As much as people love USB-C, there’s one massive flaw that becomes very obvious the moment you look at the ports on any computer. This being that there’s no (standardized) way to tell what any of those ports do. Some may do display out (Alt-Mode), some may allow for charging, but it remains mostly a matter of praying to the hardware gods. According to a recent blog post, this is where Microsoft will seek to enforce a USB-C feature set on all (mobile) computers compliant with its Windows Hardware Compatibility Program (WHCP).
This also comes after years of the USB Implementers Forum, re-branding the USB specifications, with the most recent iteration thankfully using the bandwidth (in Gbps) as specifier (meaning no ‘USB PlaidSpeed’, sadly). Claiming to follow this end-user friendly spirit, the Microsoft blog post goes on to a minimum set of features that USB-C ports should have, as detailed in the above table.
Most notable is probably that PC charging support is required, as is support for at least one external display. As for the negatives, this seems to only apply to laptops, and no actual charging requirements are set (USB-PD voltages, wattage, etc.), so what the actual impact of this will be remains to be seen.
One thing remains certain, however, and that is that by trying to make USB-C the One True Connector for literally everything, there will always remain cases where end-user expectations remain unfulfilled.
Why not make something like USB PowerDuo Pro where one cable (all wires, including shield) is used to supply N-voltage and another is used for L-voltage to voltage converter inside the device. This way it should be possible to deliver up to 3,6 kW of power.
This seems like an interesting idea but my Google-fu was not sufficient to help me understand the specifics. Would you be willing to elaborate. Delivering 3.6 kW through a little connector like USB-C seems like a quite a feat. I am impressed that the USB-IF managed 240 W. 48V already seems kind of high considering the 0.5 mm pin pitch and I doubt there is much headroom for more than 5A of current through those tiny pins.
No you have ALL the internal wires at the same potential so the .5mm pitch doesn’t matter, if you limit it to 5A you need the voltage to be 720V. Sounds like a great standard, guaranteed to kill all incompatible devices.
All the internal wires at the same potential? What does that mean?
It would have more sense to use one or two of the configurable pairs for power. If currently you can have up to 5 amps in four wires, having five or six wires would ramp up to 6.25 to 7.5 amps. That’s 300-360 watts with 48 volts per pair.
Even if it was a fancy setup where initially USB PD negotiation hikes the voltage to 48V that would require pumping 70A through that tiny connector. Not gonna happen.
Remember that regular USB PD can deliver 250W so two cables could do 500W. A more reasonable ask, therefore, is to allow multi cable charging, which is not a common thing right now, I believe. I think Tesla trucks can do it, but of course not using USB PD :)
I assume you mean something like this:
https://xkcd.com/2493/
It’s beautiful
So I guess your idea is to run the max current at max voltage, times the nr of conductors / 2…
Problem is heat: having 5A once (or twice, including return) heats up the contacts and the connector around it, but doing that multiple times will quickly mean molten connectors…
So if all wires & shield are providing power I guess there is no USB PD negotiation?
Dear God I hope nothing as monsterous as that is produced, it would be a sure way to damage any poor unsuspecting USB C device that is unfortunate enough to plug into it. I’ve heard of shoddy USB C PSUs designed for a specific product that provide a higher voltage (9,12,20) without negotiating first, but your idea sounds even worse to this uneducated layman.
Also, fwiw, almost all residential circuits in the US are 120v and 15a fused – which would only get you 1800w. Even that seems crazy to ask of a USB C cable, I don’t need or want to use USB for an electric heater or microwave.
Does anyone care about WHCP? Serious question. When I build a PC, I don’t look for WHCP certified components. If a business is buying loads of PCs, do they look for it or just buy computers that always come with the latest version of Windows on them, so must be compatible enough. Or will MS not allow OEM copies of Windows to be sold on non-WHCP devices?
It is for tablets and notebooks.
Got it? OK.
And what the manufacturers get is a nice sticker and when businesses/governments buy stuff they don’t look at much more than compatibility with (BS) certificates.
Hardware manufacturers must follow WHCP to be allowed to preload Windows, and would be practically cut off the market if they didn’t, for better or worse. This effectively enforces universal compliance, even for component manufacturers.
Needed, because right now one has to read the documentation carefully to see if a feature will work, and even then sometimes one has to just, buy, try, return to find out.
Now all we need to put the Universial back in USV is similar initiative to easily identify capabilities of USB-C cables!
Yesssss! Finally! M$ to the rescue!
