Smart test

Analysis of current detection scheme

Compared with ten years ago, electronic products now have more functions. Engineers have to design sophisticated systems, often "creatively" meeting strict power budgets in order to maintain high energy efficiency. Predictive system maintenance and protection require rapid response system response. A key function is to monitor the current consumption and voltage drop of the system.

In all current detection methods, the use of amplifiers to monitor the shunt current is the most commonly used method so far. Current detection can be implemented using a current sense amplifier (CSA) or an operational amplifier (Op Amp) with an external gain setting resistor (Figure 1). The choice of the two depends on the performance requirements and the target cost of the bill of materials (BOM).

Typical op amp current sensing requires 2 to 4 precision resistors

Ordinary operational amplifier current detection requires 2 to 4 precision resistors

NCS21xR has integrated precision resistors: NCS21xR has integrated precision resistors

From a performance point of view, the mismatch between the gain setting resistors will affect the accuracy of the current measurement, which in turn affects the size of the parallel device. Other design considerations include device specifications (input bias voltage, common mode rejection, gain error), parallel device size, shunt location, and PCB layout. We will analyze these factors in depth in subsequent blogs. Now, let us look at these factors in general.

We selected four amplifiers (LM358, NCS20081, NCS333 and NCS214R) to compare the performance advantages of general-purpose to precision amplifiers (from left to right in Figure 2 and Figure 3).

NCV214R integrates gain setting resistors for better matching and common mode performance. Others require an external resistor network. Suppose a matched resistor network is used in the performance comparison of LM 358, NCS20081 and NCS333. NCS214R provides precision (Figure 2) and extremely energy-efficient solutions (Figure 3).

With a 50 mV shunt drop: 14% offset error

50 mV shunt voltage drop: 14% bias error

With a 50 mV shunt drop: 7% offset error

50 mV shunt voltage drop: 7% bias error

With a 50 mV shunt drop: 0.02% offset error

50 mV shunt voltage drop: 0.02% bias error

With a 50 mV shunt drop: 0.12% offset error

50 mV shunt voltage drop: 0.12% bias error

Reduce offset voltage: reduce the offset voltage

Improve accuracy: Improve accuracy

System performance improves: System performance improves

Figure 2: For a fixed shunt pressure drop (for example, 50 mv), the offset error is several orders of magnitude difference

To achieve 2% offset error: 350 mV shunt drop

To achieve 2% bias error: 350 mV shunt voltage drop

To achieve 2% offset error: 175 mV shunt drop

To achieve 2% bias error: shunt voltage drop 175 mV

To achieve 2% offset error: 0.5 mV shunt drop

To achieve 2% bias error: shunt voltage drop 0.5 mV

To achieve 2% offset error: 3 mV shunt drop

To achieve 2% bias error: shunt voltage drop 3 mV

Reduce voltage drop across sense resistor: reduce the sense resistor voltage drop

Reduce power dissipation: reduce power dissipation

System efficiency improves: System energy efficiency improves

Figure 3: To achieve the system energy efficiency, for a fixed bias error, a lower shunt voltage drop will reduce power consumption

From the perspective of BOM cost, a well-matched resistor network will be expensive (~$1) to offset the cost savings of using a general-purpose op amp (~$0.10). Although current-sense amplifiers are expensive, they are likely to be cheaper than op-amp solutions when the cost of the complete solution is compared.

But wait! Not only that... Another advantage: the size of the solution. An op amp with an external resistor network will not be as small as the NCS21xR in uQFN or SC70.