designfeature
Bryan Yarborough
Applications Engineer, TT Electronics
Components and Methods for
Current Measurement
Current measurement com-
ponents and methods must
provide an accurate output
signal as well as preventing
damage to the associated
printed circuit board.
l RDCR
1:N
RBurden
Ideal current transformer circuit
Fig. 1. In the ideal current transformer ac current passes through the copper wind- 3. Transistor (Direct)
ings with very little resistive losses.
TTelec_F1
Rogowski Coil
Fig. 2. Rogowski coil is an air core design
that has a lower inductance providing a faster
response.
TTelec_F2
28 Power Electronics Technology | January 2012
C
urrent sensing is used to perform two essential circuit functions.
First, it is used to measure “how much” current is flowing in a
circuit, which may be used to make decisions about turning off
peripheral loads to conserve power or to return operation to
normal limits. A second function is to determine when it is a
“too much” or a fault condition. If current exceeds safe limits, a
software or hardware interlock condition is met and provides a
signal to turn off the application, perhaps a motor in a stalled condition or short cir-
cuit. It is essential to choose the appropriate technology with the necessary robust-
ness to properly withstand the extreme conditions that can exist during a fault.
A signal to indicate the “how much” condition and the “too much” condition is
available in a variety of measurement methods:
1. Resistive (Direct)
a. Current Sense Resistors
b. Inductor DC resistance
I/N V= (I/N x RBurden)
2. Magnetic (Indirect)
a. Current Transformer
b. Rogowski Coil
c. Hall Effect Device
a. RDS(ON)
b. Ratio-metric
Each method has advantages for current measurement, but also comes with
tradeoffs that can be critical to the end reliability of the application. They can also be
classified into two main categories of measurement methods; direct or indirect. The
direct method means that it is connected directly in the circuit being measured and
that the measurement components are exposed to the line voltage, whereas the indi-
rect method provides isolation that may be necessary for design safety.
Current Sense Resistor
The resistor is a direct method of current measurement that has the benefit of sim-
plicity and linearity. The current sense resistor is placed in line with the current
being measured and the resultant current flow causes a small amount of power to
be converted into heat. This power conversion is what provides the voltage signal.
Other than the favorable characteristics of simplicity and linearity, the current
sense resistor is a cost-effective solution with stable Temperature Coefficient of
Resistance (TCR) of < 100 ppm/°C or 0.01% /°C and does not suffer the potential
of avalanche multiplication or thermal runaway. Additionally, the existence of
low resistance (< 1 mΩ is available) metal alloy current sense products offer supe-
rior surge performance for reliable protection during short circuit and overcurrent
events.
www.powerelectronics.com
 Load
ID Metal
Drain D
ID
N+ SiO2
l G + VS
VH = V Gate VDS N P
VGS +
VGS
N+
Source S

B
Fig. 4. Power MOSFET’s on-resistance provides current sensing capability

Hall effect principle, magnetic field present
Fig. 3. Hall-Effect devices are capable of measuring large currents.
TTelec_F3
Inductor DC resistance
The DC resistance of an inductor can also be used to
provide a resistive current measurement. This method is
considered “lossless” because of the low resistance value of
the copper, typically < 1 mΩ and because it is providing a
secondary use of an existing component. In higher current
applications; a 30 amp current would provide a 30 mV
signal for a 1 mΩ resistance value. This method has two
drawbacks; first copper has a high TCR (temperature coef-
ficient of resistivity) of approximately 3900 ppm, which
causes the resistance value to increase by 39% for a 100°C
rise above room temperature. Because of this high TCR,
the temperature must be monitored and compensated to
provide an acceptable current measurement. The second
drawback is the variance in the resistance of the copper
due to dimensional changes that occur due to the conduc-
tor being wider or thinner from one lot to the next.
Current Transformer
A current transformer’s three key advantages are that it
provides isolation from the line voltage, provides lossless
current measurement, and the signal voltage can be large
providing a measure of noise immunity. This indirect
current measurement method requires a changing cur-
rent, such as an AC, transient current, or switched DC;
to provide a changing magnetic field that is magnetically
coupled into the secondary windings (Fig. 1). The second-
ary measurement voltage can be scaled according to the
turns ratio between the primary and secondary windings.
This measurement method is considered “lossless” because
the circuit current passes through the copper windings
with very little resistive losses. However, a small amount
of power is lost due to transformer losses from the bur-
den resistor, core losses, and primary and secondary DC
a voltage is induced into a secondary coil that is proportional
to the current flow through an isolated conductor. The
TTelec_F4
exception is that the Rogowski coil, (Fig. 2), is an air core
design as opposed to the current transformer that relies upon
a high permeability core, such as a laminated steel, to mag-
netically couple to a secondary winding. The air core design
has a lower inductance providing a faster signal response and
very linear signal voltage. Because of its design, it is often
used as a temporary current measurement method on exist-
ing wiring such as a handheld meter. This could be consid-
ered a lower cost alternate to the current transformer.
Hall Effect
When a current carrying conductor is placed in a magnetic
field, as shown in Fig. 3, a difference in potential occurs
perpendicular to the magnetic field and the direction of cur-
rent flow. This potential is proportional to the magnitude
of the current flow. When there is no magnetic field and
current flow exists, then there is no difference in potential.
However, when a magnetic field and current flow exists the
charges interact with the magnetic field, causing the current
distribution to change, which creates the Hall voltage.
The advantage of Hall effect devices is that they are capa-
ble of measuring large currents with low power dissipation.
However, there are numerous drawbacks that can limit their
VBAT
Load
R2
Gate drive ISENSE
KELVIN/SOURCE
R1
– OUT
VSENSE RSENSE
+
VOUT
Power ground

