See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/320456068

Design of PCB search coils for AC magnetic flux density measurement

Article in AIP Advances · April 2018

DOI: 10.1063/1.4991643

CITATIONS READS
9 2,158

1 author:

Michal Ulvr
Czech Metrology Institute
33 PUBLICATIONS 125 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

NanoMag (Nano-scale traceable magnetic field measurements) View project

All content following this page was uploaded by Michal Ulvr on 17 October 2017.

The user has requested enhancement of the downloaded file.

Design of PCB search coils for AC magnetic flux density measurement
Michal Ulvr

Citation: AIP Advances 8, 047505 (2018); doi: 10.1063/1.4991643

View online: http://dx.doi.org/10.1063/1.4991643
View Table of Contents: http://aip.scitation.org/toc/adv/8/4
Published by the American Institute of Physics
 AIP ADVANCES 8, 047505 (2018)

Design of PCB search coils for AC magnetic

flux density measurement
Michal Ulvr1,2,a
1 Department of Electromagnetic Quantities, Czech Metrology Institute,
Brno CZ-63800, Czech Republic
2 Department of Measurement, Czech Technical University in Prague,

Prague CZ-16627, Czech Republic

(Received 22 June 2017; accepted 8 September 2017; published online 17 October 2017)

This paper presents single-layer, double-layer and ten-layer planar square search coils
designed for AC magnetic flux density amplitude measurement up to 1 T in the
low frequency range in a 10 mm air gap. The printed-circuit-board (PCB) method
was used for producing the search coils. Special attention is given to a full char-
acterization of the PCB search coils including a comparison between the detailed
analytical design method and the finite integration technique method (FIT) on the
one hand, and experimental results on the other. The results show very good agree-
ment in the resistance, inductance and search coil constant values (the area turns)
and also in the frequency dependence of the search coil constant. © 2017 Author(s).
All article content, except where otherwise noted, is licensed under a Creative
Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://doi.org/10.1063/1.4991643

I. INTRODUCTION
AC magnetic flux density amplitude Bm up to 1 T can be generated by an AC electromagnet with
an air gap in the low frequency range (or up to 10 kHz for generating lower Bm value). This high
Bm value is needed for magnetometer calibration. The set Bm value in the air gap must be precisely
measured to ensure the traceability of the calibrated magnetometer. The traceability will be ensured
by a search coil with a voltmeter. When a search coil is used, the area turns (the search coil constant)
is very important for calculating the Bm value. Its inductance value or resonance frequency value,
which is determined by its impedance, is important for determining the frequency dependence of
the search coil. As the estimated electromagnet air gap length is 10 mm and the homogeneity of
the magnetic flux density within an area of (60x60) mm, the search coil that is used must be thin,
with the highest area turns value within an area of about (10x10) mm and with a low inductance
value. Hand-wound search coils are the most widely-used type of flat search coils. Their area turns
value can vary from 1 cm2 to 100 cm2 with an outer diameter up to 10 mm. However, the thickness
of these search coils varies from 1 mm to 2 mm and they are only circular in shape. Examples of
hand-wound flat search coil parameters can be found in Table I. Integrated planar coils have many
applications in sensors and in MEMS devices,1–3 and also in RF and microwave applications,4,5 in
biomedical applications,6–8 and in power applications.9–11 If photolithography is used for fabricating
planar search coils, a turn/space less than 20 µm in width (down to a few µm) can be achieved using
this technology. A higher area turns value with a smaller diameter than the PCB search coil and
with much smaller thickness than on a hand-wound search coil can be achieved with this technology.
However, the thickness of the copper will be reduced, and this rapidly increases the resistance of
the coil. And as the turns will be closer to each other, the inductance value would also be higher.
Another disadvantage of this method is its high price, which implies that usually only a single-layer
or double-layer planar coil can be fabricated. In recent years, additive manufacturing (3D printing

a
Electronic address: mulvr@cmi.cz

2158-3226/2018/8(4)/047505/9 8, 047505-1 © Author(s) 2017

047505-2 Michal Ulvr AIP Advances 8, 047505 (2018)

TABLE I. Parameters of the examples of hand-wound flat search coils.

