DDR2 DRAM High-Frequency Test at Probe (HFTAP) · 2020. 7. 1. · Wafer-Level Final-Sort Challenge...

19
6/6/2005 Southwest Test Workshop 2005 DDR2 DRAM High-Frequency Test at Probe (HFTAP) Masahide Ozawa Elpida Memory, Inc. Yoichi Funatoko FormFactor Inc., Asia Southwest Test Workshop 2005

Transcript of DDR2 DRAM High-Frequency Test at Probe (HFTAP) · 2020. 7. 1. · Wafer-Level Final-Sort Challenge...

  • 6/6/2005Southwest Test Workshop 2005

    DDR2 DRAM High-Frequency Test at Probe (HFTAP)

    Masahide Ozawa Elpida Memory, Inc. Yoichi Funatoko FormFactor Inc., Asia

    Southwest Test Workshop 2005

  • p. 2

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Presentation Outline• Introduction• DDR2 DRAM wafer-level-final-sort tests objectives

    and goals• High performance probing technology solution• New probing technology internal qualification • Customer evaluation• Summary and conclusion• Follow on work

  • p. 3

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Wafer-Level Final-Sort Challenge

    2004 2005

    Elpida : 333MHz DDR2 DRAMFFI : K3 [x30DUT]533MHz testerSWTW2005

    Device Frequency(MHz)

    300

    200

    100

    Elpida : 100MHz Mobile RAMFFI : (S200, shared driver 200MHz)143MHz testerPresented at SWTW2004

  • p. 4

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    High Performance K3 Probing SolutionHFTAPTM probing technology

    FFI K3 probing technology

    K3 performance enables DDR2 testing @ 333MHz / 667Mbs

  • p. 5

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    0.5 1.0 1.50.0 2.0

    -15

    -10

    -5

    -20

    0

    fre q, GHz

    dB(S

    (4,1

    ))

    m1

    m2

    m1freq=dB(S(4,1))=-1.004

    142.0MHzm2freq=dB(S(4,1))=-3.004

    877.0MHz

    0.5 1.0 1.50.0 2.0

    -40

    -30

    -20

    -50

    -10

    fre q, GHz

    dB(S

    (2,1

    ))

    m3

    dB(S

    (3,1

    ))

    m4

    m3freq=dB(S(2,1))=-23.964

    277.0MHzm4freq=dB(S(3,1))=-31.299

    277.0MHz

    FFI Attenuation & Cross-talk Simulation

    Attenuation Cross-talk

    Near-EndFar-End

  • p. 6

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    FFI Internal qualification Tpd Distribution by TDR Measurement

    1.42 1.43 1.44 1.45 1.46 1.47 1.48

    100.0%99.5%97.5%90.0%75.0%50.0%25.0%10.0%2.5%0.5%0.0%

    maximum

    quartilemedianquartile

    minimum

    1.4810 1.4772 1.4720 1.4640 1.4575 1.4515 1.4465 1.4396 1.4325 1.4270 1.4220

    QuantilesMeanStd DevStd Err Meanupper 95% Meanlower 95% MeanN

    1.4518250.00948680.00035361.45251911.4511309

    720

    MomentsFFI_(ns)

    Distributions

    Channel to channel skew was +/- 30ps

  • p. 7

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida Evaluation Items• Basic characteristic

    – Skew– Waveform– Tr/Tf– Jitter

    • Device evaluation– At-speed test– Correlation between wafer and package

    TCK-VDD ShmooAC parametersIdd current

  • p. 8

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation:Channel to Channel Skew

    Tpd Distribution of HFTAP Probe Cards

    020406080

    100120140160180

    -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30

    Difference from average [ps]

    Cou

    nt

    23.27.7

    3σ[ps]σ[ps]

    Tpd spec (±75ps) is satisfied.

    σ: Standard Deviation

    n=576

    (TDR Measurement)

  • p. 9

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Input Waveform (Amplitude 0.5V Vt:0.9V )

    0ps 937.5ps

    0ps 1875ps① ②

    0ps

    937.5ps

    1875ps

  • p. 10

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Tr/Tf (Amplitude 0.5V 20-80%, Vt:0.9V )

    Tr Tf

    Tr/Tf of K3 probe card

    277304

    325

    399

    0

    100

    200

    300

    400

    500

    Tr Tf Tr Tf

    Driver I/O

    Tr/

    Tf

    [ps]

  • p. 11

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Jitter

    110ps

    ※Evaluated at 533MHz

    The source of jitter is caused by attenuation, cross talk, and tester timing control.

