Beruflich Dokumente
Kultur Dokumente
CONTROL
UNIT
POWER
MANAGEMENT
UNIT
PA
PA
FPGA
IMPLANT
acoustic
waves
6
DC/DC
PA
d
VBATT
COUT
PA
PA
PA
64
EXT. ENVIRONMENT
SUBCUTANEOUS
PATIENTs SKIN
Fig. 1.
Index TermsCMOS, energy harvesting, implanted medical device, power management, healthcare, sensor networks.
I. I NTRODUCTION
Requirements for active implanted medical devices
(IMDs) deeply introduced into the human body are ruled
by the implant size (1-3 cm3 ) and by the type of coupling
used to transmitt energy [1]. RF and magnetic induced
coupling are heavily attenuated in water, whereas ultrasound is a valid option to overcome this limit despite the
complexity to choose the carrier frequency [2].
The selection of the wavelength () for delivering energy is guided by the transducer design. For instance a
spherical array transducer enables to focus the ultrasound
(US) beam without beamforming electronics. However, to
track the implant when it goes out-of-focus, each array
element needs a power amplifer (PA) to shape the acoustic
field. A good candidate in energy transmission is the classE power amplifier with a theoretical power efficiency up
to 100 % [3].
However, it is more challenging the choice of the CMOS
technology used for energy conversion in the implant. To
avoid oxide breakdown, HV transistors are employed as
the input available voltage can be expected to vary from
few millivolts to tens of volt. In medical ultrasound as the
carrier frequency is setted around 1 MHz, two series LDO
are used to further improve supply noise rejection at low
frequency.
PA
2865
VOUT
VOUT - VGOAL < V2
VOUT > VGOAL
V1 = V1,0
V2 = V2,0
VOUT
VGOAL
Increase CP
VOUT
V2 = V2 + V2,0
Increase CS
Decrease CP
VOUT
START
V1 = V1 + V1,0
Decrease CP
END
ZVS condition
VDRAIN
ZVS
VDRAIN
Increase CSHUNT
I DS
VCC
VDRAIN = ZVS
VOUT = VGOAL
Decrease CS
VDRAIN
VOUT
3.66VCC
> ZVS
RFC
VDD
VDRAIN
LS
< ZVS
B. Measurements
CS
VOUT
IDS
Ct
CSHUNT
CP
LP
Rt
Fig. 2.
model.
w/o tuning
w/ tuning
25
VGOAL = 19 Vpeak
20
15
10
5
0
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
VOUT [Vp]
A. Tuning Methodology
Design equations have been developed for the class-E
power amplifier [4], henceforth in this section a tuning
strategy is proposed. The shunt-C class-E power amplifier
is designed with Advanced Design System (ADS) and
an average value for Rt and Ct is considered. 64-PA
are fabricated on standard substrate FR4 due to the lowfrequency requirement.
Figure 3 shows the tuning methodology adopted to
calibrate 64-PAs, two main loops can be distinguished to
set respectively the output voltage (VOUT ) of the PA to
the desired voltage (VGOAL ) and to achieve ZVS for best
efficiency. To set the value for the capacitors the magnitude
|VOUT VGOAL | is compared with V1 or V2 . The
initial conditions V1,0 , V2,0 for V1 , V2 are chosen
within the range [0.1 1] V and their values also dermine
the final maximum distance between VOUT and VGOAL .
Fig. 4.
tuning.
2866
(1a)
(1b)
vrec
vSS
m10
m8
m1
m2
m3
m4
m16
65
m12 m13
CSHUNT = 550 pF
Cp = 690 pF
CS = 1 nF
f0 = 1 MHz (DC = 50 %)
Lp = LS = 68 H
50
45
40
m14
6
VCC [V]
m15
R2
BATTERY MODEL
vBATT
vb4
ESR
Rself
CBATT
VDD = 4.5 V
A. PSRR Analysis
A methodology to analyze PSRR is proposed in [6] by
mean of control-theory. Figure 7 represents an equivalent
network of the two-series LDO to analyse PSRR at low
frequency. Hence, VOUT 1 and VBAT T can be expressed as
follows:
35
2
m18
Fig. 6.
DE
PAE
m17
Rf2
V3
vSS
vb2
R1
iOUT
60
55
Rf1
m9
V3
OTA2
vref
vout1
V1
m11
m7
vout1
vb3
70
Efficiency [%]
m5
vref
OTA1
75
Fig. 5.