Nominal Outer diameter of the

Manufacturer Type constant value (cm2 ) winding (mm) Thickness (mm) Resistance (Ω)

LakeShore FNT-5P04-30 30 3.9 1.4 110

FNT-6R04-100 100 10.4 1.57 6.5
Norma 234 003 5.18 2.5 1 29.2
234 004 100.0 11.2 1 28.9

technology) can also be used for the planar coils fabrication.12–14 The printed-circuit-board (PCB)
method is limited by the technology that is used. The minimal dimension of the turn/space is 100 µm
and the smallest diameter of the drilled hole is also 100 µm.
However, the method is cheap, simple, it has good repeatability, uniform cross sections, and a
planar coil with more than 10 layers can be fabricated. New planar single-layer, double-layer and
10-layer search coils for Bm measurements were designed and produced by the PCB method, and are
characterized in this work.

II. DESIGN AND CHARACTERIZATION OF THE PCB SEARCH COILS

A. Analytical method
The two most widely-used types of PCB coils are shown in Fig. 1. The area turns value of a
single-layer square coil according to Fig. 2 a) is calculated from
N
X
S1L = D2 + (D − xi)2 , (1)
i=1

where D is outer diameter, N is number of windings, and x is the difference of the sides of the square
(Fig. 1). The value of D can be calculated as15
D = d + 2 Nw + 2(N - 1)s, (2)
where d is inner diameter, w is turn width and s is space width. For w = s there applies x = w + s.
Eq. (1) can be expressed as
N +1 1
S1L = D2 + ND2 − 2xD N + x 2 N (N + 1) (2N + 1) . (3)
2 6
Formula (1) can also be used for a circular single-layer PCB coil only multiplied by π:
 XN 
S = π R2 + (R − xi)2  , (4)
 i=1

where R = D/2. It is obvious that the area turns value of a square PCB coil is higher than the area
turns value of a circular PCB coil, so henceforth only square PCB coils will be considered here. It
is rather more complicated to calculate the area turns value of the double-layer PCB coil according

FIG. 1. Dimensions of the planar square and circular coils.

047505-3 Michal Ulvr AIP Advances 8, 047505 (2018)

FIG. 2. Layout of a) a single-layer PCB search coil (left) with d=200 µm and b) a double-layer and 10-layer PCB search
coil (right) with d=400 µm. The width of the turn/space is 100 µm and the number of turns is 25. Red wire=top layer, blue
wire=bottom layer.

Fig. 2. b), because the calculation must be separated into the sum of the areas of the squares and the
rectangles. For the bottom half of the area S 1 applies
N
X 2
S1 = D2 + (D − 2xi)2 = D2 (N + 1) − 2N (N + 1) x D + N (N + 1) (2N + 1) x 2 . (5)
i=1
3
And for the top half of area S 2 applies
N " #
X 5
S2 = D (D − x) + (D − 2xi) (D − 3ix) = D (N + 1) − N (N + 1) + 1 xD + N (N + 1) (2N + 1) x 2
2

i=1
2
(6)
The total area turns value of the double-layer PCB rectangle coil is calculated as
" #
9 5
S2L = 2 (S1 + S2 ) = 2D2 (N + 1) − N (N + 1) + 1 xD + N (N + 1) (2N + 1) x 2 . (7)
2 3
The area turns value of the multi-layer PCB coil is then determined simply from the formula (7) by
multiplying the number of layers/2. Formulas (1)–(7) can be used only for approximate calculations
of the area turns value.
Next, the resistance and the inductance value must be determined. The resistance of the coil is
calculated from the equation
l
R= ρ , (8)
wt
where t is the thickness of the copper, ρ is copper resistivity (0.0172.10-6 Ω· m) and l is coil length,
which is calculated from12
l = 2(D + w) + 2N(2N - 1)(w + s). (9)
An important parameter of the PCB search coil is the inductance of the winding. Various empirical
formulas can be applied for the rectangular PCB coil inductance value.15,16 In this case, the following
two formulas were applied for calculating the inductance value of the single-layer coil
" # !2
 2.07 (D + d) 0.18 (D − d) D−d 
LS = 0.3175µN 2 (D + d) 
 
ln + + 0.13  (Rosa) (10)
 D−d D+d D+d 
 
D+d
LS = 1.17µ0 N 2 (Wheeler), (11)
1 + 2.75 D−d
D+d
where Am is the total copper area and Atot is the total area of the inductor (coil). For these two
parameters, we can apply
Am = wl (12)
Atot = D .
2
(13)
The inductance value of the double-layer coil was calculated with the coupling factor Kc and
formula17
047505-4 Michal Ulvr AIP Advances 8, 047505 (2018)

TABLE II. The structure of the 10-layer PCB search coil.

