiations, and especially to the single event effect: theoritically and practically. Based on the litterature and the domain knowledge, 3 classes will be used to classify the components, with testing effort in relation to the known criticity.

S.E. Classes Definition

Component Classification Criteria Table

• It can be noticed that one component initially in class-0, since being potentially replaced by other components can later reach Class-1, and then Class-2 if only one reference eventually fits the requirements from test.
• Case of a single source component (key-component) being safely covered by known and efficient mitigation techniques could lead to a Class-1 classification, ignoring its single source availibility.
• The general aim of the classification is that one component known to be sensitive, with no known efficient mitigation techniques available at the level of the design, should be tested carefully, and perfectly known regarding its sensitivity and its failure modes.
• It is possible that a component share some Class 1 and Class 2 status; in such a case, the test will be optimized regarding a specific approach.

S.E. Test Process

R2E S.E. COTS CERN Validation Process .vsd

R2E Component T.I.D Validation

TID Classes Definition

Component Classification Criteria Table

T.I.D. Test Process

R2E T.I.D. COTS CERN Validation Process .vsd

R2E Component Displacement Damage Validation