Using USB for powering devices is wonderful, as it frees us from a tangle of incompatible barrel & TRS connectors, not to mention a veritable gaggle of proprietary power connectors. The unfortunate side-effect of this is that the obvious thing to do with power connectors is to introduce splitters, which can backfire horribly, especially since USB-C and USB Power Delivery (USB-PD) were introduced. The [Quiescent Current] channel on YouTube recently went over the ways in which these handy gadgets can literally turn your USB-powered devices into a smoldering pile of ashes.
Much like Qualcomm’s Quick Charge protocols, USB-PD negotiates higher voltages with the power supply, after which this same voltage will be provided to any device that’s tapped into the power supply lines of the USB connector. Since USB-C has now also taken over duties like analog audio jacks, this has increased the demand for splitters, but these introduce many risks. Unless you know how these splitters are wired inside, your spiffy smartphone may happily negotiate 20V that will subsequently fry a USB-powered speaker that was charging off the same splitter.
In the video only a resistor and LED were sacrificed to make the point, but in a real life scenario the damage probably would be significantly more expensive.
What an absolutely hilarious problem, truly shotgun to your own foot worthy.
Why not just stick 5K1 on the CC1 and CC2 lines and be done with it? Again, I’m no expert on USB-PD et al, but that seems like the simplest way to get 5V from a host
A few seconds of thinking more and I have the answer. Doh!
You can’t have the CC lines of all ports connected together or someone will negotiate a higher voltage
No, the issue is totally different than what is on the CC lines, although you are correct that there should not be multiple devices on the CC lines.
What these splitters do is simply route VUSB to two devices. There should never be more than one device connected to a single USB port. The correct way to do it would be to design the splitter as a source and sink and to negotiate power for each port separately.
The issue is that one device on the splitter can negotiate a voltage that the other device can’t handle. There are however other problems like the “main” device requesting a defined current, but that doesn’t take into account the current draw of the other device, even if voltage wasn’t an issue.
Simply put, splitting like this is against the USB standard and using this Chinese garbage will always be a gamble.
I hope this could not happen with a proper hub?
In theory a hub should either refuse to use higher voltages at all or have negotiated voltage for every output individually (or possibly some mix of the two where you are either on the 5V only dumb side or on one of the individually controlled ones). I have a few hubs on my desk with 2 or 3 QC3/PD standard supporting ports but the rest are just dumb 5V for instance.
But I’d be willing to bet many hub’s won’t handle all the possible cases they will get used in properly especially if they are the cheap ones. Though even if they were developed with the intention of working properly as its just too stupid of a standard full of weird edge cases that will trip you up… Usually edge cases in which you the user have no control or way to know what will happen in advanced either, so will get whatever you get – like your laptop charging your powerbank instead of the other way around…
I’d never bet against the existence of at least one dangerously maldesigned USB hub(or any other USB device); but I suspect that hubs are substantially protected by the fact that, for the lousy-but-mostly-works implementation you can mostly get away with slight modifications to some decade-old USB3 hub design(if you are feeling really classy and actually want the type-c connector to work in both orientations you can just wire in the data lines from two USB ports, skip the mux chip) by pretending that PD doesn’t exist and just providing enough +5v downstream that plugging in a bus powered portable HDD probably won’t brown anything out and phones and tablets will charge…eventually; while all basic USB functions will be bottlenecked by the fact that there’s only 5Gb/s to the host; but realistically the user buying your trash is probably plugging in a mouse and keyboard that consume basically zero bandwidth, some mobile device that mostly wants power; and a cheap flash drive that can’t manage USB3 speeds anyway.
The trouble for these goofy little splitters is that the minimum viable way is also the fail-deadly way; since, for things like phones, you can’t just act like a confused 15 year old USB3 peripheral and expect good results.
It probably wouldn’t take that much to slap a USB-PD implementation into the splitter that either just tells the upstream device that it’s a host and needs to provide 5v now, if you want to provide the minimum-viable safe experience for downstream devices; or which actually negotiates with downstream devices to determine highest mutually acceptable voltage; if you want to be classy about it; but the BoM and assembly difference between just connectors wired together and ‘includes IC and firmware’ is a bit of a jump. Anything involving DC-DC converters to provide both parties with what they want from a single upstream voltage is, of course, even more BoM, please heat and bulk.
i like your framing, “I’d never bet against …”. because gambling is exactly what we’re doing. and i happily gamble, not against the existence of a bad scenario, but against the possibility that i will run into it.
i’m impressed that henning below exactly highlit a real situation where this sort of question is significant…but in my life, it just isn’t happening…most likely. since i’m already gambling that there isn’t a lurking significant defect, or that i won’t drop it or mangle it or whatever, i feel like it’s not a huge change in the odds over the status quo. i don’t use many devices that use USB-PD or fast charge, coming or going. and i haven’t needed to multiplex my usb-c ports, and so on. it just seems like i’ll probably go decades successfully continuing to treat usb as “if the connector fits, it will work” without ever meeting a downside.
i’m throwing my two cents in because the concerns in this article are valid and useful to know! but “you can use a usb-c cable to destroy a device” strikes me in some sense as FUD. in real life most of us just won’t meet these limitations.
