betaFeedback

WorksEngineering

Polyimide-Based Flexible Coupled-Coils Design and Load-Shift Keying Analysis

Abstract

Wireless power transfer using inductive coupling is commonly used for medical implantable devices. The design of the secondary coil on the implantable device is important as it will affect the power transfer efficiency, the size of the implant, and also the data transmission between the implant and the in-vitro controller. In this paper, we present a design of the secondary coil on a polyimide-based flexible substrate to achieve high power transfer efficiency. Load shift keying modulation is used for the data communication between the primary and secondary coils. A thorough analysis is done for the ideal and practical scenario and it shows that a mismatched secondary LC tank will affect the communication range and communication correctness. A solution to achieve robust data transmission is proposed and then verified by SPICE simulations.

Primary CoilSecondary Coil
Outer Length llN/A14 mm
Self-inductance LL895 nH564 nH
Series resistance RsR_s1.114 Ω\Omega2.333 Ω\Omega
Quality factor @ 40.68 MHz205.3361.9

Table 1. Proposed Coupled-Coils Parameters

Introduction

Implantable medical devices (IMD) usually have a high constraint on the size, and thus wireless power transfer is widely used instead of the bulky batteries to provide power for the implants. A wirelessly powered trans-sclera electrical stimulation (TsES) system has been developed to treat retinal degenerative diseases [1]. However, the power transfer efficiency of the inductive link was only 8% at a distance of 10 mm, which may not adequate for in vivo operations. The secondary coil was fabricated on a printed circuit board (PCB), which was too rigid to be used inside the body. Moreover, implant biocompatibility could not be guaranteed. In this paper, to solve the above problem, a pair of polyimide-based flexible coupled-coils is proposed to solve the above problems.

To save the use of an additional antenna, power and data are transmitted via the same inductive coupling link. Frequency-shift keying (FSK) modulation is commonly used in IMDs applications to achieve high data rate [2], but using FSK will increase circuit complexity, and high data rate is not necessary for many applications. To minimize the system volume, load-shift keying (LSK) modulation is also commonly used [3]-[5]. However, it was found that the uplink data sent back from the implant to the transmitter is not as robust as expected [1], limiting the communication range. In this paper, the issue of using LSK scheme is studied and analyzed. A solution is then suggested and verified by SPICE simulations.

Polyimide-Based Flexible Coupled-Coils Design

IMDs are usually low-power applications and inductive power transfer using a pair of series-parallel (S-P) resonant coupled-coils is usually used. In [1] and [6], for a trans-sclera electrical stimulation system, the secondary coil is fabricated on a PCB with thickness of 0.3 mm, and after packaging with encapsulation, the implant thickness will be up to 1 mm.

Surgeons suggest to put it beneath the conjunctiva and rectus muscle, and the implant width should be less than 6 mm so as not to affect the activity of the eyeball.

Polyimide has been proven to be a biocompatible material and an excellent choice for neuroprosthetic application [7, 8]. In this research the polyimide-based secondary coil has been developed to overcome the size limitation and

improve the power transfer efficiency (PTE). The proposed secondary coil installation is shown in Figure 1.

Proposed secondary coil installation

Figure 1. Proposed secondary coil installation [9]

The two electrodes and coil are fabricated on the polyimide substrate, and the controller IC was mounted on the substrate as shown in Figure 1. Proposed secondary coil installation [9]. As the substrate is biocompatible, the biocompatible packaging is easier to achieve, and the implant thickness is reduced to 0.2 mm. With the flexible substrate, the implant can be attached to the surface of the eyeball so that the implant can be made larger without affecting the eyeball movement. The width is still kept at 6 mm, while the length is increased to 40 mm for better stimulation performance. The shape of the coil is oval to make better use of the area.

