Electronic Design

Frequency-Modulated DC-DC Converter Offers Flexibility

It's certainly easy enough to buy a switcher off the shelf, but what do you do if you need a little extra voltage with a little more control? A programmable switcher would be nice. The following design is a very simple attempt at providing a little extra voltage with flexible control, while requiring very little hardware overhead.

A basic voltage-feedback boost converter design is chosen with a vastly simplified feedback circuit. The feedback circuit requires only a comparator, some firmware within a microcontroller, a couple of resistors to select the voltage feedback, and a capacitor to set the desired maximum ripple for the load (Fig. 1a).

The output voltage is set by the resistor divider equation:

(VOUT)(R1)/(R1 + R2) = VREF

The maximum ripple is easily set by solving the linear form of the standard capacitor equation:

(I)(Δt)/(ΔV) = C

where I is the average load current. The comparator and microcontroller could be selected individually. Or, they can be found in a single package such as the PIC10F206 six-pin microcontroller or the PIC12F629 eight-pin microcontroller.

Although a voltage-feedback technique is chosen, it's not used in the traditional sense—where the duty cycle is modulated to meet the load requirements. Instead, a finite amount of energy is stored on a per-unit time basis, where the time is modulated based on the feedback to meet the load requirements.

So as the load increases, the total period is reduced and vice versa. Yet the energy stored in and transferred from the inductor always remains the same. In essence, the frequency of the switcher changes with load changes. In a basic sense, this method is something like a current-feedback design without the extra current-feedback loop.

This functionality is easily controlled in a firmware-based state machine. Only a few states are required to control the system (Fig. 2). The first state is the detection of an undervoltage event generated by the comparator. This immediately drives the FET on, storing energy in the inductor. This state is held for a fixed period of time.

Once the fixed energizing time has ended, the FET is turned off and a minimum off time (discharge time) is held. Then the comparator is enabled again for the final state, waiting for the undervoltage event. The result is something like the waveform in Figure 1b.

So, where does the "flexible control" portion of this design come in? This is, by nature, part of the firmware within the microcontroller. The state machine can be designed to be a bit more dynamic to accommodate some special needs, like soft-start, switchable output levels, or even overload or underload protection and indication. Plus, because the microcontroller is a programmable device, the firmware could be changed to accommodate product variations without adding hardware.

For example, one product may have an internal subcircuit that requires 9 V from 5 V, while another variant of the same product may have an 18-V requirement. This type of variation could simply result from a change in firmware that alters the energy stored in the inductor.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish