Amplifier Senses Currents At Large Negative Voltages

Aug 1, 2007 12:00 PM
By Akshay Bhat, Strategic Applications Engineer, and Prashanth Holenarsipur, Product Definer, Maxim

In telecom and other systems with high-voltage negative supply rails, a simple combination of an instrumentation amplifier and discrete components enables high-side current sensing for circuit protection.

Click here for the enhanced PDF version of this article including diagrams and/or equations.

High-side current-sense amplifiers (CSAs) are principally used for monitoring the current from a positive supply rail. However, applications like ISDN and telecom power supplies require CSAs that operate on a negative rail. One method for designing a negative-rail CSA uses a precision instrumentation amplifier IC and several discrete components.

A similar approach, using a dual-supply op amp to sense current through a -5-V rail, was discussed in a previous article.^[1] But here, the design is extended for sensing much higher negative-rail voltages, while using an amplifier IC that operates on a single supply. Though the example presented here is for a -120-V supply, the design can be modified for monitoring negative rails at other voltage levels.

Application Example

Fig. 1 shows a block diagram of the power-distribution network in a typical telephone exchange. A rectifier converts the ac at the power mains to dc, and the dc output from the rectifier is used to charge a 48-V lead-acid battery. The battery powers the user telephones through the telephone line, thus eliminating the need for a back-up battery at the user end.

The battery polarities are connected so that the line voltage is negative (-48 V). A negative line voltage reduces the corrosion from electrochemical reactions occurring on a wet telephone line. A telecom network also uses several dc-dc converters to derive intermediate power-supply rails from the -48-Vdc input. These intermediate rails power the switches, radios, routers, ATX computers and other electronic equipment in the telephone exchange.

A fault condition can damage the power supply if the load current exceeds the maximum rated value; therefore, output protection is required. A time-tested method is to build a circuit breaker using a CSA and a power transistor. The CSA intensifies the small voltage drop across a sense resistor that is added in series with the battery. The circuit breaker is triggered whenever the battery current rises to the maximum rated value, typically 120% to 140% of the nominal current.

The sense resistor can be placed either between the load and the ground (low-side current sensing) or between the load and the negative terminal of the 48-V battery (high-side current sensing). These two alternatives require a tradeoff in different areas. The low-side resistor adds undesirable resistance in the ground path. Moreover, not all faults can be detected using the low-side scheme.^[2]

The high-side or negative supply current sensing has to handle a larger power supply and common-mode signal, but it can detect any fault caused by inadvertent shorts with the ubiquitous ground plane. The CSA discussed in this article follows the high-side approach.

Circuit Description

The circuit in Fig. 2 shows an implementation of the negative-rail current-sensing block. It uses an instrumentation amplifier like the MAX4460 or MAX4208 and some discrete components.

The MAX4460/MAX4208 are instrumentation amplifiers with a novel architecture called indirect current feedback.^[3] This topology enables the input common-mode voltage range to cover ground (i.e., the amplifier's negative rail) unlike traditional three op-amp architecture.^[4]

In the indirect current-feedback architecture, load current through R_SENSE creates a differential voltage between IN+ and IN- of the instrumentation amplifier. This differential voltage is converted to an internal differential current by transconductance amplifier g_M1. The transconductance amplifier g_M2, again with identical transconductance, seeks to cancel this internal differential current by using a high-gain amplifier in a negative feedback loop. Due to the matched transconductance, the feedback action reproduces the input-differential voltage between IN+ and IN- across the pins FB and GND.

The MAX4460's output provides a suitable gate drive for MOSFET M₁. The voltage drop across resistor R₃ equals V_SENSE, the voltage across R_SENSE. Consequently, R₃ sets a current proportional to the load current:

I_OUT = (I_LOAD × R_SENSE)/R₃ = V_SENSE/R₃. (Eq. 1)

The drain-source breakdown-voltage rating of the MOSFET must exceed the total voltage drop between the two supply rails (+125 V, in this case). R₂ is chosen so that the output voltage lies within the desired range of the following circuit, typically an analog-digital converter (ADC). As explained later, R₂ and R₃ set the gain of the CSA. An additional op-amp buffer can be used at V_OUT, if the ADC does not have a high-impedance input.