To increase the link efficiency, the primary coil could also be made on a flexible PCB, such that it can be attached onto the skin to minimize the distance between the coils. The secondary coil is bended and attached to the surface of the eyeball. Bending will change both the self-inductances and the coupling coefficient, and the field solver HFSS is used to simulate the coupled-coils under the bending situation. The simulation setup is shown in Figure 2. Parameters of the coupled-coils are shown in Table I. The primary has series resonance and the secondary has parallel resonance. The inductive link circuit model is shown in Figure 3.

Polyimide-based flexible coupled coils simulation setup

Figure 2. Polyimide-based flexible coupled coils simulation setup (a) flat case. (b) bending case

Circuit model of inductive link

Figure 3. Circuit model of inductive link

From [10], the inductor and capacitor values under the resonant condition is related by the following equation:

ωo=1Ls2Cs21−Ls2Cs2RL2(1)\omega_o = \frac{1}{\sqrt{L_{s2}C_{s2}}}\sqrt{1-\frac{L_{s2}}{C_{s2}R_L^2}} \tag*{(1)}

and the PTE is given by

η=PoutPin=k2Q1Q22(1+Q2α+k2Q1Q2)(α+Q2)(2)\eta= \frac{P_{out}}{P_{in}} = \frac{k^2Q_1Q_2^2}{\left(1+\frac{Q_2}{\alpha}+k^2Q_1Q_2\right)(\alpha+Q_2)} \tag*{(2)}
Q1=ωoLs1Rs1,Q2=ωoLs2Rs2,α≡ωoCs2RL(3)Q_1 = \frac{\omega_oL_{s1}}{R_{s1}},\quad Q_2 = \frac{\omega_oL_{s2}}{R_{s2}},\quad\alpha\equiv\omega_o C_{s2}R_L \tag*{(3)}

When both the primary and secondary were flat and 10 mm apart, the simulated PTE of the inductive link reached 39.33%. When both coils were bended, the self-inductance decreased to 562 nH and the coupling coefficient changed to 4.2% from 5.0% and the PTE dropped to 37%. The flexible coupled-coils can maintain high PTE even when bended. Ls2L_{s2} is 562 nH, and the resonant capacitor Cs2C_{s2} thus equals to 27 pF, which is fabricated on-chip. The on-chip capacitance Cs2C_{s2} cannot be fabricated with a very exact value and at the same time, is affected by the parasitic capacitor of the loading circuits. One effect of the mismatched Cs2C_{s2} is that the LSK scheme may malfunction at the weak coupling condition and robust data transmission will be compromised. In § 3, the issue of LSK scheme is analyzed and a solution is proposed to achieve robust communication.

Analysis of LSK Modulation Scheme

Ideal LSK Modulation Scheme

We first consider the LSK modulation under ideal situation. Figure 4 shows an AC voltage source driving the primary LC tank Ls1L_{s1} and Cs1C_{s1}. The secondary LC tank receives the coupled power and delivers to the implant unit. The equivalent loading resistance RLoadR_{Load} is 12.5 kΩ\Omega, and parasitic capacitor Cp2C_{p2} is assumed to be small and can be ignored. During backscattering, the implant unit shifts RSWR_{SW} to modify the loading resistance across the secondary coil. The primary coil input current is given by (4). The primary impedance ZpriZ_{pri} and the primary resonator impedance Z11Z_{11} were defined on (5) respectively.

Simplified block diagram of IMDs and reflected impedance model

Figure 4. (a) Simplified block diagram of IMDs. (b) Reflected impedance model

I1(jω)=VsZpri(4)I_1(j\omega)=\frac{V_s}{Z_{pri}} \tag*{(4)}
Zpri=Z11+Zeq,Z11=Rs1+jωLs1+1jωCs1(5)Z_{pri}=Z_{11}+Z_{eq},\quad Z_{11}=R_{s1}+j\omega L_{s1}+\frac{1}{j\omega C_{s1}} \tag*{(5)}

The equivalent impedance model is shown in Figure 4(b), and the equivalent impedance Zeq(jω)Z_{eq}(j\omega) is given by

Zeq(jω)=ω2M2Rs2+jωLs2+1/jωCs2∥RL(6)Z_{eq}(j\omega)=\frac{\omega^2M^2}{R_{s2}+j\omega L_{s2}+1/j\omega C_{s2}\mathbin{\|}R_L} \tag*{(6)}

ZeqZ_{eq} is real at the frequency ωo\omega_o.

