QLF08_LaBel.pdf - 第17页

IC and Component Selection for Space Sy stems – Presented by Kenneth A . LaBel 17 “Guaranteed” Radiation T olerance • A limited number of semiconductor manufacturers, either with fabs or fabless, will guarantee radiation…

IC and Component Selection for Space Systems – Presented by Kenneth A. LaBel

Radiation Perspective on IC Selection

• From the radiation perspective, ICs can be viewed

as one of four categories.

– Guaranteed hardness

• Radiation-hardened by process (RHBP)

• Radiation-hardened by design (RHBD)

– Historical ground-based radiation data

• Lot acceptance criteria

– Historical flight usage

• Statistical significance

– Unknown assurance

• New device or one with no data or guarantee

RHBD Voting Approach

http://www.aero.org/publications/crosslink/summer2003/06.html

IC and Component Selection for Space Systems – Presented by Kenneth A. LaBel

“Guaranteed” Radiation Tolerance

• A limited number of semiconductor manufacturers,

either with fabs or fabless, will guarantee radiation

performance of devices

– Examples:

• ATMEL, Honeywell, BAE Systems, Aeroflex

– Radiation qualification usually is performed on either

• Qualification test vehicle,

• Device type or family member, or

• Lot qualification

– Some vendors sell “guaranteed” radiation tolerant devices

by “cherry-picking” commercial devices coupled with

mitigation approaches external to the die

• The devices themselves can be hardened via

– Process or material (RHBP or RHBM),

– Design (RHBD), or

– Serendipity (RHBS)

Most radiation tolerant foundries use a

mix of hardening approaches

IC and Component Selection for Space Systems – Presented by Kenneth A. LaBel

Archival Radiation Performance –

Ground-based Data

• Reviewing existing ground radiation test data on a IC and

it’s application has been discussed previously

– For example. Christian Poivey at NSREC Short Course in 2002

– Using a “similar” device with data is risky, but sometimes

considered (though not recommended)

• In general, the flow is shown below

Does data

exist?

Same

wafer lot?

Sufficient

test data?

Test method

applicable?

Has

process/foundry

changed?

YES

Test recommended but may be

waived based on risk

assumption

YES

Data usable

YES

Test

YES