    Jitter of K3 is similar of existing final tester

    Jitter of K3 Probe Card

    110 114122

    142

    0

    50

    100

    150

    Rise Fall Rise Fall

    Driver I/O

    Jitte

    r [ps

    ]

  • p. 12

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: At-Speed Test (1)

    High-speed WT Assembly Package Test

    PC533

    PC667

    Packaging separately divided by the result of High-speed WT

  • p. 13

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: At-Speed Test (2)• Correlation Result

    Correlation test result

    was similar to final test.

    CorrelationRatio

    K3 87.0%

    Correlation of at-speed test between wafer and package

    0% 20% 40% 60% 80% 100%

    PC533

    PC667

    Resu

    lt o

    f W

    afer

    Test

    Result of Package Test

    upgrade

    same grade

    downgrade

  • p. 14

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Correlation (core test)

    Correlated well

    These are evaluated at high temperature with Vdd set in 3points.

    White line means “Be equal” for wafer test and package test. If plot is near this white line, correlation is OK.

    Test Item: TCK-VDD Shmoo for some pattern

    CoreTest correlation of K3

    0

    10

    20

    30

    40

    50

    0 10 20 30 40 50

    Package Test Result [ns]

    Waf

    er

    Test

    Resu

    lt [

    ns]

  • p. 15

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    AC Parameter Correlation of K3

    0.0

    0.2

    0.4

    0.6

    0.8

    1.0

    0.0 0.2 0.4 0.6 0.8 1.0Package Test Result [ns]

    Waf

    er T

    est R

    esul

    t [ns

    ]

    AC Parameter Correlation of K3

    0.0

    0.5

    1.0

    1.5

    2.0

    2.5

    3.0

    0.0 0.5 1.0 1.5 2.0 2.5 3.0Package Test Result [ns]

    Waf

    er T

    est R

    esul

    t [ns

    ]Elpida evaluation: Correlation (AC parameters) 1

    ※Sub line is indicated the range of Timing accuracy (±100ps)

    These are evaluated at high temperature with Vdd set in 3points.

    Test Item: AC Parameters (setup, hold, TCK etc.)

    Compared by package test result•Different 100ps~150ps for some test item

    Available for testing with adjustment

    Zoom up

  • p. 16

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Correlation (AC parameters) 2

    These are evaluated at high temperature with Vdd set in 3points.

    ・40mV higher VIH for specificpins

    Available for testing with adjustment

    Test Item: AC Parameters (VIH/VIL)

    VIH/VIL Correlation for K3

    0.00

    0.04

    0.08

    0.12

    0.16

    0.00 0.04 0.08 0.12 0.16

    Package Test Result [V]

    Waf

    er T

    est R

    esul

    t [V

    ]

    All Itemsome VIH

  • p. 17

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Elpida evaluation: Correlation (Idd)

    Correlated well

    These are evaluated at high temperature with Vdd set in 3points.

    Test Item: Idd Current

    Idd Correlation for K3

    0

    50

    100

    150

    200

    250

    300

    0 50 100 150 200 250 300

    Package Test Result [mA]

    Waf

    er T

    est R

    esul

    t [m

    A]

  • p. 18

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Summary and Conclusion

    Correlation test result was similar to final test.

    Correlation between wafer and package test

    Available to device measurement for 333MHz

    DDR2 DRAM at device speed testing

    High-frequency wafer testing

    PC667 (333MHz) evaluation passed all criteria.

    Basic characteristics(Waveform, Tr/Tf, Jitter)

    ResultsObjectives

  • p. 19

    Southwest Test Workshop 20056/6/2005 Elpida / FFI

    Follow-on Work• Elpida Memory, Inc.

    – Further evaluation for volume production

    • FormFactor, Inc.– Higher parallelism K3 [x64 DUT & x 128 DUT] – Production qualification at > 300 MHz.– Customer evaluation for 500MHz and beyond with K5