resistance. Isf = senseFET current

Rogowski Coil Fig. 5. SenseFET uses a small portion of its parallel MOSFET cells to sense cur-
The Rogowski coil is similar to a current transformer in that rent.

TTelec_F5
www.powerelectronics.com January 2012 | Power Electronics Technology 29
 currentmeasurement

Fig. 6. (a) Shows a four-terminal lel cells and connects to the common Gate and Drain, but
resistor for current sense applica- separate Source. This creates a second isolated transistor; a
tions. (b) Shows two pad designs, “Sense” transistor. When the transistor is turned on, the cur-
(1) with isolated pad regions, and rent through the sense transistor will be a ratio comparable
(2) with a plated through-hole. to the main current through the other cells.
Depending on the transistor
product, the accuracy tolerance
range can vary from as low as 5%
or as wide as 15% - 20%. This
is not suitable for current control
applications that typically require
1% measurement accuracy, but is
intended for overcurrent and short
circuit protection.
Pad Design 1 Pad Design 2
Resistor technology benefits
Thin Film, are not typically used for
use, including non-linear temperature drift requiring com- current sense applications, but is included in this discussion
pensation, limited bandwidth, low range current detection to provide breadth to the topic. Generally these resistive
requiresTTelec_F6b.ai
a large offset voltage that can lead to error, suscep- products are for precision applications because of the resis-
tibility to external magnetic fields, and high cost. tive layer ranges from 0.000001 in. to 0.000004 in. thick.
They are quite surge-tolerant in the appropriate application,
Transistors but are not designed for the high currents typically associ-
Transistors are considered a “lossless” overcurrent detec- ated with the applications mentioned here.
tion method since they are standard control components
to the circuit design and no further resistance or power
dissipating devices are required to provide a control signal.
Transistor datasheets provide the on-resistance for the
drain-to-source, RDS(ON), with a typical resistance in the
mΩ range for power MOSFETs (Fig. 4). This resistance
comprises several components that begin with the leads
connecting to the semiconductor die through the resis-
tance that makes up the numerous channel characteristics.
Based on this information, the current passing through
the MOSFET can be determined by ILoad = VRDS(ON) /
RDS(ON).
Each constituent of the RDS(ON) contributes to measure-
ment error that is due to minor variations in the resistances
of the interface regions and TCR effects. The TCR effects
can be partially compensated by measuring temperature (a)
and correcting the measured voltage with anticipated
change in resistance due to temperature. Often times,
the TCR for MOSFETs can be as large as 4000 ppm /
°C, which is equivalent to a 40% change in resistance for
100°C rise. Generally, this method provides a signal with
approximately 10% to 20% accuracy. Depending on the
accuracy requirements, this may be an acceptable range
for providing overcurrent protection.
(b) OAR5 OAR5 OAR5
(2.94 Watts) (5.15 Watts) (5.15 Watts)
Ratio Metric Current Sense MOSFETs Side View Side View Side View
The MOSFET consists of thousands of parallel transis-
tor cells that reduce the on-resistance. The current sensing Fig. 7. Thermal images illustrating the isolation performance of the OAR and
MOSFET shown in Fig. 5 uses a small portion of the paral- OARS product.
TTelec_F7

30 Power Electronics Technology | January 2012 www.powerelectronics.com

Fig. 8. Differences in the thermal expansion coefficient between the metal
materials of the current sense resistor and the circuit board materials dissipate
forces on each part.
TTelec_F8.ai
Thick Film, is typically 0.0005 to 0.002 in. thick, which
is nearly 100 times that of Thin Film. The increased thick-
ness equates to a greater mass that is better able to carry
the relatively high currents and dissipate the heat across
the substrate, as well as better able to manage transients.
Another advantage of the Thick Film products is the flex-
ibility to request off standard resistance values because of
the process efficiency of laser trimming. The tradeoff is that
these products are not as capable of the very tight tolerances
of Thin Film products.
Foil, has a larger cross section still and is a uniform resis-
tive alloy, which is different from the Thick Film technology
that is resistive materials suspended in a glass matrix. By
comparison the foils tend to withstand larger surge tran-
sients as compared to the previous versions. The principle
advantage with this technology is the low range ohmic val-
ues with low TCRs.
Bulk alloy, this resistor technology has the greatest surge
tolerance because of its large current carrying mass. It
is available in resistance values as low as 0.000 5 Ω with
low TCR. This tends to be the best choice for high cur-
rent power supplies or where fault conditions can result in
extreme currents. These products do not have as wide of
a resistance offering as the Thick Film products, because
the resistor alloy has limited resistivities to reach high range
values, as well as needing the mechanical strength to tolerate
process handling.

