1. Switchable LTCC/Polyimide Based Thin Film Coils
microsystems
laboratory
Switchable LTCC/Polyimide Based Thin Film Coils
Loren Rieth 1,2, Siddharth Chakravarty 1, Jui-Mei Hsu 1, Richard Normann 3,
Florian Solzbacher 1,2,3, Matthias Klein 4, Michael Töpper 4, Sohee Kim 5
Micrograph of a Utah Electrode Array (UEA) with 100 equal length probes.
Utah electrode array
Integrated circuit with neural
amplifiers and/or stimulators,
signal processing and RF
telemetry electronics
LTCC based coil for wireless
power/data transfer
Integrated UEA, signal IC, and
Inductive power coil.
1 Materials Science and Engineering, Univ. of Utah, Salt Lake City, UT
2 Electrical and Computer Engineering, Univ. of Utah, Salt Lake City, UT
3 Bioengineering, Univ. of Utah, Salt Lake City, UT
4 Fraunhofer Institute for Reliability and Microintegration IZM, Berlin, Germany
5 Fraunhofer Institute for Biomedical Technology IBMT, St. Ingbert, Germany
IV. Polyimide Based Coil Design and Simulation:
Two separate coil designs are simulated:
Type I coil with a single winding layer (fewer turns)
Type II coil with two winding layers (more turns)
Initial coils will have two winding layers with a “switch” to disable one winding layer for controlled inductance and capacitance
Coils will be Cu or Au metal traces electrodeposited on polyimide/Kapton® substrates
Metal lines will be from 10 to 20 µm thick for low resistance (< 1 kΩ) and high fabrication yields
Metal line height, the spacing between lines will be from 10 to 20 µm wide to maximize inductance and minimize parasitics
Polyimide based coils will be adhered to a LTCC ferrite platelet to enhance voltage gain and Q-factor, and to reduce the magnetic field on the signal processing IC below
Type I Coil
Type II Coil
I. Objectives and Introduction:
Develop a chronically implantable inductive power coupling for in vivo microelectronics
Use electromagnetic simulations to design the coil:
Optimize coil design for 2.64 MHz operation
Maximize inductance and minimize parasitic losses (e.g. capacitance; resistance; loss tangent)
Microfabricate a thin coil with a high Q-factor (quality factor) to efficiently receive power from an external supply
Supply 3 V to power signal/telemetry IC
Receive command signals encoded on the power waveform
~ 5  5 mm2 to match UEA dimensions
Two approaches have been followed including:
An LTCC ferrite device with an integrated metal coil
A microfabricated gold (Au) coil on a polyimide substrate bonded to a LTCC ferrite platelet
Q-factor and inductance as a function of the coil line height. Line spacing is 10 µm and 75% areal coverage.
Inductance versus the distance between the coil and ferrite platelet. Single layer coil with line height and width of 20 µm, line spacing of 10 µm, and 62 turns for 75% areal density.
Quality factor and inductance versus number of turns for Type I and II coils. Maximum Q occurs when windings fill 80% of internal spiral area; line spacing 10 µm and line height 20 µm.
II. LTCC Coil Design and Testing:
Low temperature co-fired ceramic (LTCC) technology to form thin (~200 µm) ferrite platelet
Ferrite material based on Fe2O3 has high permeability (µr~200) and high resistance (R)
Ferrite concentrates the magnetic flux in the coil
The coil is formed with screen printed Ag paste in a square geometry planar coil (20 µm thick)
Coil performance measurements:
22 mm diameter external (supply) coil with 17 turns on two layers
Supply coil inductance: 6.03 µH
10 mm separation between supply and test coils
2.64 MHz supply frequency with 13.8 V peak-to-peak sine wave
Table II. Simulation results for Au coils on polyimide substrates
Q-factor and inductance as a function of coil line width. Line spacing is 10 µm and line height is 20 µm and 75% areal coverage.
Type I (single-layer thin-film Au coil on ferrite)
Type II (double-layer thin-film Au coil on ferrite)
w=s=20µm
15µm
10µm
20µm
Inductance (µH) 12 20 46 45 80
180
Series resistance (Ω) 63
112
251
126
224
502
Q at 2.64 MHz
3.0
5.9
Parasitic capacitance (pF)
0.13
0.12
13.2
13.9
14.6
Self-resonance (MHz)
127
103 68
6.5
4.8
3.1
Bandwidth (MHz)
0.43
0.32
0.33
0.34
Voltage gain, w/o tuning cap.
0.017
0.022
0.034
0.033
0.043
0.066
Voltage gain, with tuning cap.
0.051
0.099
0.18
0.24
Coil parasitic capacitance between metal traces and between the spiral and a conducting layer.
Optical micrographs of initial Au coil samples on polyimide substrates with a design optimized by simulation results.
V. Conclusions, Future Work, and Acknowledgements:
Measured inductance and voltage gain on LTCC coils (screen printed Ag traces) are too low
A new coil design based on metal traces fabricated on a polyimide substrate and bonded to a LTCC ferrite platelet have been investigated
Simulated the polyimide coil designs; results suggest it will achieve sufficient inductance and voltage gain for integrated neural prosthetic devices (and in vivo microelectronics)
Fabricated initial Au coil on polyimide substrates, and achieved more than an order of magnitude more turns than the LTCC based coil architecture
Test Type I and Type II coil performance with the reference coil to measure inductance, voltage gain, resistance, capacitance and Q-factor
Refine the coil design to most efficiently supply the signal processor IC power requirements and minimize coil size
Investigate biocompatibility of the polyimide based architecture with encapsulation layer
We gratefully acknowledge NIH support (HHSN C / N01-NS ) and discussions with Prof. Normann’s, Prof. Harrison’s, and Prof. Solzbacher’s students
III. LTCC Coil Results and Discussion:
Coil inductances are very low and would require large capacitors (~10 nF) for 2.64 MHz operation with high Q
Large capacitors are not available in Surface Mount Device (SMD) component sizes compatible with device integration
A higher inductance is needed, which requires more turns
The voltage gain is very low, which would require driving the supply coil at dangerous voltages (> 100 Vrms) to achieve a 3 Vrms supply
Optical pictures of the initial LTCC ferrite coil samples shown with a dime for scale.