Analog Devices Welcomes Hittite Microwave Corporation

Analog Devices Welcomes
Hittite Microwave Corporation
NO CONTENT ON THE ATTACHED DOCUMENT HAS CHANGED
www.analog.com
www.hittite.com
Report Title:
Qualification Test Report
Report Type:
See Attached
Date:
See Attached
Wafer Process: PHEMT-B
QTR: 07001
Rev: 02
HMC154
HMC356
HMC372
HMC373
HMC374
HMC375
HMC376
HMC382
HMC484
HMC486
HMC487
HMC489
HMC536
HMC546
HMC549
HMC590
HMC591
HMC592
HMC646
Note: This qualification was designed to evaluate the pHEMT-B process. The package type is only specific to the
MS8G which was tested by HMC484MS8G. Other package qualifications are available at www.Hittite.com.
1.0 Introduction
The definition of a qualification family, as defined by EIA/JESD47 Stress-Test-Driven
Qualification of integrated Circuits, is all devices that use the same wafer fabrication technology,
wafer fabrication process and wafer fabrication site.
This qualification procedure is designed to satisfy the reliability requirements for Hittite
Microwave Corporation’s GaAs MMIC Switches using the pHEMT-B wafer process. The
HMC484MS8G was chosen as the qualification vehicle to qualify the pHEMT-B Power process.
A complete data sheet for the HMC484MS8G can be found at www.hittite.com.
1.1 General Description
The pHEMT-B GaAs wafer process uses optical gate stepper lithography and is ideal for
switches, amplifiers & LNAs. The entire die, with the exception of bridge metal and bonding
areas, is fully passivated to help protect the device from environmental conditions.
The HMC484MS8G is a low cost SPDT switch in 8-lead MSOPG package for use in transmitreceive applications which require very low distortion at high input signal power levels, through
10 watts (+40 dBm). The device can control signals from DC to 3.0 GHz. The design provides
exceptional intermodulation performance; > +70 dBm third order intercept at +5 volt bias. RF1
and RF2 are reflective shorts when “OFF”. On-chip circuitry allows single positive supply
operation from +3 Vdc to +10 Vdc at very low DC current with control inputs compatible with
CMOS and most TTL logic families. Applications include wireless infrastructure, ISM/Cellular
portables/handsets, automotive telematics, mobile radio, test equipment.
2.0 Summary of Results
All testing has been completed. Activation energy for pHEMT-B process is 1.4 eV. Using 85 °C
as the normal operating temperature of the device, the equivalent device hours for 74,000 hours
at 125 °C is calculated to be 7.18 X 106 hours. Using a Chi-Square 90% confidence interval, the
failure rate is calculated to be better than 3.23 x10-7 failures per hour or 323 FIT.
PAR
A
TEST
QTY
IN
QTY
OUT
4.1.1
Initial Electrical
231
231
Complete
IR Reflow
231
231
Complete
HTOL, 1000 hours
77
77
Complete
Post HTOL Electrical Test
77
74
Autoclave
77
77
PASS/FAIL
NOTES
3 failures due to
mechanical
damage during
electrical
testing.
Complete
2 failures due to
mechanical
damage during
electrical
testing.
Post Autoclave Electrical Test
77
75
THB, 1000 hours
77
77
Complete
Post THB Electrical Test
77
77
Pass/No Failures
pHEMT-B Failure Rate Estimate
Based on the HTOL results, a failure rate estimation was determined using the following
parameters:
To=85 °C(358 °K)
Ta=125 °C(398 °K)
Device hours = (74 X 1000hrs) = 74,000 hours
For 0.5µm pHEMT-B GaAs MMIC, Activation Energy = 1.4 eV
Acceleration Factor = exp[1.4/8.6 e-5(1/358-1/398)] = 97 (Arrhenius Equation)
Equivalent hours = 74,000 hours X 97 = 7.18 e6 hours
Since there were no relevant failures, C=0.
Since we used a time terminated test R = C+1 = 1
The GaAs failure rate was calculated using Chi Square Statistic: λα = [(χ2)α,2r]/2t
λ60 = [(χ2)60,2]/(2X 7.18 e6)] = 1.8/1.44 e7 = 1.25 e-7 failures/hour or 125 FIT
λ70 = [(χ2)70,2]/(2X 7.18 e6)] = 2.4/1.44 e7 = 1.67 e-7 failures/hour or 167 FIT
λ90 = [(χ2)90,2]/(2X 7.18 e6)] = 4.6/1.44 e7 = 3.190 e-7 failures/hour or 319 FIT
3.0 Qualification Flow Chart
Physical Attributes
Process Tests
Physical Attribute
Testing
n=20
Initial Electrical
Parameters
n=231
IR Reflow
Simulation
n=231
Autoclave
n=77
High Temp.