vb1
m6
10
A2 (1 Ap1 ) + Ap2
A1 A2
Vref +
Vrec
(3a)
1 + 1 A1 A2
1 + 1 A1 A2
A4 (1 Ap3 ) + Ap4
A3 A4
=
Vref +
VOU T 1 (3b)
1 + 2 A3 A4
1 + 2 A3 A4
VOU T 1 =
VBAT T
R1
Where the attenuation factors are defined as 1 = R1 +R
f1
R2
and 2 = R2 +Rf 2 . The supply gains are written as
Ap1 = V1 /Vrec , Ap2 = VOUT 1 /Vrec , Ap3 = V3 /VOUT 1
and Ap4 = VBAT T /VOUT 1 . So, assuming 1 A1 A2 1
and 2 A3 A4 1, equations 3 can be rearragend as:
Vref
1
Vrec
+
1
1 P SRRLDO1
Vref
VOU T 1
1
+
2
2 P SRRLDO2
VOU T 1
(4a)
VBAT T
(4b)
1
1
1
1
P SRRLDO1 = A1 + A1 P SRR2 + P SRR1 and
1
1
1
1
P SRRLDO2 = A3 + A3 P SRR4 + P SRR3 . The parameters
1
2
, P SRR2 = AAp2
,
P SRRi are defined by P SRR1 = AAp1
A3
A4
P SRR3 = Ap3 and P SRR4 = Ap4 . By replacing
equation 4a into equation 4b, VBAT T can be expressed
Where
as follows:
VBAT T
1
2
1
1+
1 P SRRLDO2
Vref
1
P SRRLDO1 P SRRLDO2
Vrec
1
(5)
At low frequency any variation in Vrec is further suppressed at VBAT T by using two LDOs in series as shown
by the term (P SRRLDO1 P SRRLDO2 )1 .
B. Experimental Results
PSRR of the proposed LDO is measured using the
methodology proposed in [7]. At low frequency PSRR
2867
vrec
vref
A1
m10
OTA1
v1
vout1
A2
Rf1
TABLE I
S UMMARY OF LDO SPECIFICATIONS AND PERFORMANCE
R1
vSS
vout1
m17
OTA2
vref
A3
Parameter
Input voltage (V)
Load current (mA)
Output voltage (V)
Battery capacity (Ah)
Drop voltage (mV)
Ground current (A)
Load regulation @ VOU T 1 (mV /mA)
OU T 1
Line regulation V
(mV /V )
Vrec
Load regulation @ VBAT T (mV /mA)
BAT T
Line regulation V
(mV /V )
V
v3
vBATT
A4
Rf2
ESR
R2
Rself
vSS
CBATT
rec
V. C ONCLUSION
A novel system to transfer energy via ultrasound to
deep implanted medical devices has been proposed. The
driving electronic for the external transducers showed a
DE up to 71% for VCC = 3 V and it was capable of
delivering Iac = 23 W/cm2 at 11 cm by leading to a
theoretical link = 10.6%. Moreover, to recharge a microbattery in the implant, a linear dropout regulator was
fabricated in a CMOS 0.18 m HV technology. The LDO
was successfully tested and measured along with simulated
results were reported.
70
65
VOUT1
VBATT
PSRR [dB]
60
55
50
45
ACKNOWLEDGMENT
40
35
10 dB
30
25
0.1 100 200 300 400 500 600 700 800 900 1000
f [Hz]
(a)
85
R EFERENCES
75
[1] Xiaojuan Wei and Jing Liu. Power sources and electrical recharging
strategies for implantable medical devices. Frontiers of Energy and
Power Engineering in China, 2:113, 2008. 10.1007/s11708-0080016-3.
[2] A. Denisov and E. Yeatman. Ultrasonic vs. inductive power delivery
for miniature biomedical implants. In Body Sensor Networks (BSN),
2010 International Conference on, pages 84 89, june 2010.
[3] P. Reynaert and M. Steyaert. RF Power Amplifiers for Mobile
Communications. New York:Springer, 2006.
[4] M. Acar, A.J. Annema, and B. Nauta. Analytical design equations
for class-e power amplifiers. Circuits and Systems I: Regular Papers,
IEEE Transactions on, 54(12):2706 2717, dec. 2007.
[5] Mazzilli F. et al. In-vitro platform to study ultrasound as source for
wireless energy transfer and communication for implanted medical
devices. In EMBC 2010, pages 3751 3754, sept. 2010.
[6] M.S.J. Steyaert and W.M.C. Sansen. Power supply rejection ratio in
operational transconductance amplifiers. Circuits and Systems, IEEE
Transactions on, 37(9):1077 1084, sep 1990.
[7] Texas Instrument. Ldo psrr measurement simplified: Slaa414, 2009.
[8] Infinity Power Solutions. http://www.infinitepowersolutions.com.
65
PSRR [dB]
09 (Vrec )
10 (iOU T )
4.10.05 (VBAT T )
300 [8]
200 (= Vrec -VBAT T )
210
0.66 (simulated 0.6)
2.2 (simulated 0.8)
4.7 (simulated 1.6)
0.08 (simulated 0.03)
55
45
VOUT1
VBATT
VBATT (Cext = 10 F)
35
25
15
5
1.0e+04
2.5e+05
5.0e+05
f [Hz]
7.5e+05
1.0e+06
(b)
Fig. 8. Measured PSRR: (a) at low frequency [50 Hz : 1 kHz] and (b)
at high frequency [10 kHz : 1 MHz].
2868