(Jin-Yong Bae)
†
Copyright © The Korean Institute of Illuminating and Electrical Engineers(KIIEE)
Key words
Current Sensor, Split Resistor. Skin Effect, Electrical Model
1. Introduction
In recent years, the use of power conversion has continued to increase. A typical
converter and inverter system uses current sensors to measure system current(1-15). Accurate current information detection is essential for power control such as various
power supplies, chargers, converters, and inverters that are controlled at high frequency
above several hundred [kHz](1-8). In addition, power electronics provides a more stable and responsive power supply
through current control rather than voltage control(1-5). If the amount of current is relatively small, hall sensors or magnetic sensors are
used. In the related art, the magnetic saturation of hall sensors increases, making
it is difficult to sense high DC current above several ten [A], with resistor current
sensors exhibiting higher accuracy(1,6). This article gives a current sensor of split resistor placement for high-power and
high-frequency application. This paper compares previous resistor current sensors
and the proposed current sensor. In the frequency range of 10[Hz] to 50[MHz], changes
in resistance, inductance and phase-difference are analyzed. The previous current
sensor is stable in the frequency range of 10[Hz] to 1[kHz]. However, the proposed
split resistor current sensor is stable in the frequency range of 10[Hz] to 1[MHz].
This paper provides an electrical model for the resistor current sensor. The validity
of the proposed split resistor sensor is analyzed through PSIM simulation and a prototype
experiment.
2. The Previous Current Sensor
Fig. 1 shows a previous current sensor for detecting high currents above several ten [A].
Fig. 1(a) shows a 30[A] previous current sensor, and Fig. 1(b) shows a 50[A] previous current sensor. In the 30[A] current sensor, the resistor
is designed to detect the current of 30[A] as 50[mΩ], and the 50[A] current sensor,
the resistor is designed to detect the current of 50[A] as 50[mΩ]. Therefore the resistance
value of each ideal current is given by the followingequations (1) and (2) according to Ohm's Law.
Fig. 1. Previous Current Sensor
3. The Proposed Current Sensor
Fig. 2 shows the current sensor for the proposed split resistor placement. The proposed
current sensor of the split resistor placement is composed of an octagon. Alternatively,
it is designed to flow up to 50[A] using 100 small resistors.
Fig. 2. Proposed Current Sensor for Split Resistor Placement
Fig. 3 shows the equivalent circuit model of the current sensor for the detection of high
current.
Fig. 3. Equivalent circuit of current sensor
where
${Z}$ : Impedance of current sensor
${R}$ : Pure Resistance of current sensor
${L}$ : Inductance of current sensor
${C}$ : Capacitance of current sensor
In the case of the capacitor component C = 0 in Equation (4), it can be expressed as Equation (5).
In the case of the capacitor component L = 0 in Equation (4), it can be expressed as Equation (6).
4. The Analysis of Skin Effect at High Frequency
For the precision of the resistor current sensor, skin effect must be considered for
frequency of several ten [kHz].
The skin effect is that the induced electromotive force is generated inside the conductor
because of the direction of the current changes rapidly at high frequency. That means
the high frequency current flows through the surface of the conductor. In general,
the skin depth can be expressed as Equation (7)(10-12).
where,
$μ$ : copper permeability $4\pi\times 10^{-7}$[H/m]
$σ$ : copper conductivity $5.8\pi\times 10^{7}$[S/m]
The length of the wavelength can be expressed as in Equation (8).
where
$\lambda$ : wavelength
$f$ : frequency
$c$ : velocity of light
Therefore, the length of the wavelength($\lambda$) and the skin depth($δ$) according
to the frequency($f$) is as follows.
Fig. 4 shows the skin depth of metals with frequency(9-10).
Table 1. Wavelength($\lambda$) and skin depth($δ$) according to the frequency
Frequency($f$)
|
Wavelength($\lambda$)
|
Skin depth($δ$)
|
1 [GHz]
|
0.3 [m]
|
2 [μm]
|
100 [MHz]
|
3 [m]
|
6.6 [μm]
|
1 [MHz]
|
300 [m]
|
66 [μm]
|
50 [Hz]
|
6000 [km]
|
9.3 [mm]
|
Fig. 4. Skin depth of metal according to the frequency
5. Simulation and Experiment Results
In this paper, Newtons4th Ltd's frequency response analyzer is compared to the current
sensors of the previous 30[A] and 50[A] current sensors with the proposed 50[A] split
resistor placement of high current detection. The PSM 3750, frequency response analyzer
was used to measure the change in resistance(R), inductance(L), and phase-differences
at frequencies of 10[Hz] to 50[MHz].