If only Google and Apple would follow. By introducing MORE arbitrary things that ONLY they use!
Seriously. Can’t we just come up with 1 single, UNIVERSAL standard that is like 20 years ahead of the used AND needed specs? I don’t care if the cable or the head will be as thick as my forearm.
Perhaps rather than trying to push power down a data wire, we could adjust a Tesla power charge cable to take faster data?
That would be good for running an Nvidia GPU.
USB-C style outer shell as ground, so no flipping. Three inner pins, two metal for power (better than thin wires), and the third optical both for speed, upgradable technologically, and protective from any electrical faults including interference.
https://xkcd.com/1406/ Connects to everything (except USB C)
Still do NOT understand what was wrong with the USB-A, wide enough plug, long enough contacts, etc. IMHO, Leave it alone and make everything backward compatible, and allow vendors to expand the use to deliver higher power. Phones may need USB C, but the rest is perfectly fine with the USB-A, good standard, well established globally, etc.
Ideally, leave things where they are, and invent something BETTER, same plugs, USB-A or USB-C, maybe add high-power inductively coupled thing instead of wearing out thin contacts, leave the original USB-C to deliver the handshaking and set the limits for the inductive coupling, etc. How exactly? That’s where engineers should be driving the bean counters, and not the other way around.
As far as microsoft enforcing goes I am not exactly sure what the results will be; I remain highly skeptical they won’t just destroy what currently works more-or-less-realiably-well, with their proprietary add-ons that would cripple what they don’t want.
Having said all that fluff, why don’t we just have some kind of inductively-coupled magnetic plugs? No contacts to wear out, etc. If I am to redesign what I NEED vs what vendors can make, I’d start with a flat mating surface that’s disabled by default, so it won’t react to, say, bluetooth headphones near. I’d also design simple and reliable open source protocol for the handshake that would set up voltage, amperage, etc. Doesn’t have to be stuck with 5V, could be 12V or 24V, whichever floats through the thinnest wires possible. Actually, it could be any voltage/amperage, and this would also make sure the battery is charged up to specs, and not under or over, and measure the final result – intelligent charging so to speak. I’d also make sure batteries are ID-ed with chips that identify what charging voltage they need, how to tell when they are full, etc etc. Two-way coupling, not one-way-ram-power-through-hope-it-works, so the charger reads the instructions from the battery and adjusts accordingly.
But what am I doing speaking common sense. It is not needed, microsoft does all the thinking.
Microsoft adding requirements to their hardware compatibility program is not the same as them dictating port requirements for the whole world.
I think you may be missing some of the point of USB C. The A type connector is great, but the new specification for the C type comes with added benefits:
It’s reversible. You are apparently the only human who doesn’t mind plugging an A type connector in backwards.
The USB PD spec only applies to USB C, so you won’t be using a type A connector to charge something that needs more than 5v 2.1a, like a laptop (I’m sure you could, but it’s not in spec so no one is making devices or chargers that do so)
USB C breaks from the host/peripheral model (A/B), allowing both sides to negotiate their roles. USB OTG and some out-of-spec hacks allow for connecting a peripheral using a USB A port, but the USB C standard eliminates any confusion or hackery on this regard.
It’s a robust connector – I work with corporate devices that see tons of abuse, and have not seen a bad USB C connector. Broken and bent cables aplenty, but it has been much more durable than comparable connectors like lighting (and especially micro USB B). USB A is probably more durable, and probably will continue life for a very long time, just not as the predominant end user connection.
I like USB C. I’m not seeing unicorns and rainbows, but it’s a heck of a lot better than the alternatives we have available, including trying to drag USB A into a universal role.
USB C is too small and fragile to have on a desktop. I’m always worried I’m going to rip the port off the board if I bump the cable. I don’t like seeing them on anything larger than a phone. It would have been great if it was the size of a USB A connector.
The USB C charging connector on my laptop is already starting to wear out and it’s not even that old. Wiggling the cable makes it pop up a notification that it’s on battery. The port is soldered on the motherboard and will require a good hot air rework station to replace.
They should be mandating that all USB C connectors are to be installed on an inexpensive, user replaceable daughterboard and require replacements to be sold for at least 10 years.
Awww, and I was looking forward to both LudicrousSpeed and PlaidSpeed, too! ;)
PlaidSpeed???
What is that? My Google-fu only finds references to a Tesla model.
A cultural reference so old and obscure that you’re better off forgetting it exists.
Spaceballs.
One of the greatests from Mel Brooks.
The other ones also rock.