Product Specific Features

High--current applications require the resistance value to
be very low to minimize power losses and yet provide the
necessary signal level to provide a voltage signal high enough
to exceed noise levels. These low ohmic values often times
need a four-terminal connection to reduce errors that may
result due to the contact resistance that occurs when the
part is mounted to the board.
The CSL (Fig. 6a) offers four terminals by design, but
other standard surface mount devices can benefit from
a four-terminal pad design. These parts offer physically
separated connection points for current and voltage, which
reduces contact related measurement error. In the case of
the CSL the current will flow through the inner pins and

www.powerelectronics.com January 2012 | Power Electronics Technology 31

 currentmeasurement

voltage is sensed on the Measurement Accuracy Isolation EMI (Tamper Robust Size Cost
outer pins and is rec- Method Resistance)
ommended for best Resistive (Direct)
accuracy with the Sense Resistor High No High High Small Low
LRF3W to be config- Inductor DC resistance Low No Moderate High Small Low
ured as a cross flow
Transistor (Direct)
arrangement with cur-
RDSon Low No Moderate Moderate Small Low
rent on opposing dia-
Ratio Metric Moderate No Moderate Moderate Small Moderate
metrically corners (e.g.,
pin 2 to pin 3). Magnetic (Indirect)
The pad layout (Fig. Current Transformer High Yes Moderate High Large Moderate
6b) creates separate Rogowski Coil High Yes Moderate High Large Moderate
regions for measuring Hall Effect High Yes High Moderate Moderate High
signal voltage from the
current carrying por- formance. Observe the temperature of the
tion, which reduces error. Pad Design 1 solder joint with respect to the hot spot.
illustrates one method that creates an iso- These temperatures are based on reaching
lated pad region within the pad layout, but thermal equilibrium, however in the appli-
this design may reduce the pad area below cation these results may be extended to
necessary limits to carry high currents be considered as the thermal performance
through the copper trace. Pad Design characteristics for an overcurrent protec-
2 uses a plated through hole to connect tion condition. The FR4 will not exceed its
under the pad and connects to an internal temperature rating, though extreme circuit
or outer trace for measurement; this maxi- conditions exist.
mizes the pad space to carry current to
the resistor. The contact point places the Solder Joint Stress
signal line as close to the current channel as The OARS product family’s elevated and
possible; minimizing measurement error. curved construction permits the resistor
Fig. 9. LRF3W from TT electronics uses side to flex, reducing the stress generated by
Thermal Isolation terminations to achieve a 3W rating. differences in thermal expansion coeffi-
The OARS (Open Air Resistor Surface cients between the heat producing metal
Mount) is a unique design that elevates the hot spot of and the dissipating circuit board material. Surface mount
the resistive material well above the circuit board material. components that are flat and parallel to the circuit board
This places the hottest region of the part into the available will apply shear forces to the solder joint that can lead to
airstream, which dissipates the maximum amount of heat failure or changes in performance. In high thermal cycling
energy to the air instead of the PCB. applications, the OARS has been preferred to other simi-
This provides a two key advantages for thermal design, lar all metal construction parts because of this flexibility
which affects the PCB material and the other neighboring feature (Fig. 8).
power or semiconductor components. Typical FR4 PCB The LRF3W (Fig. 9) from TT electronics provides several
material is only rated to 130°C; a power resistor that is design benefits derived from its 1225 aspect ratio with the
traditionally against the board could cause damage to the termination along the long side of the component. The side
material during power excursions or reduces the upper termination extends the power rating to 3 Watts, eliminates
temperature performance limits of the circuit. An elevated the need to reduce the circuit traces as required by the tra-
current sense prevents damage to the circuit material and ditional 2512 footprint, and reduces solder joint stresses due
permits the solder joint to run cooler. The second benefit to differences in the temperature coefficient of expansion
by dissipating the heat to the air instead of the PCB is the between the ceramic and PCB material. The 1225 aspect
improved performance of nearby heat-affected devices. reduces the distance between the center / hot spot region of
These effects may include lifetime rating, power handling, the part and the thermally dissipative circuit board material.
luminous output, accuracy, and reliability. This permits a high 3W rating and reduces the stresses on
The thermal images shown in Fig. 7 help illustrate the the solder joint due to the differences in temperature coef-
isolation performance of the OAR and OARS product. ficients of expansion between the ceramic substrate and the
These tests were conducted on FR4 board material with no thermal mass of the board material.
ambient airflow; airflow would improve system thermal per-

32 Power Electronics Technology | January 2012 www.powerelectronics.com