# Is it possible to build laptop power adapters that are considerably smaller and lighter?

I am considering an alternative to the well know 85W Apple power adapter (or lets call it a brick, as it is big and heavy).

I am wondering if there is any technology for producing smaller power adaptors of the same power that that would be much much smaller and lighter? graphene maybe?

• the powersupply is mostly transformer and capacitors. graphene could perhaps help with the size of the capacitors, but the transformer can only be made smaller if the running frequency is increased. – Jasen Jan 26 '15 at 10:38
• What makes you think Apple aren't at the cutting edge of this already? Given their aggressive technological development of things like phone glass and aluminium laptop bodies. – pjc50 Jan 26 '15 at 10:51
• @pjc50 "What makes you think ...?" -> It's a matter of priorities. If they must make bleeding edge techo products to achieve their primary aims Apple may do so, but if they can make it pretty and expensive looking and sexy to the extent that the true believers flock after it, as they often can and as they invariably do, the technical performance will only follow if it adds about no cost. – Russell McMahon Jan 26 '15 at 12:36
• Wait, people thing the apple power brick is big? I think they're positively tiny. You should see the 210W power brick a "Desktop replacement" laptop I had used. It was literally the size of a actual brick. – Connor Wolf Jan 26 '15 at 12:42

Using inductively coupled supplies, higher frequencies allow reduced core sizes. This requires more capable switching devices. At present that would probably mean IGBT's in your consumer supplies.

Capacitive coupling is capable of magic if the switching frequency can be made MUCH higher - maybe a factor of 1000. People are starting to deliver. As CPUs that run in the GHz range have become commonplace the ability to switch power at such frequencies at half decent prices has followed, albeit reluctantly.

Consider a 0.01 uF capacitor - somewhat higher than the gate capacitance of a rather annoyingly capacitive MOSFET.
Energy transfer for full charge discharge = 0.5 x C x V^2 x Frequency
At a switching rate of 1 Ghz and for say 5V swing this gives
E = 0.5 x C x V^2 x f = 0.5 x 10^-8 x 25 x 10^9
=~ 120 Watts.
Hmmm You do not want to discharge or charge the cap by more than a small percentage per cycle and to work over mains to equipment voltage ratios you need many stages and ... all of which makes life hard.
BUT once you get to 10 Ghz that's 120 Watt with a 0.001 uF / 1 nF, and if you allow early stages 30V swing (you don't) then that's about 4 kW for a 1 nF or 400 Watts for a 10 pF or 40 Watts for a 1 pF. At 1 pF you start to be able to make some very nice low loss caps and "rather compact" and ... .

Give it 10 years and the 10 GH cap switch supplies will be competing with the 10 MHz+ inductive switchers for compactness. Maybe.

You can buy data couplers now that power their isolated portion by capacitive power transfer in the GHz range.

Google is running a design competition to build an inverter with a power density of 50 W per cubic inch.

It will be a while before your laptop PSU fits into an Altoids tin though. The density for the competition is still a way-out goal, nothing can do that yet. It's also a many-kW inverter, so economies of scale will dilute things like the connectors, PCBs etc.

Finally, a 90% efficient PSU still has to dissipate 10 Watts. If it got much smaller, it would have to get much more efficient too, otherwise it would get unreasonably hot (or need a fan).

Edit The winners of the Google Little Box Challenge managed ${1423 W/in^3}$, a 2 kW solar inverter fitting into only 14 cubic inches. Not much bigger than your current laptop power supply, eh?

• It seems the the website is down. – sorin Jan 11 '17 at 18:25

Crank up the switching frequency. There are methods of building switching power supplies in the RF regime, but they are more complicated to design and are more likely to produce EMI. I do not know of any of these converters operating in the 100W regime, however.