Thank me after the bingewatch. ;)
Thanks all; I did not realise it was such a niche comment.
Mel Brooks output is like the proverbial curate’s egg — good in parts. I must confess that I did not enjoy Spaceballs.
i dont even like spaceballs and i get the reference.
This standard doesn’t go far enough in my opinion. “PC Charging Support: yes” only knocks out 1 of 4 bad modes.
1. Laptop doesn’t charge over USB C at all (eliminated, good!)
2. Laptop only charges if there’s a 20V 5A profile (seriously? Can’t accept less current?)
3. Laptop only charges if there’s a 20V profile (15V 3A not good enough for you?)
4. Laptop charges on 5V but clocks CPU at minimum speed to save power, becoming near useless
Macbooks are the only devices I’ve seen that accept any voltage and retain full performance by draining the battery if required, which is what all laptops should do along with warning the user if they’re using a low-power charger.
XKCD comic about there’s too many variants, we need a standard, then there’s another variant dot png
I can see why they felt the need to do something; but this seems like it breaks one of USB’s important capabilities: being the cheap port that can be supplied in quantity for undemanding peripherals.
Obviously it has bitten off being the expensive and capable port as well(with fairly high wattage PD, DP alt mode; and USB 4 including thunderbolt’s PCIe behavior with the trademarks filed off); but the desirability of cheap ports hasn’t really gone away.
With this standard; USB 3-capable devices now have to run DP and the full power delivery circuitry to allow the device to be charged to every type-C port. The only allowed way to give someone just a 5Gb/s port to plug a flash drive or a mouse dongle or whatever into is to use type A.
USB4 devices are in the same situation, but more so: they need all that and the 4 PCIe lanes(or a lot of PCIe switching; the spec does not actually require that all the ports remain capable of either video or PCIe; just that you can start plugging in anywhere).
If you are in it for the backward compatibility this incentive to retain type A may not be a bad thing; but if you want to standardize the set of cables you pack a set of requirements that essentially declare that ‘just’ USB is forbidden to be type-c seems very likely to push larger devices toward retaining port mixes longer; and devices too large to readily accomodate type As to potentially just omit type-Cs they don’t feel like wiring fully.
Could they really not have preserved a split between ‘is basic USB’ and ‘is the everything port’?
i still like usb2 mostly because its four wires and easy to build your own devices. 99% of the time when i want more ports, 2.0 is good enough. the use cases for really fast ports is maybe one or two per system (its like the firewire port i liked to have but never used). but most of what you plug in is low speed peripherals. keyboards, controllers, printers, etc. things that usually either sip power or need their own plugs. if you need four pcie lanes the device is better off being inside the chassis rather than dangling off a port.
the main thing driving usbc was portable device users where you need one port that does everything. desktop users still prefer robust ports that do one thing very well and let other ports do the other things. dedicated power and video cables for things that need them. its nice to be able to plug your screen into a usb port.
usbc is also terrible for industrial applications and why its a good thing to keep type a and b ports around. unless you want to mold a usbc connector into a d-block with a couple dedicated power connectors. my 3d printer managed to vibrate its micro usb connector right off the controller board and it made me realize why 2d printers seem to be the only thing that uses a b connector. whatever improvements the c port got, i doubt its going to fare any better in a similar application without some through-hole lugs to resist metal fatigue of the exotic solder formulation.
so im fine if they keep 2.0 around and usbc ports are rare on a mobo. usb3 on a-ports is fine for external drives. again if you need four pcie lanes, m.2 is a thing. mechanically speaking one port to rule them all is a pipe dream. 2 or 3 different sizes with increasing capabilities with size kind of makes sense with the big ports also having dedicated power lines for industrial applications.
Ironically I had an OEM motherboard that claimed no display outputs, yet it had USB-C displayport alt-mode.
I did have to find the specification page on archive.org, the OEM had deleted it.
That is a super secret security feature.
No noobie hacker can decrypt your data if they can’t see it!
And how exactly does this help users who switch to Mac?
I mean, Windows 11 isn’t exactly popular and the rest who can’t afford a Mac (or don’t want it because it’s different) is going *nix route anyway.
Which in turn has no dependency on x86_64 platform to begin with.
WHCP also doesn’t help Raspberry Pi users, either.
Users who hold on to Windows 10 (no idea why) don’t want Windows 11 hardware, either, because it might be Windows 10 incompatible.
This will probably just mean that laptops and tablets will have at most 2 USB ports, except perhaps for the very high-end models.