The next generation of wall chargers

Still waiting for the next generation of USB-C hubs. Currently, almost all cause Wi-Fi issues, most will have Windows complain about "limited display connection".

Ironically, the best USB-C charger for something like a phone (laptops come with chargers, anyway) I've found is simply the Raspberry Pi 4 USB-C charger - very cheap ($10), small, plenty power (15W), fixed cable and no fucking blue Anker LED that can illuminate your bedroom.

Monoprice has a really nice line of multi-port chargers called "Obsidian" whose feature list includes "no lights"

In particular, the 60W USB-C + 3x12W USB-A brick is really nice.

I know that "silicion" isn't just pure silicon given dopings but I can't help but ponder the environmental and toxicological differences between the two. Gallium nitride seems loke it might be more toxic and gives me a sneaking suspicion they will be a "compact floreacent tube" solution that is a flash in the pan decade standard that gets replaced by something far less toxic anyway (LEDs). However I know that I don't know the details so I cannot judge only raise concerns which I hope were already addressed.

I understand the logistics issues involved in standardization of repeatably used sockets but USB chargers are so very "unhygenic" from a security perspective. Proper design from the data end shouldn't allow any unexpected changes but that is error prone in practice. It is sort of like sharing cups - in theory the immune system should take care of any pathogens but in practice it is a needless risk.

Getting into engineer dreams, an ideal design would involve readily apparent subsets between charge and data pins even if they occupy the same socket. So any charger honeypots to plant or extract data would be obviously untrustworthy.

GaN has a higher breakdown voltage than Si, and a higher drift velocity, which allows a faster switching speed (for the equivalent bias).

For power supplies and switched mode RF amplifiers, you need to switch the transistor on and off as fast as possible; ideally an open or short circuit. During the time is is transitioning from on to off, it is dissipating power, thus losing efficiency in your circuit.

If you can operate at a higher bias, your passive support circuitry can operate at a lower current (lower I^2*R losses) not to mention smaller due to higher frequency (smaller inductors and capacitors).

The faster switching also generates more harmonics, which can align in phase to cause higher voltage transients, thus you need a higher breakdown.

The issue with GaN is fabricating it. It is a compound semiconductor grown with epitaxy (layer by layer) and has to be lattice matched to a carrier substrate, unlike Si which is diffusion doped into the bulk Si substrate. So GaN is maybe at 8” wafers while Si is much bigger. Some GaN is on Si carrier (cheaper but lower thermal conductivity) and some in SiC (expensive but better thermal conductivity). The goal is to grow it on diamond, but again the lattice matching is a problem.

The longer term goal is diamond semiconductors. Maybe 30 years from now that will replace GaN, as GaN is doing to GaAs and Si for certain applications.