You don’t equal most of people and trying to generalize your personal experience usually doesn’t work.
It’s not FUD to point out that doing things out of spec is dangerous.
Klh – i am not most people but i speculate that most people don’t buy these unusual splitters, because i’ve never seen one in the wild. and with some confidence, i can declare: most people don’t read hackaday. and with the same faith as if it were axiomatic: most people will happily mix-and-match any connector that fits. :)
I think the standard can be confusing but its fine. I think companies tend to do the right thing else obviously they will blow up your device. This isn’t really a corner you can cut. Also, I think this is why you rarely see multi port usb C hubs like you do with USB A. Because it takes a lot more work.
No. Hubs (done correctly) act as a sink and source at the same time. They negotiate the correct voltage and current. Most likely the hub will negotiate the highest possible voltage/current that the power source can provide, and then advertise this voltage/current to the ports (current proportionally).
These Chinese trinkets simply route VUSB to another port, which is against the USB standard. Even back before USB-PD and Type-C, it was never okay to just daisy-chain ports.
This brings up a question, I have had unanswered for a long time.
I have a laptop PC with one USB-C port, that is the only way for incomming power.
The port also has DisplayPort Alternate output, and works with a portable monitor, that has only one port, USB C.
So I have been looking for a device (not a typical dock with seperate DP out), that can take power in, feed it to the PC, and let the PC send video (and power) to the monitor.
But I have found nothing. Is there some logic I am missing or are there something in the spec, that forbids it?
I have seen Dell monitors that provide power & network to a laptop while taking video from the laptop over a single USB-C connection.
If I’m understanding your question correctly they do exist: I have no experience with the one below, this is not a recommendation or an endorsement; but it has USB-C power in, supplies power to the computer it is connected to over USB-C and exposes that computer’s DP alt-mode video on a USB-C connector with DP alt mode (along with an HDMI and VGA port, for versatility).
https://www.startech.com/en-us/universal-laptop-docking-stations/dkm31c3hvcpd
It just seems to be an uncommon niche: portable monitors that accept input power and can pass it through often cost very little more than ones that don’t do that; so people seeking a portable setup often just go with one of those, or have more than one video-capable USB-C port; so the vast majority of the little ‘travel hub’/’travel dock’ things ignore providing USB-C video out, which they presume you already have, in favor of supporting a couple of options that you possibly don’t have but might need; while desktop hubs overwhelmingly seem to feed the video signal to DP or HDMI connectors; presumably because you still pay a premium for stationary USB-C monitors so people prefer docks that don’t require doing that.
I don’t know why it’s quite as obscure as it is; and it sure carries a major premium over the more typical ‘travel dock’ scenario; but (normally) if startech is doing something it’s aimed at business types who care more about ease of procurement than cost; so they don’t tend to do anything too wild in terms of crazy standards violating hacks.
Thank you, this is what I have been searching for (plus a few things not needed right now).
Both monitor and PC are about 5 years old; the monitor is not used often enough that I have looked at something newer. And the low to mid priced laptops I see, almost all have only one USB C port.
I would probably go for the StarTech as well, if concerned for a company that probably does things the right way.
I ended up going for this one.
https://www.amazon.com/dp/B0D8SZ3MC9
This is a common use case in the AR glasses community. I have an Nreal/Xreal Air that need power and DP alt. My Steam Deck only has one USB C port, the dock has no C ports.
none of that could ever happen to me because i have different usage patterns, so i’m just daydreaming here
but does your laptop charge at all if you hook the usb-c port up to a regular 5V-only charger? like with a cheapo wall wart that only does 5V, maybe with an A-to-C cable? that would be the relatively safe way to hedge your bets, if it works at all.
No, it doesn’t.
We had that situation where I work, facilities set up a hot desk area but didn’t understand how USB C works so they bought a bunch of cheap adapters that only supported 5V.
Took the (off site) helldesk a little while to work out why people were reporting faulty laptops
Cheap stuff only offers 5v and often 500mA. Laptop requires a lot more voltage and current, you need PD capable charger.
i’m not surprised it needs a PD-capable charger but my last 3 laptops each use about 2.5W on average during my actual usage patterns. shrug i definitely agree it would not be awesome to operate near that limit all the time but i didn’t know if it would be possible or not.
Most laptops are 12v (3 x lithium battery). Some will charge at 12v, some will operate but not charge.
(That second point is handy, meaning you can run it from a 12v powerbank and once it’s drained switch over to the internal battery.)
Yes, at 5V it will take 0,5 to 1,9A as available. MB TDP is 10W, charging takes some time that way. OEM PSU is marked 5, 9 and 12V – 24W.
The monitor is 5V, so the cheap splitters would work if they passed DisplayPort, which I do not think they do.
But I am happy to have now found docks that will pass DP, so that will be the way to go.
