Power Tips #153: How to generate a regulated negative output from a negative input using a boost controller

How to deal with the reality that a standard buck controller power stage won’t work for negative-input conversion.

Telecommunications equipment, industrial test and other applications require a negative input to negative output voltage conversion. Because dedicated controllers for this topology are rare, you need a workaround to generate a stable output.

One solution I’ve found is to connect a boost controller’s GND pin to the negative input rail, which repurposes the device as a negative-input, negative-output buck controller and eliminates additional gate-drive circuitry. It then becomes possible to have a level-shifted feedback network regulate the output. So in this power tip, I’ll discuss two approaches using a traditional switch-mode power-supply: one using a buck controller with external field-effect transistors (FETs) and one using a buck converter with integrated FETs.

Comparing standard and negative-input, negative-output buck controllers

The power stage of a standard buck controller (Figure 1) closely resembles a negative-input, negative-output topology.

Figure 1 This simplified, standard buck controller schematic resembles that of a negative-input, negative-output topology.

A buck controller operates by applying a pulse-width modulation waveform to an inductor-capacitor filter. Each switching cycle starts when the main switch turns on, increasing the inductor current. Current flows from the input capacitor through the inductance to the output capacitor and back to the input capacitor. During the off time, the current commutates to the low-side diode (or switch) and the inductor current decreases.

Why a standard buck controller power stage won’t work for negative-input conversion

A negative-input, negative-output buck controller behaves very similarly to a standard buck controller. The main difference is that all currents flow in the opposite direction.

You cannot use a standard buck controller power stage, though, because of the orientation of the diode and metal-oxide semiconductor field-effect transistor (MOSFET) (with its internal body diode). Rotate these components as shown in Figure 2.

Figure 2 This simplified schematic details a negative-input negative-output buck controller.

Note that the output voltage cannot become more negative than the input voltage.

As an example, with a –48V input and a 50% duty cycle, the controller generates a –24V output. The controller’s control law decreases the negative output voltage toward the level of the negative input by increasing the “on” time of the main FET. So at a theoretical 100% duty cycle, the output voltage nearly equals the input voltage of –48V.

A standard buck controller will also not work here because both the input and output voltages are negative. For a negative-input, negative-output buck controller, the main FET connects to –Vin, and the cathode of the diode connects to GND. However, a boost controller works if you connect the GND pin to the negative input – a necessary step because otherwise all internal signals would be negative, creating a problem. Another reason is that a boost controller uses a ground-referred gate driver. Connecting the GND pin to the input voltage allows you to drive the FET without additional circuitry.

Using a nonsynchronous boost controller as a negative-input, negative-output buck controller

Figure 3 shows an example schematic using the Texas Instruments (TI) nonsynchronous boost controller to drive the main FET of a negative-input, negative-output buck controller.

Figure 3 This simplified schematic showcases Texas Instruments’ LM5155 boost controller.

Because –Vin is connected to the GND pin, all internal signals reference –Vin. Since –Vin typically varies across an input voltage range, this behavior can cause difficulties when enabling or disabling the device, regulating the output, or other protection features. Typically, you will need a level shifter (for example, an isolated type or one with bipolar FETs) to regulate the negative output voltage.

Configuring a boost converter for negative-input operation

A boost converter with internal switches can also work in theory, because the source of the main switch connects to the GND pin and the drain of the rectifier FET connects to the Vout pin.

Figure 4 shows a block diagram of a boost converter and the connection of the switches to the integrated circuit (IC) pins.

Figure 4 This boost converter block diagram includes the switches’ connections to the integrated circuit pins.

The challenge in using a converter is that many signals are internal. Some ICs integrate the output voltage divider, which makes regulating a negative output voltage difficult. Because all internal voltages reference the negative input, the output voltage would follow the input voltage. In that case, you can use the COMP output instead of the internal feedback. Connecting an optocoupler as a level shifter between COMP and GND provides one method to regulate the negative output.

Figure 5 shows how to connect a boost converter to a negative-input, negative-output buck power stage. The GND pin connects to the negative input, and the Vout pin (or FB pin) connects to power-stage ground. You can short the FB pin and use the COMP pin with an optocoupler to regulate the output. Keep all voltages, including current-sense signals, below the maximum limits of the boost converter.

Figure 5 This simplified schematic employs the boost converter shown earlier.

Design considerations

You can use a nonsynchronous boost controller such as TI’s LM5155 or TPS40210 as a simple, cost-effective solution for generating a negative output from a negative input. To increase the efficiency, replace the diode with a MOSFET, though doing so requires a synchronous boost controller that drives two switches. Negative voltages can easily cause confusion. In particular, you must check all internal signals and verify that no voltages are exceeding the controller’s maximum rating.

Florian Mueller is a systems engineer and Member Group Technical Staff in TI’s Power Supply Design Services group. He has a master’s degree in electrical engineering from the Technical University of Haag, Germany. Florian’s main focus lies on industrial high-voltage designs for different end equipment.

Source link