Finally, by increasing the internal operating frequency for these switching supplies, designers have been able to decrease overall size or, alternatively, pack a higher power rating into the same size unit. For example, a 125W supply from N2Power has a power factor correction (PFC) circuit operating at 87kHz and a half-bridge output at 43kHz. By moving both to 125kHz in the 160W design, the volume of the associated magnetics was reduced by one third, thus allowing the 160W unit to fit into the same 3-inch by 5-inch 1U footprint as the 125W supply.
Active components also are playing a major role in the shrinking of supplies in several ways. IC vendors are developing better analog components for power management, especially helpful for complex functions such as PFC. The result is better accuracy and performance at both high and low line voltages. Early generation PFC circuits could achieve power factor up to 0.99 (nearly a perfect unity value) at lower AC-line voltage of 90/110V AC, but the correction dropped to as low as 0.75 at 240V AC. Using the latest analog ICs, designs can now maintain high PFC across the full voltage range, resulting in fewer corrective components and thus smaller supply size.
The most dramatic change for supplies architecture is the inclusion of digital control circuitry in the supply’s inner workings. In recent years, processors have been used in the secondary side of the supply to monitor key points and performance, to establish some operating parameters, and to manage a communications interface port. But the primary-side control loop remained analog.
With the availability of high-performance, low-cost DSPs, digital control now extends to the primary loop, which increases the flexibility in control and operating points, including on-the-fly adaptive control and dynamic operating changes. It also affects size, by putting more of the control functions in less space, due to fewer needed ICs and passive components to implement the hardware control-loop strategy. For example, a supply using DSP control can do three-phase AC line control with about the same footprint as a single-phase unit. The DSP can also provide the required PFC with no further footprint penalty.
Don Knowles joined N2Power as the vice-president of engineering 12 years ago after more than two decades in power electronics design and manufacturing sectors, spanning industrial, ICT, and medical electronics. He holds a degree in Electronics from American River College, Sacramento, Calif.
Thanks for reporting on this. I remember hearing about 20 years ago that power supply designers were trying to make their products smaller, but it was a tough job to make that happen. Looks like progress has been made since then.
Ann, As pointed out in the article, todays electronics are getting smaller in size but the power requirements are going up. It's great that power semiconductor manufacturers like International Rectifier are developing technologies that meet both design requirements. I agree, its nice to see the arrival of this semiconductor technology in power electronics.
mrdon, those are exactly the reasons I was hearing 20 years ago that power supplies needed to get smaller, let alone more efficient. IR has long been in the forefront of advances in power supply design and manufacturing. At one time (maybe more than one?) industry pundits have said that power supplies were a major roadblock in advancing the state of electronics.
Ann, You and TJ are absolutely correct about small size vs cost. One of my biggest challenges in developing LED Lighting fixtures while working at Hunter Fan was power supply efficiency, small size, and cost.To achieve a quality power supply, efficiency cost is part of the Product Development financial equation. I was able to obtain superb power supply efficient (99%) and small size using good PFC based components and proper PCB thermal management techniques, but the cost was something the Bean Counters couldn't stomach. Unfortunately, efficiency was the design parameter they could leave without.
mrdon, it's bad enough that commoditization means cheaper--often too cheap--materials and/or construction. But when the very thing that's one of a product's key characteristics--in this case a power supply's efficiency--is sacrificed by purchasing/accounting, that just beggars belief. I mean, trading off specs and features is part of the process but sounds like you had an exceptionally bad experience.
TJ makes a good point about price, of course. On a different price note, back when, some power supply designers said they could make advances closer to what was needed in size and efficiency, but that it would cost an arm and a leg to do so.
Not only are the power supplies getting smaller, but many of them have become commodity items, with the result thatr there is a lot of price pressure. Unfortunately that tends to lead to a reduction in quality from many suppliers, while others chgarge a bit more for a better product. Consider that some of the better suppliers are able to sell a supply for amost twice the price of what others offer, a product with similar size and power ratings. So evidently in some areas quality and robustness still matter.
Lower power/smaller size is only expected. What is troubling about the trend for the DIY/learning crowd, how does the average human place a 0603? Not easily. Will through hole components stay, or will pre-mounted small form factor take over? I am sure the price or old-style components will increase. The size and style of the through-hole resistor hasn't changed in decades, and they stay relevant.
William, that's exactly what's been needed--for power supplies to become commodities so they'd be cheap enough and small enough. At least that was the push 20 years ago. But at that time, I didn't hear anyone complaining about what would happen when components were cheapened to reach that goal. Now I think we're all aware a lot more, as both designers and consumers, of the effects of this problem, whether it's cheapo plastic gears in clocks or sub-efficient power supplies.
What I find is that to provide the quality in my product that justifies the price I need to use a power supply that is quite a bit more expensive than the cheapest one that would work. But the adequate margins abd better construction have meant that no failure have occured in ten years. That has been quite good for the products reputation. It IS INDEED cheaper to do things right the first time.
William, I agree. But apparently the math needed to figure that out is too complicated for some companies, or they are too short-sighted. In this case, one hardly needs 20/20 hindsight to come to your conclusion.
Ann, there are a whole lot of products that are simply not worth repairing. Others aren definitely worth repairing and happen to be conveniently repairable. THAT did not just happen: designing for repairability is cloesly linked to designed for assembly. Only just a bit more effort.
But it also has an extra benefit, which is a design using components available from multiple sources. So that when I can't get parts frpom one maker, I can use parts from another maker. That is quite handy.
Ann, sorry about that. I don't do much sarcasm, nor irony. Quite possibly I am way to serious about things.
I am quite critical of both laziness and stupidity, though. I do catch grief for that on occasions.
Really though, there are a whole lot of companies that appear to be successful that have products that are very challenging to even diagnose, let alone to service. Some of them get into the "made by monkeys" section of this fine publication, some don't.
But sarcasm and irony are so much fun :) I'm also very annoyed by products that don't work as they should and/or or difficult/impossible to repair. It makes you wonder why anyone ever bothered to make them in the first place.
The legacy endpoint devices that control our critical infrastructure (utility systems, water treatment plants, military networks, industrial control systems, etc.) are some of the most vulnerable devices on the Internet.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This radio show will show what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.