For low-power application, the load resistor RLR_L would be large, for example, RL>100 ΩR_L>100\ \Omega or even larger than 1 kΩ\Omega, so that the factor Ls2/Cs2RL2L_{s2}/C_{s2}R_L^2 is much smaller than 1. As discussed in [11], there is only minor advantage in making the equivalent impedance real, while the analysis is made much simpler with good accuracy by designing Ls2L_{s2} and Cs2C_{s2} to satisfy ωo=1Ls2Cs2\omega_o=\frac{1}{\sqrt{L_{s2}C_{s2}}}.

We further assume that the output voltage is adequately filtered, which requires that ωoCs2RL≫1\omega_oC_{s2}R_L\gg1. Moreover, RLR_L is much larger than the parasitic resistor Rs2R_{s2}. Hence, the imaginary

part is negligible compared to the real part ReqR_{eq}, and ReqR_{eq} is given by

Req≈ω02k2Ls1Ls2(1+Q2QL)Rs2,QL=ω0Ls2RL(7)R_{eq} \approx\frac{\omega_{0}^{2} k^{2} L_{s1}L_{s2}}{(1+Q_{2}Q_{L})R_{s2}},\quad Q_{L}=\frac{\omega_{0}L_{s2}}{R_{L}} \tag*{(7)}

According to (7) ReqR_{eq} increases monotonously w.r.t. RLR_{L}. Ideally, backscattering changes ReqR_{eq} from RLR_{L} to RL∥RSWR_{L} \mathbin{\|} R_{SW}. The equivalent impedance ZeqZ_{eq} and ZpriZ_{pri} will decrease so I1I_{1} will increase. By detecting the amplitude of I1I_{1}, the transmitter side can decode the uplink signal.

The Secondary LC Tank Mismatch Effect

In practice, it is difficult to make Ls2L_{s2}, Cs2C_{s2} and RLR_{L} to satisfy (1). The receiver circuits such as the over-voltage protection (OVP) block and the rectifier will introduce parasitic capacitance Cp2C_{p2} to the secondary LC tank. The OVP circuit needs large diodes to sink excessive power, and the rectifier diodes have to be large enough to reduce conduction loss, and both together lead to a large Cp2C_{p2}. The secondary coil is designed to have high quality factor Q2Q_{2} to improve PTE, and Ls2L_{s2} is designed to be 564 nH and Cs2C_{s2} to be 27 pF. From actual measurement, the parasitic capacitance Cp2C_{p2} can be up to 12 pF [1], which is not negligible and affects the equivalent impedance.

The equivalent impedance Zeq(s)Z_{eq}(s) is shown in (10). Ls2L_{s2}, Cs2C_{s2} and RLR_{L} cannot satisfy (1) due to Cp2C_{p2}. Hence, ZeqZ_{eq} is not monotonously changing with RLR_{L}. Moreover, the primary impedance ZpriZ_{pri} is not monotonously increasing with RLR_{L}. During backscattering, the secondary coil switches from heavy load to light load and the change in the primary impedance ΔZpri\Delta Z_{pri} is given by (8):

ΔZpri=Zprilight−Zpriheavy(8)\Delta Z_{pri}=Z_{pri_{light}}-Z_{pri_{heavy}} \tag*{(8)}
ΔI1=I1light−I1heavy(9)\Delta I_{1}=I_{1_{light}}-I_{1_{heavy}} \tag*{(9)}