1st layer coil 18 µm

substrate 100 µm
2nd layer
coil 11.7 µm
substrate 100 µm
3rd layer
coil 11.7 µm
substrate 100 µm
4th layer
coil 11.7 µm
substrate 100 µm
5th layer
coil 11.7 µm
substrate 100 µm
6th layer
coil 11.7 µm
substrate 100 µm
7th layer
coil 11.7 µm
substrate 100 µm
8th layer
coil 11.7 µm
substrate 100 µm
9th layer
coil 11.7 µm
substrate 100 µm
10th layer
coil 18 µm

L = L1 + L2 + 2Kc L1 L2 ,
p
(14)

N2
Kc = , (15)
0.184X 3 − 0.525X 2 + 1.038X + 1.001 1.67N 2 − 5.84N + 65 0.64

where the self inductance values L 1 and L 2 (L 1 = L 2 for the same coils on both layers) of each layer
is calculated according to (10) or (11), and X is the thickness of the board on which the coils are (i.e.
the distance between the two coils). The inductance value of the 10-layer PCB search coil, according
to the structure in Table II., can be calculated by the modified formula17

L = 10LS + (4Kc1 + 14Kc2 + 4Kc3 + 12Kc4 + 4Kc5 + 10Kc6 + 4Kc7 + 8Kc8

, (16)
+ 4Kc9 + 6Kc10 + 4Kc11 + 4Kc12 + 4Kc13 + 2Kc14 + 4Kc15 + 2Kc16 )LS
The coupling factor values are as follows: K c1 = 0.9, K c2 = 0.905, K c3 = 0.829, K c4 =0.833,
K c5 = 0.774, K c6 = 0.777, K c7 = 0.731, K c8 = 0.729, Kc9 = 0.695, K c10 = 0.697, K c11 = 0.664,
K c12 = 0.666, K c13 = 0.638, K c14 = 0.640, K c15 = 0.615 and K c16 = 0.592.
The frequency dependence of the PCB search coil constant K S can be determined approximately
from18
f2
!
KSAC = KSDC 1 + 2 , (17)
fr
where K SDC is the search coil constant (m2 ), f is the frequency of the magnetic flux density generated
in the AC electromagnet (Hz), and f r is the resonance frequency value of the search coil (Hz). The
change of K SAC from K SDC is 1% for f /f r = 0.1, and 4% for f /f r = 0.2. Formula (17) is then valid
only in its approach for small changes of frequency f from resonance frequency f r .

B. Finite integration technique method

Single-layer, double-layer and 10-layer PCB search coils were designed and simulated in the CST
Studio environment by the finite integration technique (FIT).19 Search coil windings were designed
with parameters and layout according to Fig. 2 on the assumed PCB FR-4 substrate with dimensions
of (12x12) mm. A low-frequency (LF) domain solver with full wave analysis and adaptive mesh
refinement to achieve higher accuracy was used for the simulations. A stair backward wire was used
in the design of the single-layer search coil (in the bottom layer) to close the square loops in the top
layer (Fig. 3). This is also important for the analytical calculation of the area turns value according to
047505-5 Michal Ulvr AIP Advances 8, 047505 (2018)

FIG. 3. Model design of the single-layer PCB search coil in the CST Studio – top layer (left), bottom layer (right).

formula (3). The inductance value L was calculated from the energy W stored in the coil for a given
current I as
2W
L= 2. (18)
|I |

The area turns value K s or its frequency dependence was calculated from the induced voltage U RMS
on a resistor of 1 MΩ - when the peak magnetic field strength value H=112.5 A/m (BRMS =0.1 mT)
was applied perpendicular to the search coil (on the z axis) - according the formula

FIG. 4. Resistor shielding design: total view (top), cross section (middle) and the effect of shielding during the simulations
when H peak =10 kA/m was applied (bottom).
047505-6 Michal Ulvr AIP Advances 8, 047505 (2018)

|URMS |
KS = (19)
2πBRMS f

A 1 MΩ resistor represents the inner resistance of a common multimeter. However, the resistor
caused a serious error in the area turns calculations when it was exposed to the applied magnetic field
strength (the area turns value was closed to zero). The shielding of the resistor was therefore designed
(Fig. 4) with these parameters: conductivity of 5.108 S/m, relative permeability of 50000, and a hollow
cylindrical shape.

