Electronic Design

Backstory: Full-Wave Active Rectifier Requires No Diodes

When Electronic Design asked me to write about my Idea for Design, first seen in this year’s August 13 issue, it seemed a simple enough task. We’re all familiar with design. Most of us do it in some form or another almost every day. But, ideas? We all have them, but where do they come from? And what is the essence of a good idea? Simplicity? Elegance? Performance, novelty, creativity?

Thomas Edison once said that genius is 1% inspiration and 99% perspiration. For me, inspiration is the key. Having a good idea and turning it into a great design requires a seed of inspiration. As a consultant engineer specializing in analog design, most of the circuits I design follow a customer’s specification. Creating a product that performs to the customer’s requirements is a challenge that requires inspiration if I am to satisfy those requirements with a circuit that performs reliably, efficiently, and at the right price. And that’s usually where the good ideas come from.

But it isn’t always that way. Sometimes, an idea materializes simply from the desire to do something differently, or better, or cheaper. It can happen to me quite suddenly. I tend to frown, quizzically, and look very distant. Julie, my nearest and dearest, can always tell. “You’re getting one of those ideas, aren’t you?” she says. I usually nod, sheepishly, and then run off to my study. I keep lots of paper and sharpened pencils handy. You never can tell when an idea might strike.

So it was with the rectifier idea. I’d been going through my burgeoning collection of electronics literature, reflecting on the merits of different active rectifier circuits. Some of them were half-wave, others full-wave. Most of them used diodes inside an op amp’s feedback loop. Others used clever arrangements of op amps and analog switches. All very neat, all very effective. But something nagged at me. If I were to feed an alternating signal swinging symmetrically about ground into a single-rail op amp, what would the output look like? Wouldn’t the negative peaks be clamped to ground? And wouldn’t that, effectively, be half-wave rectification?

I ran off to my study and started scribbling. Julie, bless her, brought me tea and a slice of pizza. The first part was easy. Op-amp IC1a, configured as a unity-gain follower operating on a single positive rail, would, indeed, follow the positive-going segments of the input waveform and would level off at ground during the negative-going portions. I added resistor R1 to protect the op-amp’s non-inverting input from excessive current flow during large, negative-going input swings.

So far, I had a promising idea for a half-wave rectifier that didn’t need any diodes or analog switches. Super. But I wanted full-wave rectification. Nothing for it but to add a second op amp, IC1b. By combining the second op amp in a unity-gain inverting function with the output of the non-inverting follower, the overall function would simply invert the negative-going input peaks when IC1a’s output was at ground to produce a positive-going version and would effectively allow IC1a’s output to follow through during the positive-going peaks, producing full-wave rectification.

Genius? Absolutely not. All I did was to take some wellestablished techniques and combine them in a way that suited my idea. Certainly, the resulting circuit is fairly unique to me, but the foundations it is based on are not. “If I have seen further it is by standing on the shoulders of giants,” said Isaac Newton. The giants of analog design are the likes of George Philbrick, Bob Widlar, and Barrie Gilbert, and latter-day luminaries like Bob Pease and Jim Williams continue to guide my way.

So, everything was looking good for big signals, but what about smaller signals? I wanted to add gain so signals of a few hundred millivolts could be boosted to a few volts. I knew that adding R4 and R5 to IC1a would ams needed for the required amount of plify the positive portion of the input signal, but the ratio of R2 to R3 that dictates the inverting gain of the second stage would have to be changed so the positive and negative peaks would be amplified equally. Time for some math—nothing fancy, just a little algebra. By equating expressions for the overall gain during positive and negative excursions, I derived a simple equation that could be used to determine the resistor valuegain.

Is there a route map for good design? Possibly. But there are different ways to turn that spark of an idea into a living, breathing circuit. There are no hard and fast rules. Find the way that works best for you. I like to scribble, tweak, do the math if necessary, then maybe run a few PSpice simulations. Only when things are looking good at this stage do I move to the breadboard. I rarely rely on simulation results to prove a circuit. For me, the breadboard is the final arbiter. For really simple circuits, I sometimes use the matrix-type breadboards where you simply plug in the components, but they have their limitations. I much prefer to solder components onto a piece of copper-clad board. As well as providing a secure anchor for the parts, it functions as a low-impedance ground plane.

What would be my advice for good design? Try the three R’s—read, research, revise. Plenty of good literature is available on electronic design, not just in textbooks, but also in manufacturers’ application notes and magazines like Electronic Design. Thorough research is essential, particularly when it comes to choosing the right components. I used the OPA2374 dual op amp in the rectifier circuit only after carefully assessing characteristics such as phase-reversal behavior and overload recovery time. Finally, revise. There’s no such thing as a perfect design. Circuits can always be improved.

And remember to think “out of the box.” It’s an overused phrase, but important nonetheless. That’s effectively what I did with the rectifier circuit by using a single-rail op amp in an unusual way. Most designs would strive to avoid saturating an op amp’s output, whereas the rectifier circuit actually exploits that behaviour. I like to design circuits just for the hell of it. It’s fun, and I continue to learn from the process.

So, is this idea thing an addiction? Probably. As long as I keep getting a buzz from analog design and that sweet rush of anticipation whenever I warm up my soldering iron and turn on my oscilloscope, I’ll carry on turning ideas into designs and, with a little inspiration, I hope to continue sharing them with you.

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