Fig. 5 shows the resistance(R) variation analysis of the proposed split resistance current
sensor and the previous current sensor.
Fig 5. Analysis of the resistance ration change of the proposed split resistance current sensor and the previous current sensor
In the case of the previous 30[A] and 50[A] current sensors, the resistance(R) is
generally stable when the frequency of the range is 10[Hz] to 500[Hz]. However, it
was confirmed that when performed, the resistance(R) exponentially increased from
1[kHz]. This is unusual, when the resistance decreases instantaneously at a frequency
of about 1[MHz], and rapidly rises in a higher frequency range. The proposed 50[A]
split resistor current sensor, however, has a characteristic that the frequency is
almost constant in the range of 10[Hz] to 1[MHz]. Therefore, it can be seen that the
proposed current sensor is an accurate advantage for current above hundred [kHz]s.
Also the proposed current sensor may include a characteristic that has the resistance
decreasing instantaneously at 5[MHz] and the resistance rapidly increasing from the
higher frequency range.
Table 2 shows the resistance variation analysis according to the frequency. Referring to
Fig. 6 and Table 1, with the resistance of 10[Hz] set to 1 which is the reference resistance, the change
of the resistance according to the frequency was examined. With resistance based on
10[kHz], the proposed 50[A] current sensor arrangement has a 1.7% resistance variation.
Also, when the resistance is based on 10[kHz], the previous 50[A] current sensor has
a 75.9% resistance variation, and in the case of the previous 30[A] current sensor
has a 95.4% resistance variation.
Table 2. Analysis of the resistance variation on frequency(Rx[Hz]/R10[Hz])
Fig. 6. Inductance of the proposed 50[A] current sensor and the previous 50[A] and 30[A] current sensor with frequency(LX[Hz])
Table 3. Analysis of the inductance variation on frequency(Lx[Hz])
When the resistance is based on 100[kHz], the proposed 50[A] current sensor arrangement
has a 6.1% resistance variation whereas the previous 50[A] current sensor has a 399.6%
resistance variation. The previous 30[A] current sensor has a 429.5% resistance variation.
With resistance based on 1 [MHz], the proposed 50[A] current sensor arrangement has
a 2.0% resistance variation. The previous 50[A] current sensor has a 435.1% resistance
variation while the previous 30[A] current sensor has a 762.5% resistance variation.
Alternatively, the previous 50[A] and 30[A] current sensor has large margin of error
at 100[kHz] and 1[MHz]. Therefore, the proposed 50[A] current sensor can have a higher
current accuracy in the range of 10[Hz] to 1 [MHz] compared to the previous 50[A]
and 30[A] current sensor.
Fig. 6 and Table 3 show the variation in inductance(L) of the proposed 50[A] current sensor and the
previous 50[A] and 30[A] current sensor.
The inductance(L) of the proposed 50[A] current sensor and the previous 50[A] and
30[A] current sensor had similar characteristics. It can be seen that it has the highest
inductance(L) of the frequency of 100[Hz], and thereby it is confirmed that it performs
variation in 200[nH] to 430[nH].
In this study, the simulation was performed using PSIM based on the measured resistance,
inductance, and phase-difference. Alternatively, the results were confirmed by experiments
in the range of 10[Hz] to 50[MHz].
Fig. 7. PSIM Simulation Circuit
Fig. 8. The voltage and current waveforms of the proposed current sensor and the previous current sensor (10[Hz])
Fig. 9. The voltage and current waveforms of the proposed current sensor and the previous current sensor (100[kHz])
Fig. 10. The voltage and current waveforms of the proposed current sensor and the previous current sensor (1[MHz])
Fig. 11. The voltage and current waveforms of the proposed current sensor and the previous current sensor (10[MHz])
In particular, the simulation was performed based on the model shown in Fig. 3.
Fig. 7 shows the PSIM simulation circuit and Fig. 8 to Fig. 11 display the voltage/current experiment waveforms of the proposed 50[A] current sensor
and the previous 50[A] and 30[A] current sensor. The proposed 50[A] current sensor
is verified by simulation and experiment and has a phase-difference between voltage-current
phase smaller the previous 50[A] and 30[A] current sensor.