III. MEASUREMENT RESULTS

A single-layer (according to Fig. 2. a)), double-layer and 10-layer (according to Fig. 2. b)) search
coil were designed and produced on an FR-4 type PCB board type 10 cm in length, 1.5 cm in width,
and 0.5 mm thickness (the 10-layer coil is 1.2 mm in thickness), see Fig. 5. The dimensions of the
search coils are: w = s = 100 µm, N = 25, t = 18 µm, d(1-layer) = 196 µm, d(2-layer) = 398 µm
and d(10-layer) = 450 µm. Copper thickness of t=18 µm applies to the single-layer and double-layer
search coils, and to the 10-layer coil only for the first layer and for the last layer. In the remaining
8 layers it is 65% of this value (11.7 µm according to Table II.). The diameter of the drilled holes is
100 µm and the restring is 75%. The top and bottom wire from the coil to the connector are conducted
one above the other to avoid an additional loop turn. The calculated and measured parameters of the
fabricated PCB search coils are presented in Table III. The measured parameters of the hand-wound
NORMA search coils are given in Table IV. for comparison with the parameters of the fabricated PCB
coils. It is clear from Table III. that there is good agreement between the measured resistance value,
the calculated resistance value according to (8) and the simulated value from the CST Studio. Only
the measured resistance value of the single-layer PCB search coil is much higher than the calculated
value and the simulated value. This is probably caused by the PCB manufacturing process due to
the reproduction tolerances.20 The effect of the via to the area turns value can be neglected, because
the area of one via is only 0.04 mm2 . The thickness of the copper and the trace width have a very
big influence on the resistance and inductance values. As concerns the inductance of the PCB search
coils, there is very good agreement between the calculated values according to (10) or according to
(11) and (14)-(16), the CST Studio values, and the values measured by an LCR bridge. A relative
difference of less than 8% was determined for the single-layer PCB coil, a difference of less than
3.2% for the double-layer PCB coil, and a difference of less than 6.2% for the 10-layer PCB coil.
The PCB search coil constant value was calibrated using the variable mutual inductance method,21
with a DC electromagnet used as the magnetic flux density standard. The search coil DC constants
were calibrated for B=1 T with expanded uncertainty of 0.08%. The differences from the calculated
DC value according to (3) and according to (7), the simulated values from the CST Studio and the
measured values are less than 0.7% (for the single-layer search coil), less than 1.8% (for the double-
layer search coil) and less than 0.8% (for the 10-layer search coil). This also indicates very good
agreement.

FIG. 5. Fabricated PCB search coils: single-layer SC1L (top), double-layer SC2L (in middle) and 10-layer SC10L (bottom).
047505-7 Michal Ulvr AIP Advances 8, 047505 (2018)

TABLE III. Calculated and measured parameters of fabricated PCB search coils.

Value
Parameter
SC1L (single-layer) SC2L (double-layer) SC10L (10-layer)

Calculated KS value (cm2 ) 8.57 16.03 81.4

KS value from CST Studio (cm2 ) 8.58 16.41 81.0
Measured KS value (cm2 ) 8.52 16.31 81.6
Calculated resistance (Ω) 4.9 9.75 68.9
Resistance from CST Studio (Ω) 4.7 9.8 63.4
Measured resistance (Ω) 6.2 9.6 62.7
Calculated inductance (Wheeler) (µH) 2.57 9.44 221
Calculated inductance (Rosa) (µH) 2.69 9.84 229
Inductance from CST Studio (µH) 2.82 9.57 223
Measured inductance (µH) 2.79 9.75 210
Measured resonance frequency (MHz) 7.3 3.8 3.1
Coil diameter D (mm) 9.99 10.2 10.25
Coil length l (m) 0.51 1.02 5.1

TABLE IV. Measured parameters of hand-wound NORMA search coils.

