Fabrication, response and stability of miniature piezoresistive - - PowerPoint PPT Presentation

▶
fabrication response and stability of miniature
SMART_READER_LITE
LIVE PREVIEW

Fabrication, response and stability of miniature piezoresistive - - PowerPoint PPT Presentation

Dec 19, 2023 327 likes •788 views

Assessment of TFRs for piezoresistive sensors Fabrication, response and stability of miniature piezoresistive force-sensing thick-film cantilevers Thomas Maeder, Caroline Jacq, Stefane Caseiro and Peter Ryser cole Polytechnique Fdrale de

slide-1
SLIDE 1

1 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Thomas Maeder, Caroline Jacq, Stefane Caseiro and Peter Ryser

École Polytechnique Fédérale de Lausanne (EPFL), Switzerland

Assessment of TFRs for piezoresistive sensors

Fabrication, response and stability of miniature piezoresistive force-sensing thick-film cantilevers

slide-2
SLIDE 2

2 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook
slide-3
SLIDE 3

3 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook

Outline

slide-4
SLIDE 4

4 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Typical thick-film piezoresistive sensor

n Typical elements

n Sensing bridge n Offset trim n TCO trim n Differential amplifier

n Typical values (±)

n Offset ~30 mV/V n Response ~2-3 mV/V n TCO ~1 µV/V/K


(50 K : ~0.05 mV/V, ~2% F.S.)

n For 0.1% F.S.:

n Offset reduction ~10'000× n Stability (bridge) ~2-3 ppm

1 - Introduction

slide-5
SLIDE 5

5 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Cantilever force cell – principle

n Piezoresistive bridge n Thick-film resistors n Gauge factor KL ~12

slide-6
SLIDE 6

6 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Cantilever force cell – distances

Geometry

n L : for stress n d+ : positive signal (avg.) n d– : positive signal (avg.) n d : signal (overall) n b : cantilever width n h : cantilever thickness

d = 1

2 d + − d –

( )

σ = 6 b⋅h2 ⋅ L⋅ F εr = 6 b⋅h2 ⋅ E ⋅d ⋅ F

Nominal stress: Effective sensor strain:

r = KL ⋅εr

Response (signal / supply): (E = substrate elastic modulus; KL = piezoresistive longitudinal gauge factor)

slide-7
SLIDE 7

7 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Classical cantilever

1 - Introduction

Pros

n Full active bridge n Little thermal drift

Cons

n Double-side, complex

fabrication

n More difficult resistor

matching (separate prints)

n Layers on top side n Sensitive to horizontal

forces

L = 8 mm

Top Bottom

R2

–

R1

–

R2

+ R1 +

d+ = +6 mm d– = –6 mm

d / L = 75%

slide-8
SLIDE 8

8 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Single-side cantilever (type 1)

1 - Introduction

Pros

n Single-side, simple n Good resistor matching

(single print)

n Blank top side n Little thermal drift

Cons

n Half bridge, less

sensitive

n Sensitive to horizontal

forces

L = 6 mm

Bottom

d– = –4.75 mm

d+ = 0 mm (no stress)

d / L =40%

slide-9
SLIDE 9

9 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Single-side cantilever (types 2/3)

1 - Introduction

Pros

n Single-side, simple n Good resistor matching

(single print)

n Blank top side n Horizontal force

compensation Cons

n Half bridge, sensitivity

further reduced by "retrograde" resistors

n Buried conductors?

L = 8.08 mm

Bottom (w/o diel.)

d– = –6.75 mm d+ = -1.25 mm

d / L =34%

Bottom (with diel.)

slide-10
SLIDE 10

10 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Substrates (blank) – static fatigue

1 - Introduction

n Very good performance for ZrO2:Y (YSZ) & ZTA n Glassy (Al2O3 96% & LTCC) : poorer

slide-11
SLIDE 11

11 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Substrates (load cell) – static fatigue

1 - Introduction

n Strong degradation of high-strength substrates (ZrO2 & ZTA) n ZrO2 & ZTA better with single-side cantilevers (blank top side)

slide-12
SLIDE 12

12 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Substrates (load cell) – static fatigue

1 - Introduction

n Strong degradation of high-strength substrates (ZrO2 & ZTA) n ZrO2 & ZTA better with single-side cantilevers (blank top side)

(82) (170)

slide-13
SLIDE 13

13 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

LTCC structured cantilever

1 - Introduction

Bottom

Pros

n Single-side n Good resistor matching n Higher signal by structuration

n Concentration of compression n In practice ~2x

n Horizontal force compensation

Cons

n LTCC process critical for thin,

sensitive cantilevers (shrinkage matching, warpage)

n Resistor compatibility n Drift???

slide-14
SLIDE 14

14 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

LTCC cantilever – drift ?

1 - Introduction

n Moderate, consistent

signal

n No apparent drift

slide-15
SLIDE 15

15 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

LTCC cantilever – drift ?

1 - Introduction

n Abnormally high

signal

n Strong variations

between samples

n Significant drift

slide-16
SLIDE 16

16 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

YSZ cantilevers – drift?

