- cross-posted to:
- technology@lemmy.world
- cross-posted to:
- technology@lemmy.world
This popped up randomly in my feed today, and I found it to be pretty interesting and informative.
tl;dw: All USB-C cables have a microchip inside them which runs a small bit of software that tells the devices its plugged into exactly what they’re capable of, such as their power rating and transfer speeds. When you plug the cable into your device, it reads the data from this chip, which then dictates how much data/power it is allowed to transmit along the cable.
The problem is that when you use a USB-C extension cable, the device you’re plugging into can only see the chip data from the first cable; the cables beyond that first one are completely invisible to your device. And if your first cable is rated for 200 watts, and your extension is only rated for 100 watts, your device will still send 200 watts down the line, without ever realizing that it’s overloading the extension cable and creating a possible fire hazard.
You must log in or # to comment.
No, all USB-C cables do not have hardware to negotiate power (voltage and current). USB-C is simply a connector type like USB-A and USB-B. What you’re referring to is called USB-PD. This is when the charger and device “handshake” to agree on the optimal voltage and current, then adjust dynamically as needed. Yes extensions can be a problem for things like laptops or anything that can draw large current over high voltage. It’s really not any different than trying to run a table saw over a cheap extension cord meant for a lamp or a set of Christmas lights. If you’re using one of the usb extension cables to charge your phone, it’s fine.
This is also why you should be very careful with what you plug into power strips. Your 15A circuit may be able to supply all the power being drawn from all the things plugged in, but the power strip may not. Do you want an electrical fire? Because that’s how you get an electrical fire.
But also, the 15A breaker is only capable of supplying 12A continuously. The breaker is rated based on heat, (it trips when it gets too hot), and the 15A rating is an “instant” trip rating. Meaning 15A creates enough heat to “instantly” (it takes a few seconds to heat up, but the same is true for the wires in your walls that would cause house fires) trip.
It’s only capable of dissipating enough heat for ~12A of continuous usage, but many devices take more power when they’re first turned on. Like a box fan may only need 1.5A to run, but may take 2A to initially get the blades moving.
All of this is to say that if you truly need 15A of continuous usage, install a 20A breaker (and wiring, and outlets) instead. Because a 20A breaker is capable of handling 15A continuously.
I’m sure they exist but I’ve never seen one that didn’t have a breaker for this very reason.
Weird, I extended lightning cables fine.
But lightning isn’t usb-c.
New mission to doom the world is to give away usb-c cables that will activate order 66 after random time.
Why can’t the extenders pass along the same data the original cord is sending?
They do (if they carry all wires), it’s just that they don’t add their own current rating into account.
So then do that
I guess there is a way to implement this true to USB-C standard, either with a simple comparison of reported capabilities or a 1-port hub. Somebody would have to make a cost-effective chip that does that though, and manufacturers are happy just printing “max 3 A” on the packaging (if even that).
The USB consortium has responded to this sort of need over time, but they have to do it in a way that is completely backwards compatible. It might be possible to do this in a new release of the standard, but it may also be that millions of cheap cables would be rendered useless in the process.
Non-compliant does not mean useless. They will continue to work with devices that use the old standard. PD is already a mess, and especily devices without displays will often fail silently without indication that it’s the wrong cable or adapter.
Are they actually “not allowed” because the framework laptops all have these but they’re inside a box and not a cable.
3rd party, universal cable != Circuit boards/connectors designed for very specific hardware (used internally)
They are if you have a retimer/redriver. Most extension cables do not.
I can’t find a picture of the inside of the module, but I guarantee you it’s not just a straight through cable. It’s got active circuitry in there.
The USB C module does not have any active circuitry and is essentially a short extension cable
The USB A module has a retimer/driver but they tend to have issues with AMD
I’ve looked inside there’s like a dozen things soldered to the PCB in there.
Looks pretty passive to me
https://frame.work/products/usb-c-expansion-card?v=FRACCQ0004
On Framework Laptop 13 Intel platforms, that means Thunderbolt 4/USB4
Framework Laptop 16 additionally supports up to 48V/5A charging
Impressive. It is short enough that I’m not surprised it works. I have a 4 inch passive cable that successfully does TB4/PD100. The signal integrity is pretty butts though and high interference will cause problems. I’m not ballsey enough to try anything higher. Looks like there’s maybe some passives on the other side of the one they currently sell, but that doesn’t count.
Why is it designed like that? Why cant the devices on either side talk to each other about power requirements etc instead of going through an extra chip?
Because the cable itself is a factor. Just because the supplier says “I can give you 200 watts”, and the secondary device says “I can take 200 watts” doesn’t mean the cable between is able to carry that.
It’s not even just power but other things like DisplayPort over USB, Ethernet over USB, Thunderbolt, and whole slew of other things a theoretical USB-C cable can do, but not necessarily every USB-C cable.
It’s important to note that not all cables connect all the wires. You don’t need them all if your cable is only for charging. You need more if your cable transfers data.