I’ve got a UC-Green “dock” hub which takes in power from my MacBook charger, and provides power to charge my MacBook. The same cable from the MacBook carries video back to the monitor (via the hub). The monitor is mains powered, because it’s a big one, but the hub happily fast charges other devices, so it’s probably work for you.
Another win for USB-C. Barrel connectors in shambles.
Well… spliters are kind of extensions in my opinion, and extensions are not within standard for reasons…
Maybe combining 1000 functions in one connection wasn’t so smart after all :D
-> https://hackaday.com/2022/12/27/all-about-usb-c-illegal-adapters/
I don’t see a problem with this. Splitters are out of spec. This isn’t related to USB C or thr PD protocol. Usb is point to point. If you split it you cant split the data or the negotiation channels.
Also, no a hub is not stupid enough to negotiate with one port and apply that voltage to all ports. I cant imagine anyone would put time into developing a product and have such a terribly out of spec implementation.
This is just a reminder to engineers to think before you act.
indeed.
just because you can violate the spec today doesn’t mean it won’t bite you in the ass tomorrow.
I bet you can find one on timu that’ll do it 😂
I mean you can find SD cards that are fake with just enough an MCU to present the proper size to the device they’re plugged into.
I put I put VCC across all the ports George just like you asked! 😂😂😂
Maybe someone can correct me, but doesn’t a PD downstream facing port still have to supply 5V for signalling purposes even when a higher voltage is negotiated for power? So you could in principle have a hub that passes through PD for one port while keeping the others on 5V, without needing its own DC-DC conversion hardware.
(Assuming the 5V rail has the same current spec as old-school USB, i.e. at least 500mA and usually more)
I thought they always wanted the current limiting outside the device and in charging brick. Because of the heat that can generate
Nope. Proper USB-C ports are off until something is plugged in and negotiates. This is annoying when you plug in something dumb and it doesn’t work.
A lot of chargers ignore this and always provide 5v, I’ve even got a couple of phones that do this as well (one being a Sony!).
My pixel toggles power on and off while it checks for a connection if there is no Rd resistor present. Not sure what this is trying to support. Maybe devices that cant set an Rd without first being powered on. i.e. no dead battery support.
The CC channels are used for the power negotiation. Those operate at a voltage even lower than 5v. Not related to the supplied power voltage. So i guess thats a yes.
I think the issue is that if you say “this is a splitter” it’s obviously dangerous, but if it’s in a nice box you might not realise it’s a splitter internally. I’d certainly assume my hub isn’t a splitter…
The fact that we need an article on this site about that…
You’re telling me my 5 volt device isn’t rated for 20 volts?!
Hold on let me go plug it into my Tesla battery.
Is it the aluminum in the air the fluoride in the water? What’s going on here?
The problem with USB-C is that it oversimplifies things. Nowhere (except on very small print at the back of some devices) is explained what voltages and currents are generated and required.
Almost no cables have any markings whatsoever what they support.
USB-A, Mini, Micro just “worked”, at least after the first years when all kinks where ironed out. Now I need to look very hard and check what are my dollar store cables to charge my keychain flashlight and what are my expensive cables and expensive charger to charge my tablet.
There is no way to explain that to the layman.
The simple standardization of coloured plugs like purple/aqua for the PS/2 mouse and KB and the different audio cable colours did wonders, as did the blue for USB-3.
USB-C should’ve had something similar, a simple set of markings on host/charger, cable ánd device to show what is the minimum and maximum requested and delivered, both in terms of power and protocols.
Indeed USB-C should have a mandatory resistor style colour banding or something so every device made to spec advertises correctly what it is capable of. Still a dumb spec that is going to cause no end of trouble for those average users that expect if it fits its works, know that the port should do x, y and z all at once, but don’t understand the spec at all to troubleshoot why it isn’t actually doing any of those..
Of course that doesn’t get you away from the malicious actor selling 20 pence cables as the premium super pricey ones just by putting the right colours on the connector… But at least the stuff that isn’t actively a scam will be easier to identify.
USBC is a standard. All cables MUST minimally support 20V and 3A without any markings. I don’t see any confusion here. The devices negotiate and agree on the power settings. If you want a higher power transfer then you need to know the capabilities of both devices. USB didnt mandate you print capabilities on your device which I think is fair as that impacts the look of your product.
If that splitter is having same connection on both sides, then it is obvious that the voltage will increase on demand on both side. Such splitters are first of all against USB standards right? A proper type C USB hub should not have this issue I believe.
The latest USB-C charging cables from Amazon, Temu and elsewhere are listed as being USB 4.0 rated, and are marked with “USB 4” on them, with one side also showing the watts being used by the device that cable is plugged into, which is also determined by whether your USB-C PD single or multiport offers 20, 30, 65+, 100+ and 200 or more watts GAN charger, and how many devices you are charging at the same time.
So it is possible to get USB 4.0 cables that deliver the advertised speeds and have a very cool digital display showing watts and more!