[036] Forward Converter Design - Part III Further Testing

Further testing on the forward converter.  

Introduction

This article continues the series in which Dr. Ridley documents the processes involved in getting a power supply from the initial design to the full-power prototype. In part III, initial power is applied to the circuit with a resistive load to verify proper operation of the primary circuitry. 

Power Supply Testing with Analog Controller

In the previous article of this series, several problems were found on the control board prior to applying power to the input. The first important lesson was that trying to debug a digital controller at the same time as trying to debug a power stage was not a reasonable approach. The digital controller was replaced with a standard analog controller in order to debug the power stage while minimizing the number of variables.

After several years of power supply design, you become quickly aware that you cannot “breadboard” designs with proto-boards and long wires. Critical noise-sensitive areas MUST be laid out on a PCB to provide minimal path lengths for high frequency connections, and ground planes for shielding. In all projects, you want to minimize the development times. Decisions must be made about whether to stop testing and return to board layout, or provide compromise interim solutions in order to collect data as quickly as possible. One trick I have used frequently is to build power supply test circuits from a collection of previously-designed printed-circuit boards. For this project, a control circuit containing a 3825 controller was cut out of a large PC board, and connected to the power circuit after removing the digital controller parts. This allowed testing to continue after just a few hours work.

Applying Bulk Power to the Circuit

The schematic of the power stage is shown in Figure 1, together with the signals for interfacing with the control circuit. Snubbers are omitted from Figure 1 for clarity.  (This circuit was shown in full in the previous article in this magazine [1].)

Figure 1: Power stage and control interface schematic. Snubbers are omitted for clarity. The power transformer and secondary circuits are highlighted in blue – these will be removed for initial power stage testing.

After testing the gate drive waveforms, and correcting circuit problems in part II of this article, power was applied to the circuit. There are a couple of choices for doing this, one of which is to use a DC bench power supply to apply a fixed voltage. Such a supply must be capable of providing high voltages, up to 400 VDC, and high currents in order to operate at low line and full load. For a 350 W design, a 1 kW bench supply is typically needed, which is expensive, bulky, and potentially hazardous.

I always prefer working directly from an AC source at the very beginning of a project. This will often bring out problems with the AC connections that you do not want to delay finding until later in the project. An isolated AC source provides a good compromise of safety and instrumentation options, as discussed in [2].  

Figure 2 shows the circuit for the initial power testing. The transformer of the power supply was removed, and replaced with a power resistor. This allows us to test the following elements of the power supply:

Instrumentation connections and grounding

Gate drive waveforms with voltage applied to FET

Turn-on thresholds and gate drive levels

Current sensing circuits

FET operation with voltage applied