Hacker News Reader

Comment by lampington | Hacker News Reader

Does anyone know what it that's led to the big improvements in power supplies over the last ten years or so? Wall warts have shrunk tenfold.

View on HN

Wall warts used to work as a step down transformer and a linear regulator (http://en.wikipedia.org/wiki/Linear_regulator). The step down transformer is the big brick like piece of the wall wart.

More modern transformers are switch mode supplies (http://en.wikipedia.org/wiki/Switched-mode_power_supply). Switch mode supplies (like the OP describes) are smaller, run cooler, and (nowdays) cheaper than linear supplies.

The main thing that's happened is IC manufacturers have made switch mode controllers cheap, and reliable. As a result, most wall warts are going away, to be replaced by switched supplies.

There isn't any significant hardware-wise difference in most SMPSs between today and say 20 years ago. The main cause for essentially every wall wart being SMPS is simple economies of scale and globalisation: * Difference between power supplies for different countries amounts mostly to plug shape, you don't need different transformers for 110 and 230V. * The whole power supply is smaller and lighter, so it's cheaper to ship it around the world * SMPS is simplest way to meet requirements on cos phi, efficiency and idle power in some jurisdictions.

In end when you are designing product that does not have any special requirements on power quality (like precision instruments) or customer expectation that it should be big and mostly empty (home AV stuff) it is simply cheapest to order SMPS wall-warts from China in bulk.

It makes me happy that there's some technical discussion on this thread. One thing I'd like to add is that linear power supplies are better for AV systems because switching power supplies have a lot more RFI and electrical noise, which can be a problem when you're dealing with small analog signals. (This noise is also why cheap wall chargers can cause your phone's touchscreen to malfunction, since they have very small signals which can be overwhelmed by noise.) So they aren't using linear power supplies just to make the AV box more impressive, but for functional reasons. (See www.ti.com/litv/pdf/snaa057b which explains why unregulated linear power supplies are good for audio systems.)

My point with AV equipment was that these things tend to have somehow standardized sizes and mostly empty cases. Thus shipping cost tend to be dominated by "volumetric weight" and not real device weight. EMI/RFI from SMPS is valid concern for these applications but not as significant as it is often claimed to be.

More significant issue is cost: you need power supply that supplies slightly weird voltage rails that essentially has to be custom, even off the shelf SMPS modules tend to be significantly costlier than wall warts and you don't have any reason not to just stick big transformer in there (even more so because you essentially don't care about regulation of the more loaded rails)

WRT 20 year old supplies, there are pessimistic, yet true, comments to be made WRT the likelihood of the UL registration being fake, or the RFI suppression components being replaced by jumper wires, and value engineering to run hotter to require quicker replacement (electrolytic tech WRT handling heat hasn't improved much.

There is one tiny correction to your post that I can make; the reason why you order SMPS wall arts from China is because there are 20 places to compete on "12 volt, 2.1 mm coaxial jack, 1.5 meter cord" so you can play mfgrs off against each other, but there's not so many places willing to play internal mounting games WRT exactly where each plastic mounting post is located.

IF internal SMPS had something as effectively standardized as NMEA electric motor mounts, then you MIGHT see more internal power supplies because competition would be cheaper.

I think also the EU put in a regulation limiting idle current draw (phantom load). The linear units leak when not in use, you can feel the warmth. Switching units don't leak significantly. That forced switchers while they were still slightly more expensive.

If the switcher was $0.50 more expensive, you were getting a linear even though it would cost you an extra $2/year in electricity. Consumers demand cheap initial cost in their electronics. The EU regulation just forced the upfront cost.

Ah, right, thanks! So old wall warts were like the power supplies the OP describes right at the start, and the change we've is that newer wall warts are like a scaled down version of the ones he spends the rest of the article describing.

And I would add that switching power supply tech has been around for decades (hello, desktop PC tower cases!). It just wasn't used as often in consumer devices because they were more complex (more parts, trickier to design) and there simply weren't many consumer devices around that required so much power _and_ portability.

I would argue laptops were the killer app as far as switchers are concerned. Once they showed up and got popular, designers newly had reasons to design switchers more often.

A lot of the old wall warts are "Type 2", which means transformer, rectifier, reservoir capacitor -- no regulation there at all. With light or no load, the output voltage can soar (such as, 18v DC no-load from a nominal 12v DC @ 300 mA wart). The switchers are regulated, so you get what it says on the tin.

As transistor switching speeds get faster, the switching supply can operate at higher frequencies, which means less energy has to be stored per cycle, which means smaller components can be used.

Reply on news.ycombinator.com