Nominal constant Measured constant Resonance

value (cm2 ) value (cm2 ) Diameter (mm) Resistance (Ω) Inductance (µH) frequency (MHz)

5.18 5.1570 2.5 29.2 63.2 3.1

100.0 99.7997 11.2 28.9 287.3 3.0

A comparison of the calculated (according to (17)), simulated and measured frequency depen-
dence of the PCB search coil constant is shown in Fig. 6, Fig. 7 and Fig. 8. A single-layer
Helmholtz type solenoid22 and multimeter 3458A type was used for measuring the search coil con-
stant frequency dependence. The search coil constant was calculated for a given frequency and
BRMS =0.1 mT according to (19) and the expanded uncertainty of measured frequency dependence
was 0.3% up to 50 kHz. The differences from the calculated frequency dependence, the simu-
lated frequency dependence from the CST Studio and the measured frequency dependence are
less than 1%. It is evident that all three PCB search coil types are frequency independent up to
20 kHz.

FIG. 6. Frequency dependence of the fabricated single-layer PCB search coil constant.
047505-8 Michal Ulvr AIP Advances 8, 047505 (2018)

FIG. 7. Frequency dependence of the fabricated double-layer PCB search coil constant.

FIG. 8. Frequency dependence of the fabricated 10-layer PCB search coil constant.

IV. CONCLUSION
The design of a single-layer, double-layer and 10-layer PCB search coil for AC magnetic flux
density measurement has been proposed, and a full characterization has been presented. The design
of the PCB search coils was adapted for use in an AC electromagnet air gap of 10 mm. The analytical
design and the FIT design were compared with the fabricated PCB search coils. A single-layer PCB
search coil with a constant value of 8.52 cm2 and 0.5 mm in thickness, a double-layer PCB search
coil with a constant value of 16.31 cm2 and 0.5 mm in thickness, and a 10-layer PCB search coil with
a constant value of 81.6 cm2 and 1.2 mm in thickness, were fabricated. PCB search coils can have a
better value of certain parameters (thickness, inductance vs. area turns, diameter) than the parameters
of hand-wound search coils. However, it was necessary to fabricate the 10-layer PCB search coil in
order to achieve a higher area turns value (approaching to the nominal value of 100 cm2 of hand-wound
search coils). Thickness of the 10-layer PCB search coil is then comparable with the thickness of
hand-wound search coil. The results show very good agreement between the theoretical (calculated),
 047505-9 Michal Ulvr AIP Advances 8, 047505 (2018)

simulated (FIT) and measured values of resistance, inductance and the search coil constant. It has
also be proven that the search coil constants of all three PCB coil types are frequency independent
up to 20 kHz. The relative difference of the constant of the 10-layer PCB search coil is about 1.6%
at a frequency of 50 kHz. The fabricated search coils can therefore easily be used for AC magnetic
flux density measurements in the given frequency range up to 10 kHz.

ACKNOWLEDGMENTS
The author would like to cordially thank P. Hazdra from the Department of Electromagnetic
Field, Faculty of Electrical Engineering, Czech Technical University in Prague, for his assistance
with the simulations in CST Studio software.
1 R. E. Pawinanto, J. Yunas, M. M. Said, M.M. Noor, and B. Y. Majlis, “Design and Fabrication of PCB Based Planar
Micro-coil For Magnetic MEMS Actuator,” Proceedings of IEEE International Conference on Semiconductor Electronics,
Kuala Lumpur, Malaysia, 27.-29. August 2014 (IEEE, 2014), pp 487–490.
2 S. Duric, L. Nad, B. Biberdžic, M. Damnjanovic, and Lj. Živanov, “Planar Inductive Sensor for Small Displacement,”

Proceedings of 26th International Conference on Microelectronics, Niš, Serbia, 11.-14. May 2008 (Electron Devices Society
of IEEE, 2008), pp. 345–348.
3 T. M. Liakopoulos and C. H. Ahn, “A micro-fluxgate magnetic sensor using micromachined planar solenoid coils,” Sensors

and Actuators 77, 66–72 (1999).

