Electronic Design
Low Cost Was Primary Goal In DC-DC Converter Design

Low Cost Was Primary Goal In DC-DC Converter Design

When Dwayne Reid first approached the design of a dc-dc converter for an LED lighting controller, he had one clear goal in mind: low cost (see “Build Your Own Ultra-Low-Cost Isolated DC-DC Converter”). His product’s production volumes would be modest (in the hundreds) and with such volumes he could not afford any but the most common and widely available parts.

Reid developed this design for LED architectural and display lighting, which called for the use of the DMX (digital multiplex) signaling standard. The DMX specification calls for all local controllers in the lighting network to provide isolation to avoid ground noise from affecting the control signal’s integrity.

An opto-isolator and the control signal’s RS-422 receiver needed supply power. However, off-the-shelf dc-dc converters would cost around $20 in his quantities—too much for his product. He needed an isolated-output converter with as low a cost as possible, and he pursued that low-cost goal diligently.

He reports, for instance, that once he realized he could use a simple 1:1 transformer to provide the isolation he needed, he dug through his sample parts bin and found a handful of common-mode chokes. Recognizing that chokes are essentially 1:1 transformers, he looked at them more carefully to see if they would serve, finding that they could handle his requirement of 5 V at 50 mA with no problem.

Similarly, he decided to use the LM393 dual comparator in the design to save cost. He already needed a comparator to create a current-mode switcher. The dual comparator allowed him to use the second half to replace what would normally be an operational amplifier in the ramp generator. This gave him the equivalent of two devices for the cost of one.

The use of an LED to establish a voltage reference for the comparator represents yet another step toward shaving the cost. A final slice came from leveraging the comparator’s open collector output to drive a PNP transistor rather than using a MOSFET transistor in the final power stage.

Reid reports that he has had many e-mails about the design since it first appeared in print. Some were complementary about its inventiveness. Some were challenging. “A number of people told me that couldn’t work. I told them to breadboard it and see for themselves. None of them ever got back in touch,” he said.

The design has been used in production for the LED lighting controllers and has been re-used in several of Reid’s other product designs. He has not found a way to optimize the parts count any further, but still looks for opportunities to cut cost.

“I have used different magnetic components,” said Reid. “I use whatever I can purchase at low cost, even surplus inventory parts, as long as I can purchase enough to cover the projected product lifetime.”

Reid is proud of his design. “It really is non-intuitive,” he said. “It looks at first glance as if it just can’t work. But work it does—and very well.”

Hide 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.