Zener diode D₁ protects the instrumentation amplifier from overvoltage damage, while providing sufficient supply voltage for its operation. If the sense current increases above the rated value during a fault condition, then the output voltage goes negative. Diode D₂ protects the ADC from damage by limiting the negative voltage at output to one diode drop.

Design Steps

The above design can be adapted to add high-voltage, negative supply current-sense monitoring capability. This flexibility is illustrated by choosing -120 V as the negative rail. By using the following straightforward steps, one can design a CSA for a different supply rail.

Step 1: Specify the zener regulator

It is important to bias the zener on a point in its transfer characteristic that gives a low dynamic resistance (i.e., well into its reverse breakdown region) to prevent PSRR errors. Fig. 3 shows a plot of the zener current versus the zener voltage for a standard zener diode configured in reverse bias.

Data shows that the zener voltage is not well regulated close to the breakdown voltage. A general rule then is to select the bias point to be about 25% of the maximum current specified by the power rating. This bias point gives a low dynamic resistance without wasting too much power. The bias point is set to the desired value by choosing the resistor, R₁, based on the following equation:

I_R1 = (V_CC + |V_NEG| — V_Z)/R₁ = I_S + I_Z, (Eq. 2)

where VCC is the positive-rail supply voltage, V_Z is the regulated zener voltage, |V_NEG| is the absolute value of the negative-rail voltage, I_S is the supply current for the MAX4460 and I_Z is the current through the zener diode.

R₁ must have a suitable power rating and be able to withstand the large voltages across it. Alternatively, one can use a series-parallel combination of lower-wattage resistors to ease these constraints.

Step 2. Select the transistor

The n-channel MOSFET, or JFET, must have a drain-to-source breakdown voltage rating greater than |V_NEG| + V_CC. This is an important constraint if the negative supply voltage is high.

Step 3. Choose R_SENSE

Select R_SENSE so that the full-scale sense voltage across R_SENSE is less than or equal to 100 mV. R_SENSE must be able to dissipate the I²R losses. If the resistor's rated power dissipation is exceeded, its value may drift or it may fail altogether.

Step 4a. Select R₃

There is considerable flexibility in choosing R₃. A good selection is influenced by two observations: As R₃ is reduced, Eq. 1 implies that for a fixed gain, the dissipated power increases. In addition, the thermal noise and leakage current of the FET set the upper limit on the selected value of R₃.

Step 4b. Select R₂

The ratio of resistors R₂ and R₃ equals the voltage gain of the resulting CSA. The output voltage is given as:

V_OUT = V_CC — I_OUT × R₂. (Eq. 3)

From Eqs. 1 and 3, we get: V_OUT = V_CC — (V_SENSE × R₂ / R₃).

Differentiating with respect to V_SENSE:

Voltage gain, A_V = -R₂ / R₃. (Eq. 4)

The negative sign represents the inverting relationship between the output voltage and the input sense voltage. From Eq. 4, R₂ can be determined.

Results

Fig. 4 plots the resulting typical output voltage as a function of the sense voltage. The following typical parameters can be inferred for the CSA: Input referred offset voltage = (5 V - 4.9831 V) / 49.942 = 338 µV, where gain = -49.942. The accuracy of the resulting CSA is a function of the tolerance of the resistors. All the resistors (except for R₁) must have a tolerance of 1% or better.

References

1. Yang, Ken, “Precision Circuit Monitors Negative Supply Current,” Power Electronics Technology, Sept. 1, 2005, p 78.

2. Maxim Integrated Products, Application Note 746, “High-Side Current-Sense Measurement: Circuits and Principles,” March 26, 2001.

3. Huijsing, Johan Hendrik, and Shahi, Behzad, “GM-Controlled Current-Isolated Indirect-Feedback Instrumentation Amplifier,” U.S. Patent 6,559,720, Oct. 26, 2001.

4. Maxim Integrated Products, Application Note 4034, “Three is a Crowd for Instrumentation Amplifiers,” April 12, 2007.