4 B. Y. Liu, H. W. Zhang, X. L. Tang, and H. Su, “RF Integrated Planar Inductor Fabricated on Ferrite Substrates,” Proceedings

of 2006 International Conference on Communications, Circuits and Systems, Guilin, China, 25.-28. June 2006 (IEEE, 2006),
pp. 2623 – 2624.
5 J. N. Burghartz and B. Rejaei, “On the design of RF spiral inductors on silicon,” IEEE Trans. Electron Devices 50, 718–729

(2003).
6 A. B. Islam, S. K. Islam, and F. S. Tulip, “Design and optimization of printed circuit board inductors for wireless power

transfer system,” Circuits and Systems 4, 237–244 (2013).

7 J. Uei-Ming and M. Ghovanloo, “Design and optimization of printed spiral coils for efficient transcutaneous inductive power

transmission,” IEEE Trans. Biomed. Circuits Syst. 1, 193–202 (2007).

8 Y. Mi, X. Tang, S. Rui, Y. Chu, C. Bian, C. Yao, and C. Li, “Development of High dB/dt pulsed magnetic field generator

based on printed circuit board archimedes spiral coil for biomedical applications,” IEEE Trans. Plasma Sci. 44, 1879–1887
(2016).
9 E. J. Brandon, E. E. Wesseling, V. Chang, and W. B. Kuhn, “Printed microinductors on flexible substrate for power

applications,” IEEE Trans. Compon. Packag. Technol. 26, 517–523 (2003).

10 E. Waffenschmidt, B. Ackermann, and M. Wille, “Integrated ultra thin flexible inductors for low power converters,” in

Proceedings of IEEE 36th Power Electronics Specialists Conference, Recife, Brazil, 12.-16. June 2005 (IEEE, 2005),
pp. 1528–1534.
11 M. Ludwig, M. Duffy, T. O’Donnell, P. McCloskey, and S. C. Ó. Mathùna, “PCB integrated inductors for low power DC/DC

converter,” IEEE Trans. Power Electron. 18, 937–945 (2003).

12 Y. Yan, Ch. Ding, K. D. T. Ngo, Y. Mei, and G.-Q. Lu, “Additive Manufacturing of Planar Inductor for Power Electronics

Applications,” in Proceedings of 2016 International Symposium on 3D Power Electronics Integration and Manufacturing,
Raleigh, USA, 13.-15. June 2016 (IEEE, 2016).
13 N. Lazarus, C. Meyer, S. Bedair, H. Nochetto, and I. Kierzewski, “Multilayer liquid metal stretchable inductors,”

Smart Mater Struct. 23, 1–10 (2014).

14 S.-Y. Wu, C. Yang, W. Hsu, and L. Lin, “3D-printed microelectronics for integrated circuitry and passive wireless sensors,”

Microsystems & Nanoengineering 1 (2015).

15 M. K. Kazimierczuk, High Frequency Magnetic Components, 2nd Edition (Wiley, Chichester, 2014), pp. 472–519.
16 S. S. Mohan, M. M. Hershenson, S. P. Boyd, and T. H. Lee, “Simple accurate expressions for planar spiral inductances,”

IEEE J. Solid-State Circuits 34, 1419–1424 (1999).

17 J. Zhao, “A new calculation for designing multilayer planar spiral inductors,” Electronics Design, Strategy, News 3, 37–40

(2010).
18 V. E. Černyšev: “Rasčot častotnoj pogrešnosti měr magnitnoj indukcii”, Trudy metrologičeskich institutov SSSR 140(200),

Leningrad, Izd. standartov 1972 (in Russian).

19 M. Clemens and T. Weiland, “Discrete electromagnetism with the finite integration technique,” Progress In Electromagnetics

Research 32, 65–87 (2001).

20 S. Stalf, “Printed inductors in RF consumer applications,” IEEE Trans. Consum. Electron 47, 426–435 (2001).
21 M. Ulvr, “A system for calibration of search coils with area turns up to 100 m2 ,” Proceedings of Conference on Precision

Electromagnetic Measurements 2012, Washington DC, USA, 1.-6. July 2012 (IEEE, 2012), pp 320–321.
22 M. Ulvr, “Setup for generating an AC magnetic field from 3 kHz up to 100 kHz,” IEEE Trans. on Magnetics 51 (2015).

View publication stats