Operating Life
n=77
Final Electrical
Parameters
n=231
Bond Pull
n=10
(30 wires)
Temperature
Humidity Bias
n=77
Die Shear
n=10
SEM
Inspection
n=5
CrossSection
Inspection
n=5
4.0 Test Procedures
4.1 MMIC related tests - These tests are designed to demonstrate that the GaAs MMIC devices produced with the
0.5µm PHEMT process are capable of maintaining the specified parameters throughout their useful life under rated
operating conditions. The HMC484MS8G die was chosen to qualify the pHEMT-B process. The results of these tests
qualify by similarity all other products using the same wafer process. A complete data sheet for the HMC484MS8G can
be found on our catalog website.
4.1.1 Initial Characteristics - 231 HMC484MS8G devices were electrically tested for DC and critical RF parameters. at
ambient temperature (+25 °C). There were no failures in this test.
4.1.2 IR Reflow Simulation - 231 devices from 4.1.1 are subjected to the following temperature profile in air:
This test was performed at Hittite.
4.1.2.1 Autoclave - 77 devices from 4.1.2 were subjected to 96 hours of un-biased Autoclave (121°C/100% RH, 14.7
psig). This test was performed at Hittite.
4.1.2.2 High Temperature Operating Life - 77 devices from 4.1.2 were subjected to 1000 hours of accelerated burn-in
at Tj=125 °C with a 5 volt constant bias applied to Vdd (pin 4) and pins 1,2,7,8, and paddle backside grounded to force
both paths of the switch into the off state. This test is performed at Hittite.
4.1.2.3 Temperature Humidity Bias (THB) - 77 devices from 4.1. were subjected to 1000 hours of temperature
humidity bias (THB) with a 5 volt constant bias applied to Vdd (pin 4) and pins 1,2,7,8, and paddle backside grounded
to force both paths of the switch into the off state. This test is performed at Hittite.
4.1.6 Final Characteristics - All devices from 4.1.2 were electrically tested to DC and critical RF parameters at ambient
temperature (+25 °C). Any out of specification parameter is considered a failure. This test was performed at Hittite.
There were no failures in this test.
5.0 Physical Attributes
5.1 Bond Pull -10 die from the wafer (4 quadrants and center) were selected. The die were mounted to test coupons
with conductive epoxy. Wires were thermosonic bonded from the die to the coupon. 3 wires on each of the ten samples
(30 total) were pulled to destruction to the requirements of MIL-STD-883 Method 2011, condition D. Variables data
was recorded. This test was performed at Hittite. There were no failures in this test.
5.2 Die Shear – The 10 die from 5.1 were sheared to destruction to the requirements of MIL-STD-883 Method 2019.
Variables data was recorded. This test was performed at Hittite. There were no failures in this test.
5.3 Metal and Dielectric Thickness – 5 die from each wafer lot (4 quadrants and center) were selected. The die were
mounted, cross-sectioned and polished. Metal and dielectric thickness was recorded for each layer and compared to
expected values. There were no anomalies. This test was performed at Oneida Research Services in Whitesboro, NY.
(Reference Photo 1)
5.4 SEM Inspection – 5 die from each wafer lot (4 quadrants and center) were selected. The die were mounted and
sequentially delayered to perform a comprehensive inspection to the requirements of MIL-STD-883 Method 2018.
There were no anomalies. This test was performed at Oneida Research Services in Whitesboro, NY. (Reference Photos
2-4)
Photo 1
SEM micrograph of structure, Wafer Lot
FPA63816M, Wafer 089, Die 38401, 2500X, 20 kV
Photo 2
SEM micrograph of typical ohmic metallization,
Wafer Lot FPA63816M, Wafer 089, Die 38401, 5000X, 20 kV
\
Photo 3
SEM image of typical gate metallization, Wafer
Lot FPA63816M, Wafer 089, Die 38401, 10000X
Photo 4
SEM image of metal to metal layer contacts,
Wafer Lot FPA63816M, Wafer 089, Die 38401, 1000X
Photo 5
SEM micrograph of airbridge, Wafer Lot
FPA63816M, Wafer 089, Die 38401, 2000X, 20 kV