ASIC Procurement

Download Report

Transcript ASIC Procurement

GLAST LAT Project
March 24, 2003
GLAST Large Area Telescope:
Gamma-ray Large
Area Space
Telescope
Tracker Subsystem
WBS 4.1.4
5D: ASIC Procurement
5D
Tracker Peer Review, WBS 4.1.4
1
GLAST LAT Project
March 24, 2003
Overview
• LAT-PS-1279: General Overview of Tracker ASIC Procurement
• Design specifications: LAT-SS-00169, LAT-SS-00170 (released)
• Wafer procurement from MOSIS/Agilent
– LAT-PS-1201 (released): Front-end readout chip (GTFE)
– LAT-PS-1222 (released): Readout controller chip (GTRC)
– Fabrication completed this month (submitted December 2002)
• Wafer probing and ASIC screening at UCSC
– LAT-PS-1250 (release in progress)
• Wafer lapping, dicing, and inspection
– LAT-PS-1321 (draft spec in progress)
• Wafer lot acceptance testing (MIL-STD-883E, Method 5007.6)
• Radiation testing (See Presentation 2B)
– LAT-TD-1325 (draft)
• Destructive physical analysis
• Electronics qualification and acceptance testing
5D
Tracker Peer Review, WBS 4.1.4
2
GLAST LAT Project
March 24, 2003
Wafer Procurement
• Vendor: The MOSIS Service of the USC ISI.
– Checking of design files, including DRC on the layout
– Subcontracting for fabrication of the masks (Dupont)
– Subcontracting the wafer fabrication
• HP/Agilent 0.5 m, 3-metal, epitaxial process (AMOS14TB)
– Thorough electrical testing of process monitors on each wafer
• MOSIS guarantees that the wafers meet the Agilent process
specifications
• MOSIS provides the test results in the form of e.g. physical
transistor parameter, sheet resistance, etc.
– Extraction of Spice model parameters for each wafer lot
• BSIM3 V3.1 models for design verification, especially in case
problems arise
– Shipment of wafers to SLAC
• LAT specifications: LAT-PS-1201 and LAT-PS-1222 (released)
5D
Tracker Peer Review, WBS 4.1.4
3
GLAST LAT Project
March 24, 2003
Wafer Procurement Status
• The flight wafers are locked in a dry-nitrogen cabinet in the wafer-probing
clean-room at UCSC.
• One wafer of each chip type was sacrificed to yield chips for testing.
5D
Tracker Peer Review, WBS 4.1.4
4
GLAST LAT Project
March 24, 2003
ASIC Screening
• Procedures, QA provisions, and travelers: LAT-PS-1250
• Detailed descriptions of the test vectors:
– LAT-TD-247 for the GTFE
– LAT-TD-248 for the GTRC
• Carried out in a cleanroom at UCSC
– Conforms with the LAT contamination control plan (LAT-MD-404)
– Conforms with ESD controls in LAT-MD-228 and NASA-STD8739.7
– Cascade automated probe station
– Custom probe cards interface to the VME readout
– Bad dice are automatically ink marked
– 100% testing of all functionality
– GTFE performance testing (to extent possible without SSDs)
– Data archived in an MS Access database
5D
Tracker Peer Review, WBS 4.1.4
5
GLAST LAT Project
March 24, 2003
EM ASIC Screening
• Only limited wafer probing was achieved with the Engineering Model
ASICs:
– The GTRC probe card had fatal problems (such as the probes
being put on backwards by the vendor) and did not work. A new
card has since been built and is functional at the full clock rate.
– The GTFE probe card could only operate at 2 MHz, due to
excessive capacitive loading on the IC output drivers. The chipto-chip outputs could not be tested at all. The probe card design
has been updated and new cards built to correct these problems.
It now operates at 20 MHz+ and tests all functionality.
– The probe station operation was not integrated with the test
program. We have since updated out software to enable fully
automated processing of an entire wafer plus automatic marking
of bad dice.
• Incomplete IC testing caused problems during the EM MCM
assembly. The wafer test system is now much improved and is
ready for flight production.
5D
Tracker Peer Review, WBS 4.1.4
6
GLAST LAT Project
March 24, 2003
Flight ASIC Screening
• The Cascade automated probe
station at the Santa Cruz Institute
for Particle Physics.
• Located in a small clean room that
is dedicated solely to wafer probing.
• In addition, it is under a downflow
hood with extra HEPA filters.
• The work area has recently been
upgraded to conform to LAT
contamination control and ESD
control requirements.
• Workers have been formally trained
in ESD and cleanroom procedures.
• Wafers are stored in a locked drynitrogen cabinet in this room.
5D
Tracker Peer Review, WBS 4.1.4
7
GLAST LAT Project
March 24, 2003
Flight ASIC Screening
• Status:
– Final versions of both probe
cards are in hand and tested.
– Software details are being
finalized and tested and the
software put under configuration
control.
– The procedure (LAT-PS-1250) is
in the review/release process.
– The clean room has been
updated for procedures, ESD,
etc. and was inspected.
– A date is set to train personnel in
clean-room and ESD
procedures.
5D
New GTRC probe card under
test in a manual probe station.
Tracker Peer Review, WBS 4.1.4
8
GLAST LAT Project
March 24, 2003
Wafer Lapping & Dicing
• Specification, LAT-PS-01321, in progress.
• A single vendor will do the following:
– Lap the wafers from 610 microns to about 270 microns. This
improves the wire bonding on the MCM
– Dice the wafers to our specified die size
– Pick the dice, separate the chip types, and separate bad dice
from good dice (according to ink markings) into waffle packs.
– Visually inspect the dice to reject
• Scratches or chipped edges
• Inked dice
• Dice of the wrong chip type
– Ship the completed waffle packs to Teledyne
– Perform wafer lot acceptance testing on each lot according to
MIL-STD-883E, Method 5007.6
5D
Tracker Peer Review, WBS 4.1.4
9
GLAST LAT Project
March 24, 2003
ASIC Qualification & Acceptance
• Qualification and acceptance testing of the ICs cannot be done on
the bare dice in any practical way.
• Therefore, the qualification and acceptance tests are those
pertaining to the completed MCMs (Presentation 7C).
• Qualification:
– 38 MCMs from the initial production, including 2 for DPA.
– DPA of some bare dice.
• Acceptance: all MCMs
– Complete functional tests of finished MCMs
– Noise, threshold, and gain performance (without the SSD load)
– Thermal cycles to acceptance levels
– Burn-in
• MCMs failing acceptance testing or not 100% functional after burn-in
(except for possibly <<1% bad individual amplifier channels) must be
discarded.
• Radiation testing (Presentation 2B): ICs mounted on mini MCMs.
5D
Tracker Peer Review, WBS 4.1.4
10