1 - Introduction

Anelasticity in YSZ

n Ferroelasticity n Problematic for elastic substrate…

Pan & Horibe, Acta Mater. 1997

slide-17
SLIDE 17

17 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook
slide-18
SLIDE 18

18 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Substrates

1 - Introduction

Tested substrates

n All pre-fired n Not structured, same layout

Code Substrate material Thickness [µm] A 3YSZ (Kerafol) 45 B 3YSZ (Kerafol) 90 C Al2O3 96% (Kyocera A-476) 400 D Al2O3 96% (CeramTec Rubalit 708S) 150 E ZTA (CeramTec Rubalit HSS 2-14-02-004) 250 F ZTA (CeramTec Rubalit HSS4-38/3 S2) 320 G LTCC (Heraeus CT700) 470 / 710 H LTCC (Heraeus Heralock HL2000) 180 / 270 I LTCC (DuPont 951) 270 / 410

slide-19
SLIDE 19

19 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Layouts

Tested layouts 1) Short cantilever, half-bridge 2) Long cantilever (no tracks under resistors) 3) Long cantilever (tracks under resistors) 1 2 3

slide-20
SLIDE 20

20 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Fabrication

n Resistors (DP 2041) on dielectric:

n 3YSZ : ESL 4931 (for steel -> CTE ~ YSZ) n Others : ESL 4913 + 4917 (low CTE)

n 3YSZ : 45 µm critical, 90 µm OK n Al2O3 / ZTA : OK down to 150 µm (ZTA recommended) n LTCC : flatness critical (DP951 ≳ HL2000 > CT700)

slide-21
SLIDE 21

21 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook
slide-22
SLIDE 22

22 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

A – 45 µm 3YSZ – layout 3

n Very low heat conductance (45 µm thick, k ~ 2-3 W/m/K) n Thermal drift max ~1% (for 2'000 ppm F.S.)

slide-23
SLIDE 23

23 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

B – 90 µm 3YSZ – layout 2

n Same material, 2x thickness n ½ thermal drift

slide-24
SLIDE 24

24 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

D – 150 µm alumina – layout 2

n Very low thermal drift even for thinnest Al2O3

slide-25
SLIDE 25

25 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook
slide-26
SLIDE 26

26 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

A – 45 µm 3YSZ – layout 3

n High signal level, consistent n No visible drift (<±5 ppm) n Linear signal, ~43 ppm/mN

slide-27
SLIDE 27

27 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

B – 90 µm 3YSZ – layout 2

n High signal level, quite consistent n Linear signal, ~20 ppm/mN n Slight drift?

slide-28
SLIDE 28

28 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

B – 90 µm 3YSZ – layout 2 (loading)

n High signal level, quite consistent n Linear signal, ~20 ppm/mN n Slight drift?

slide-29
SLIDE 29

29 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

B – 90 µm 3YSZ – layout 2 (unloading)

n High signal level, quite consistent n Linear signal, ~20 ppm/mN n Slight drift?

slide-30
SLIDE 30

30 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

B – 90 µm 3YSZ – layout 3 (unloading)

n Apparent drift similar for both layouts

slide-31
SLIDE 31

31 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

D – 150 µm Al2O3 – layout 2

n Expected magnitude vs 90 µm YSZ (B) & 400 µm Al2O3 (C) n Very clean signal

slide-32
SLIDE 32

32 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

D – 150 µm Al2O3 – layout 2 (unloading)

n Expected magnitude vs 90 µm YSZ (B) & 400 µm Al2O3 (C) n Very clean signal

slide-33
SLIDE 33

33 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

H2 – 180 µm LTCC HL2000 – layout 3

n High signal, large variations n Visible zero drift (not anelastic) – damage ? n No apparent dependence on layout

slide-34
SLIDE 34

34 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

H2 – 180 µm LTCC HL2000 – layout 3

n High signal, large variations n Visible zero drift (not anelastic) – damage ? n No apparent dependence on layout

slide-35
SLIDE 35

35 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

H3 – 270 µm LTCC HL2000 – layout 2

n Thicker: mostly similar behaviour n Some "clean" samples

slide-36
SLIDE 36

36 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

H3 – 270 µm LTCC HL2000 – layout 2

n Thicker: mostly similar behaviour n Some "clean" samples

slide-37
SLIDE 37

37 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

I2 – 270 µm LTCC DP951 – layout 2

n Different LTCC : similar behaviour

slide-38
SLIDE 38

38 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

I2 – 270 µm LTCC DP951 – layout 2

n Different LTCC : similar behaviour

slide-39
SLIDE 39

39 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

I2 – 270 µm LTCC DP951 – layout 2

n Different LTCC : similar behaviour

slide-40
SLIDE 40

40 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

I2 – 270 µm LTCC DP951 – layout 2

n Increase of drift with apparent signal -> anomaleous

slide-41
SLIDE 41

41 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outline

Outline

  • 1. Introduction
  • 2. Manufacturing
  • 3. Thermal drift
  • 4. Force response & signal stability
  • 5. Conclusions & outlook
slide-42
SLIDE 42

42 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Conclusions

n Thin cantilevers on many substrates, including LTCC n Same manufacturing process:

n Post-fired, single-side n Two-layer, piezoresistors on thick-film dielectric n Resistors allowed above buried tracks (variant 3) or not (variant 2)

n Results:

n Al2O3 / ZTA : clean signal, thermal drift not a problem n 3YSZ : possibly slight anelastic drift & thermal effects due to very low

thermal conductance of cantilever

n LTCC : signal mostly unstable (some clean samples)

n Cause ? Low thermal expansion ?

slide-43
SLIDE 43

43 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

Outlook

n Elucidate drift mechanism on LTCC

n Perform progressive loading tests n Check for resistor damage n Try on LTCC with high CTE – should avoid instabilities

n Extended analysis of new design

n Performance & economics vs existing cantilever n Sensitivity to side loads n Lowest practical force ranges (deflection, manufacturing…)

slide-44
SLIDE 44

44 IMAPS/ACerS 12th CICMT, Denver, 19-21.4.2016

The end Thank you for your attention !