Without the parasitic capacitor, ΔZpri\Delta Z_{pri} is a monotonous function of RLR_{L}, and ΔZpri\Delta Z_{pri} is positive under all coupling condition. However, due to Cp2C_{p2} of the secondary LC tank, ZpriHZ_{priH} will be smaller than ZpriLZ_{priL} at the weak coupling condition. In the numerical example, the parameter set up were following the structure shown in Figure 4(a), Ls1L_{s1}, Ls2L_{s2}, Rs1R_{s1} and Rs2R_{s2} were set according to Table I. Cs1C_{s1} and Cs2C_{s2} were set to satisfy (1). RLR_{L} was 12.5 kΩ and RSWR_{SW} was set to 500 Ω. As shown in Figure 5, when kk is lower than 0.05, ΔZpri\Delta Z_{pri} is negative, which means that during backscattering when the secondary loading is switched from light load to heavy load, the amplitude of the primary coil current I1I_{1} will decrease instead of increasing, and according to (9) ΔI1\Delta I_{1} is positive at weak coupling. The theoretical results were shown in Figure 6. With the coupling coefficient increasing, ΔI1\Delta I_{1} will change from positive to negative, so the uplink signal generated by the LSK scheme will flip consequently.

Primary impedance magnitude difference versus coupling coefficient

Figure 5. Primary impedance magnitude difference versus coupling coefficient

Primary resonator input current magnitude difference versus coupling coefficient

Figure 6. Primary resonator input current magnitude difference versus coupling coefficient

The situation will be worse if there is also a mismatch in the primary LC tank. We change the simulation set up. The primary resonant capacitance Cs1C_{s1} is larger than the designed value by 1% and CpC_{p} is still 12 pF. The results of the stimulation are shown in Figure 5 and Figure 6, which are labeled as “Primary mismatch”. From the stimulation results, it can be seen that due to the primary mismatch, the uplink signal will flip at k=0.09k=0.09 instead of 0.05, and hence the effective communication distance will further decrease.

Zeq=ω2k2Ls1Ls2∗RL+Rs2+ω2C22RL2Rs2−jω(Ls2−C2RL2+ω2C22Ls2RL2)(RL+Rs2)2+ω2[Ls22+RL2C2(Rs22C2−2Ls2)],C2=Cs2+Cp(10)Z_{eq}=\omega^{2}k^{2}L_{s1}L_{s2} \ast\frac{R_{L}+R_{s2}+\omega^{2}C_{2}^{2}R_{L}^{2}R_{s2}-j\omega\left(L_{s2}-C_{2}R_{L}^{2}+\omega^{2}C_{2}^{2}L_{s2}R_{L}^{2}\right)}{(R_{L}+R_{s2})^{2}+\omega^{2}\left[L_{s2}^{2}+R_{L}^{2}C_{2}\left(R_{s2}^{2}C_{2}-2L_{s2}\right)\right]},\quad C_{2}=C_{s2}+C_{p} \tag*{(10)}

Simulation Results

The theoretical results are shown in Figure 7. Two cases were simulated by using SPICE to verify the theoretical results. For both cases, CpC_p was set to 12 pF, and kk at 0.06. For the proposed solution, Cs1C_{s1} is set at 17.03 pF, which is smaller than the designed value by 1%. The switch was driven by the “SW” signal, and it is active high. When the switch turns on, the secondary loading will change from light load to heavy load, the primary current I1I_1 should increase but as shown in Figure 8(a), I1I_1 was larger when the secondary was at light load. The flipping issue has been eliminated in Figure 8(b).

Detuned primary LC tank input current magnitude difference versus coupling coefficient

Figure 7. Detuned primary LC tank input current magnitude difference versus coupling coefficient

Simulation results for flipping and potential solution

Figure 8. Simulation results (a) flipping case. (b) potential solution

The secondary parasitic capacitance CpC_p will introduce an negative imaginary part into the equivalent impedance ZeqZ_{eq}. However, if Cs1C_{s1} is made larger than the designed value, Z11Z_{11} will contain a positive imaginary part that could be used to cancel the imaginary part of ZeqZ_{eq}, so that ΔZpri\Delta Z_{pri} may become negative at weak coupling.