(1) 100[kHz]
▷ The proposed 50[A] current sensor :
6.612[deg] voltage-current phase-difference
▷ The previous 50[A] current sensor :
21.057[deg] voltage-current phase-difference
▷ The previous 30[A] current sensor :
36.047[deg] voltage-current phase-difference
(2) 1[MHz]
▷ The proposed 50[A] current sensor :
85.751[deg] voltage-current phase-difference
▷ The previous current 50[A] sensor :
89.725[deg] voltage-current phase-difference
▷ The previous current 30[A] sensor :
89.513[deg] voltage-current phase-difference
(3) 10[MHz]
▷ The proposed 50[A] current sensor :
97.243[deg] voltage-current phase-difference
▷ The previous 50[A] current sensor :
99.062[deg] voltage-current phase-difference
▷ The previous 30[A] current sensor :
103.975[deg] voltage-current phase-difference
From Fig. 8 to Fig. 11, simulation and experiments show that the proposed 50[A] current sensor has a smaller
voltage-current phase difference than that of the previous 50[A] and 30[A] current
sensor.
6. Conclusion
This paper presents a split resistor current sensor for stable resistance up to 1[MHz].
Where previous current sensors consist of a single resistor, the proposed split resistor
current sensor is divided into 100 detailed resistors. The current sensor can be flexibly
selected according to the current level. In the proposed current sensor of the split
resistor arrangement, there is stable resistance(R), and the voltage-current phase-
difference shows a relatively small technical characteristic compared with the previous
current sensor. Therefore, experimental results show that the proposed split resistor
current sensor is reliable over a wide range of frequencies and achievement is more
exact than previous current sensors.
References
Petter J., McCarthy J., Pollak P., Smith C.C., Nov 1991, Survey of DC Current Measurement
Techniques for High Current Precision Power Supplies, IEEE Nuclear Science Symposium
and Medical Imaging Conference 1991, Vol. 2, pp. 961-967
Mammano Bob, 1999, Current Sensing Solutions for Power Supply Designers, Unitrode
Seminar Notes SEM 1200
Xiao Chucheng, Zhao Linyin, Asada T., Odendaal W.G., Van Wyk J.D., Oct 2003, An overview
of Integrable Current Sensor Technologies, IEEE Industry Applications Conference,
2003, 38th IAS Annual Meeting, Conference, Vol. 2, pp. 1251-1258
Masi M.G., Peretto L., Tinarelli R., Sep 2012, Metrological Characterization of a
Current Sensor for Smart Grids, IEEE Energy Conference and Exhibition(ENERGYCON 2012),
pp. 806-811
Radun A., Feb 1995, An alternative Low-cost Current Sensing Scheme for High-current
Power Electronics Circuits, IEEE Transactions on Industrial Electronics, Vol. 42,
No. 1, pp. 78-84
Pankau J., Leggate D., Schlegel D., Kerkman R., Skibiniski G., Nov/Dec 1999, High
Frequency Modeling of Current Sensor, IEEE Transactions on Industry Applications,
Vol. 35, No. 6, pp. 1374-1382
Patel A., Ferdowsi M., Oct 2009, Current Sensing for Automotive Electronics – A Survey,
IEEE Transactions on, Vol. 58, No. 8, pp. 4108-4119
Ripka P., 2010, Electric Current Sensors: A Review, Measurement, Science and Technology,
Vol. 21, No. 11, pp. 1-23
Armstrong Keith, Skin Effect and Surface Currents, In CONPLIANCE Internet Site (https://incompliancemag.com
/article/skin-effect-and-surface-currents/)
Bae Jin-Yong, Nov 2019, Experimental Analysis of Split Resistor Placement, Proceeding
of The ISAAC 2019
Montgomery Carol G., 1947, Techniques of Microwave Measurements, McGraw-Hill Book
Gu Wen-Jine, Liu Riu, 1993, A study of volume and weight vs. frequency for high-frequency
transformers, PESC '93 Record, 24th annual ieee, pp. 1123-1129
Pearson Electronics, USA , EMI Current Probe Selection Guide, Internet Site (www.pearsonelectronics.com)
Filipski P.S., Boecker M., 2006, AC-DC current shunts and system for extended current
and frequency ranges, IEEE Transactions on Instrumentation and Measurement, Vol. 55,
No. 4, pp. 1222-1227
Mariscotti A., May 2010, A Calibration setup for AC and DC Current Sensors, IEEE Instrument
and Measurement Technology Conference, 2010, pp. 1409-1413
Biography
Jin-Yong Bae received a B.S., M.S., and Ph.D. from the Electrical Engineering Department
of Dongguk University, Seoul, Korea in 1998, 2002, and 2005, respectively.
He worked as a patent examiner of the electric power and semiconductor technology
examination division KIPO (Korean Intellectual Property Office).
He is currently a professor in the Department of Electric Vehicle Control Engineering
at Dongshin University, Naju, Korea.
His interests include power converter analysis and optimum design for EV(Electric
Vehicle), soft-switching converters and inverters, application for LED lighting and
solar power systems, etc.