kuleuven poster .pdf

A Potassium Magnetometry Based Current Source for the nEDM
experiment at PSI
∗
Koss ,

Peter A.
G. Bison, V. Bondar, C. Crawford, E. Wursten and N. Severijns
∗KU Leuven, Instituut voor Kern- en Stralingsfysica
Magnetometry

Motivation
Our neutron electric dipole moment (nEDM) experiment is situated at the Paul Scherrer Institute (PSI). The
nEDM measurement procedure uses the Ramsey method of time separated oscillatory fields. This method
requires a very stable ~B0-field during a measurement cycle (180s) [1]. Ideally, the ~B0-field should stay stable
on the scale of the ~E -field reversal time (an hour). At the moment we have several ways to control the field
stability:
We shield the setup with a Mu-metal shield (shielding factor ≈ 104).
We use a surrounding field compensation system to diminish external field perturbations.
We monitor the field in our setup with Mercury and Cesium magnetometers.
We have a very stable current source for the main field (~B0) of our experiment.
The stability of our present current source is 10−7 on 17 mA. We would like to improve on that by exploiting
the sensitivities of atomic magnetometers (easily 10−8) and convert them to a current stability [2].

Idea
feedback control

DAQ
Magnetometer design. Our Potassium magnetometers use free spin precession (FSP) signals [4]. A free
spin precession may be initiated by AM-pumping or by an RF-pulse. Both schemes are possible in the module
shown on the left. The magnetometer module is designed to fit in the field confining coil shown previously.

I/V

B

pump
nEDM
B0-coil

Mu-metal
shield

source coil

K vapor
cell

magnetometers

Current source concept. A commercial low noise current source provides the base current (e.g.
Magnicon). Changes in this current are registered as changes in the magnetic field in a dedicated coil which is
connected in series to the nEDM B0-coil. A magnetometer array in the source coil is able to detect these
changes. A dedicated data acquisition system (DAQ) generates an appropriate feedback response for the
detected drift [3].

2 £ 10 2

6

2 £ 10 2

5

1 £ 10 2

4

1 £ 10 2
9 £ 10 1

3

7 £ 10 1

2

5 £ 10 1

1

3 £ 10 1
1

2

Longitudinal (cm)

3

6 × 10 −5

2

4 × 10 −5

1

2 × 10 −5

0

−2 × 10 −6

1

−2 × 10 −5

2

−4 × 10 −5

3

−6 × 10 −5

4

−8 × 10 −5

5

6

7

8

9

10

Transverse (cm)

11

sin

LPF
CORDIC
atan

cos

LPF

magnitude
phase

ASD (pT)

8 × 10 −5

ADC
signal

Normalized difference, B B−cBc

4

ref. phase

1 × 10

4

RF-pulsed mode. A circularly polarized DC laser beam pumps the medium along the magnetic field. A
weaker DC probe beam is maintained perpendicular to the pump beam. An RF-pulse flips the polarization
into a dynamic configuration which results in a free spin precession (signal shown on the left).

Field confining coil
−4

1 £ 10 1

Probe power, PProbe (¹A)

Online vs Offline

102

ROI at h = 12.5 cm

3

Hz

7

fT

2 £ 10 2

Sensitivity, ¾CRLB (p

Brf

probe

8

)

RF-FSP sensitivity at 1 ¹T and 48± C

Pump power, PPump (¹A)

current
source

B

offline fitted data
online demodulated

101

100

10

1

100

101

102

Integration time, (s)

103

Data acquisition system. We intend to get a live readout of the frequency information in the FSP signal.
For this we use a lock-in algorithm which is implemented on an FPGA. Such a system can be scaled to
readout the hundreds of FSP signals as they will be generated in the n2EDM setup. The graph on the right
shows an Allan plot of the measured data which characterizes the frequency stability of the readout.

Outlook

12

Scalar potential method for coil design. The design method is based on the magnetic scalar potential
Φ. The blue lines on the left are isopotentials of Φ but may be seen as the top view of a current loop. The
righthand graph shows a typical region of interest (ROI, black rectangle on the left) field map. The values are
normalized to the value at the center of the ROI, Bc .

The PCB field confining coil is the heart of this feedback controlled current source. Unlike in [2] we hope to
only feedback control on the current drift while external field perturbations are discriminated by the coil
geometry. In this way we hope that the stability of the current source is only limited by the sensitivity of the
magnetometer.

References
[1] C. Baker, G. Ban, K. Bodek, M. Burghoff, Z. Chowdhuri, M. Daum, M. Fertl, B. Franke, P.
Geltenbort, K. Green et al., The search for the neutron electric dipole moment at the Paul Scherrer
Institute, Phys. Procedia, 17, 159 (2011).
[2] V. Y. Shifrin, C. G. Kim and P. G. Park, Atomic magnetic resonance based current source, Rev. Sci.
Instrum. 67, 833 (1996).
Printed circuit board (PCB) coil. The picture on the left shows a PCB coil which was presented in [3].
The picture on the right shows the top PCB panel of the coil. This new type of coil is a self supporting
structure which uses large multi-layered PCBs. The coil can be opened and closed via non-magnetic PCB
headers.
KU Leuven

[3] P. A. Koss, C. Crawford, G. Bison, E. Wursten, M. Kasprzak and N. Severijns, PCB Coil Design
Producing a Uniform Confined Magnetic Field, IEEE Magn. Lett. PP, 1 (2017).
[4] Z. D. Grujić, P. A. Koss, G. Bison and A. Weis, A sensitive and accurate atomic magnetometer based
on free spin precession, Eur. Phys. J. D 69, 135 (2015).
peter.koss@kuleuven.be