Fwd: Oral Exam Announcement: Daniel Pivonka

Daniel Pivonka pivonka at stanford.edu
Wed Mar 14 17:09:22 PDT 2012

---------- Forwarded message ----------
From: Student Services <studentservices at ee.stanford.edu>
Date: Fri, Mar 9, 2012 at 9:46 AM
Subject: [ee-doctorate] Oral Exam Announcement: Daniel Pivonka
To: ee-students at lists.stanford.edu

 *A mm-sized Wirelessly Powered and Remotely Controlled Locomotive Implant

*Daniel Pivonka
Department of Electrical Engineering
Advisor: Prof. Teresa Meng

Thursday, March 15th, 2012
3 pm (refreshments at  2:45 pm)
Paul G. Allen Auditorium (CIS-X 101)


*Fully autonomous implantable systems with locomotion can revolutionize
medical technology, serving applications ranging from diagnostics to
minimally invasive surgery.  In addition, they open up possibilities for a
variety of emerging technologies, allowing for new applications that were
previously impossible.  The key challenges in achieving such a device are
the limited power budget imposed by the miniaturization of the device, the
high power consumption of fluid propulsion techniques, and the need for
highly efficient data transfer and circuit design.  In addition, the device
must be bio-compatible and operate safely, with the primary risks arising
from RF exposure levels and the propulsion mechanism.

Most implantable devices require a battery or an inductive link as a power
supply, and with current technologies miniaturizing these powering methods
to the mm and sub-mm regime is not possible.  Additionally, existing
propulsion techniques often rely on piezoelectric materials and have
complicated designs, use significant power, and have increasingly low
thrust efficiency as they are scaled.  Alternatives use magnetic structures
in complex magnetic fields, but these devices move slowly even when used in
MRI.  In this work, a wireless powering method allows for a mm-sized
antenna to receive up to 500μW at depths up to a few cm in tissue.  Also,
new propulsion methods have been developed that employ electromagnetic
forces experienced by moving currents in a magnetic field, and these allow
for a simple design and efficient thrust generation.  Two methods operate
with this principle: the first uses magnetohydrodynamics (MHD), and the
second oscillates a structure that generates thrust from asymmetries in the
fluid drag characteristics.  Both methods have been evaluated via
incompressible Navier-Stokes fluid simulations, and the results show that
mm-sized devices can reach ~cm/sec speeds with ~mA of current.

To demonstrate these principles, a prototype chip was fabricated in the
TSMC 65nm process.  The entire system was integrated on chip with the
exception of a 2mm x 2mm receive antenna.  The power carrier is modulated
with a minimal depth AM pulse-width technique to reduce the impact on power
delivery.  The on-chip circuitry includes a matching network with adaptive
loading, a rectifier and regulator for the power supply, a demodulator for
clock and data recovery, and a digital controller to interface with the
propulsion system.  The chip functions with either propulsion method, and
can deliver 1-3mA from a 0.2V driver depending on the received power.
Propulsion dominates power usage, with the active circuitry consuming less
than 10% of the total power budget.  The 3mm x 4mm prototype achieves
.53cm/sec speeds in fluid with a .06T field using approximately 250μW, and
receives up to 25Mbps from a 2W 1.86GHz signal.

EE students mailing list
ee-students at lists.stanford.edu

ee-doctorate mailing list
ee-doctorate at lists.stanford.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://snf.stanford.edu/pipermail/labmembers/attachments/20120314/da47f480/attachment.html>

More information about the labmembers mailing list