Current sensing is used in many applications as a way to monitor system performance and protect system components. It’s essential.
In applications involving motors, the technique is also used to determine the speed and direction of the motor. Generally, motor applications will use a method called indirect current sensing. This method involves using magnetic induction in coils such that the monitoring current branch is isolated from the load. Any applications involving lower power (up to tens of Amps) use direct current sensing. This method involves using a special precision sensing resistor to generate a voltage that is then measured (Ohm’s law).
The main applications of direct current sensing involve overcurrent detection, current control, and current management. For example, overcurrent detection is used to determine if there is a short in the circuitry. Current control can be used in feedback systems such as power supplies and switching converters to regulate the output current. On the other hand, current management applications will generally use current sensing as a way to monitor and control battery operated systems for maximum battery life.
Primary Operating Principles
The fundamental idea behind direct current sensing is based on creating a small voltage by Ohm’s law. The current to be monitored is applied to a resistor and generates a voltage that is then measured.
The problem of this method is that an undesirable voltage drop is created. The larger the voltage drop, the more power the resistor dissipates. If the resistor’s power rating is exceeded, the device can fail.
As a result, it is necessary to keep the voltage drop across the resistor to a minimum value of tens to a few hundreds of millivolts. For this reason, many of the current sense resistors used in applications are anywhere from 10 milliohms to a few ohms in value. These generally come with a very tight tolerance in order to generate reliable measurements.
In addition to a resistor’s tolerance and resistance values. there are a few additional parameters to consider when selecting a sense resistor.
One is the Temperature Coefficient of Resistance (TCR). The TCR describes how the resistance may change with temperature. The larger the TCR, the more resistance will change as temperature drifts.
An important consideration is the TCR of the copper pads that the resistor will solder to, which can be well above the TCR of sensing resistors and introduce larger inaccuracies in the measurement.
To mitigate the effect from the copper pad, designers will generally use a layout method known as a Kelvin connection. This method connects the traces to the inner side of the pads and effectively avoids the effects due to contact resistance. In fact, many companies offer 4-terminal sense resistors that take advantage of the Kelvin principle. The Kelvin Principle can be seen in the image below:
To measure the generated voltage, a shunt resistor is generally placed in series with the load. A differential amplifier is then used to monitor the generated voltage and apply it to a measuring device such as an analog to digital converter (ADC) or feed the output back to supporting circuitry.
Two common methods exist for achieving this measurement, and they are known as high side current sensing and low side current sensing. High side current sensing places the current sense resistor between the power supply and the load. Low side current sensing places the current sense resistor between the load and ground. Both methods offer advantages and disadvantages leaving the designer to determine what is best for the specific application.
The method of low side current sensing eases the requirements needed from the amplifier, making the selection and implementation easier. Since the sensing resistor is referenced directly to ground, the generated voltage that needs to be measured is very small. Additionally, since the measured voltage does not have a large common-mode voltage present, common-mode rejection is not of concern.
For these reasons, low side current sensing is generally easier and cheaper to implement. However, the drawbacks of low side current sensing include the inability to detect a short in the load, which can be a very important fault detection notification.
Since the resistor is in between the load and ground, the load circuitry is no longer referenced directly to the ground. This may or may not be of a concern depending on how sensitive the circuitry is.
The method of high side current sensing has the advantage of directly measuring the current being drawn from the source. Not only can it detect when a short happens, but also it allows the load to be referenced directly to ground--which is generally highly desirable.
The disadvantage is that the differential amplifier required will have to operate at potentially high common-mode voltages and also require good common-mode rejection.
For these reasons, many devices now exist today, known as current shunt monitors (CSMs) or current-sense amplifiers (CSAs). These devices are generally designed to support common-mode voltages beyond their operating voltages.
In addition, they feature low DC offsets as well as fixed gains. Many include integrated resistors to improve matching, which reduces error and improves common-mode rejection. Temperature stability is generally implemented as well to allow for precise measurements over a broad range of operating temperatures.
The type of output must also be considered when choosing the amplifying device. The output may come in the form of current, voltage or even a digital output. Furthermore, the gain needed for accurate readings is available in different methods. The gain is generally either fixed and pre-determined or adjustable through external components or programmable gain settings.
When choosing the gain, the maximum voltage drop of the current sense resistor must be considered. For instance, if the largest voltage drop expected across the sense resistor is 100mV, then the designer may want a gain of 50V/V to interface with a 5V ADC input. The tradeoff between fixed versus non-fixed gain devices is that fixed devices allow better matching for the internal components and thus a more accurate gain.
On the other hand, the external gain setting devices allow for greater flexibility. Programmable gain devices take advantage of both of the features mentioned above. However, they now require additional traces to interface and program the device and can potentially cost the designer more.