To avoid cancelling between ZeqZ_{eq} and Z11Z_{11}, the primary resonator could be detuned. By making Cs1C_{s1} lower than the designed value, Z11Z_{11} was guaranteed to have a negative imaginary part, such that the uplink signal flipping issue could be eliminated.

Conclusions

In this work, a polyimide-based flexible coupled-coils was developed for the TsES system, and the inductive link has been modeled in HFSS and fabricated by FPC. The PTE was 39% when the coils were separated by 10 mm. LSK scheme with parasitic parameter has been analyzed and a solution was proposed.

Acknowledgment

This research is in part sponsored by ….

[1][12][13]This work
Link structure2-coil2-coil2-coil2-coil
Substrate materialFR4FR4PDMSPolyimide
Operation frequency (Hz)40.68 M5 M4 M40.68 M
Relative distance (mm)10101210
Outer diameter of primary coil (mm)20703724
Outer diameter of Secondary coil (mm)4.42016×106×14
Distance/secondary coil diameter2.30.511.6
Sim. or Meas.Meas.Meas.Meas.Sim.
PTE@ perfect alignment8%85.8%21%39%

Table 2. Inductive Links Comparison

References

  1. [1]X. Li et al., "Optic Nerve Stimulation System with Adaptive Wireless Powering and Data Telemetry," Micromachines, vol. 8, (12), pp. 368, 2017.DOI
  2. [2]M. Ghovanloo and K. Najafi, "A wideband frequency-shift keying wireless link for inductively powered biomedical implants," IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 51, (12), pp. 2374-2383, 2004.DOI
  3. [5]S. Mandal and R. Sarpeshkar, "Power-efficient impedance-modulation wireless data links for biomedical implants," IEEE Transactions on Biomedical Circuits and Systems, vol. 2, (4), pp. 301-315, 2008.DOI
  4. [6]L. Cheng et al, "A 40.68-MHz Active Receiver With Hybrid Adaptive On/Off Delay-Compensation Scheme for Biomedical Implantable Devices," IEEE Transactions on Circuits and Systems I: Regular Papers, 2019.DOI
  5. [7]P. J. Rousche et al, "Flexible polyimide-based intracortical electrode arrays with bioactive capability," IEEE Transactions on Biomedical Engineering, vol. 48, (3), pp. 361-371, 2001.pubmed.ncbi.nlm.nih.gov/11327505
  6. [8]R. R. Richardson Jr, J. A. Miller and W. M. Reichert, "Polyimides as biomaterials: preliminary biocompatibility testing," Biomaterials, vol. 14, (8), pp. 627-635, 1993.pubmed.ncbi.nlm.nih.gov/8399958
  7. [9]R. S. Snell and M. A. Lemp, Clinical Anatomy of the Eye. 2013.
  8. [10]K. Van Schuylenbergh and R. Puers, Inductive Powering: Basic Theory and Application to Biomedical Systems. 2009.
  9. [11]Y. Lu and W. H. Ki, CMOS Integrated Circuit Design for Wireless Power Transfer, Chapter 3, Analysis of Coupled Coils, Springer, 2017.DOI
  10. [12]U. Jow and M. Ghovanloo, "Design and optimization of printed spiral coils for efficient transcutaneous inductive power transmission," IEEE Transactions on Biomedical Circuits and Systems, vol. 1, (3), pp. 193-202, 2007.
  11. [13]A. Qusba et al, "On the design of microfluidic implant coil for flexible telemetry system," IEEE Sensors Journal, vol. 14, (4), pp. 1074-1080, 2014.DOI

Paper details

vlnc:W_n53e7gvm
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Engineering

Not recorded for this paper

View statement
Looking for related papers…
Contents