News for hackers. Stuff that matters.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by whatever1 on Thursday September 01, 2022 @08:52PM (#32685330)
I don’t care about the name. Can you please write the version on the cable? I have a dozen of USBc cables in my drawer and I have no idea what each one is capable of.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by derefr on Thursday September 01, 2022 @09:12PM (#32685455)
1. It's either a regular USB-C cable or a Thunderbolt (USB4) cable. If it's a Thunderbolt cable, it'll have a large, long, square strain-relief section behind the connector (there are active logic components inside this part.) Otherwise it won't.

2. The charging speed is fundamentally constrained by how thick the cable is. So thick cables are very likely fast-charge cables (they wouldn't spend the extra money on lower-gauge copper if they weren't going to need it for something), and thin cables definitely aren't.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by gregmac on Thursday September 01, 2022 @10:02PM (#32685757)
Just a caution: I've cut apart several microusb cables (using the ends for power, and cables for sensors, etc, just because I have a bin full of them - mostly for esp32-based projects). The thing I've noticed is cable thickness doesn't line up with conductors thickness. I've opened "large" cables to find tiny 28awg wire inside and a very thick jacket. Maybe just my cables and limited sample size, but it's like 50% of the time. A lot of cables in my bin are the random ones that get included in the box with something else.

I so far haven't chopped up any usb-c cables, but I don't see why to expect any different.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by derefr on Friday September 02, 2022 @12:59AM (#32686772)
> I've opened "large" cables to find tiny 28awg wire inside and a very thick jacket.

In a way, this is actually fine too — the conductor itself can be higher-resistance, as long as there's enough shielding to prevent the heat the conductor puts out from setting anything on fire. (Basically, following the same rules for sheath thickness as in-wall cabling uses.)

The only real practical difference, for most people, will be that anything electrically connected to the "thick sheath, thin conductor" cable, will get very hot. (And also, if you use such a cable to connect a device to a portable battery bank, you'll find you get your battery bankk will run dry sooner than expected, since your cable is acting as an additional load.)

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by wiremonger on Friday September 02, 2022 @01:16AM (#32686856)
Small tidbit: for cables that are permanently installed in buildings, the insulation and jacket material is more important than the thickness. Typically these installations are inspected by the local jurisdiction and the electrical code is very specific about what gauge conductors have to be used in which circumstances, so the inspection will generally prevent anything egregious enough to cause spontaneous combustion due to heat from electrical resistance.

The bigger issue is that the insulation material will degrade over time if it's consistently exposed to temperatures that are too high, and at that point it becomes a safety hazard because the conductors are no longer properly protected. So typically cables with a higher temperature rating will have insulation and/or an outer jacket made with different materials.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by zurn on Friday September 02, 2022 @03:49AM (#32687878)
Doesn't the shielding help prevent heat output only with short duration power use peaks? If we imagine the conductor is heating at the same rate 24/7, all of the heat must flow out of the cable at the same rate it's getting generated unless the conductor's temperature can increase indefinitely.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by derefr on Friday September 02, 2022 @11:43AM (#32692110)
If the insulator material is truly a (thermal) insulator, and thick enough, then the heat will "prefer" to flow out the ends of the cable rather than making its way through the skin of the insulator. (Remember, electrical conductors are usually also great thermal conductors; an insulated copper cable is essentially a solid-state heat pipe.) Pick up one of these thick-insulation cables, even after hours of operation — the cable itself won't be hot. But the ends sure will be!

For mains leads, this usually means you're "grounding" heat into:

1. the device itself, usually at the power supply. The heat dead-ends between the coils of a transformer, this being why high-wattage PSUs have fans.

2. your wall — specifically, your wall junction-box power sockets, at the point that the device's power lead's male connector touches the faceplate of the wall socket. This is why you'll often see slight electrical singing on such sockets when you're operating e.g. a portable air conditioner or microwave on a 120V 16A socket: despite most of the heat following the thermal conductive path of the cable into your house wiring, some of the heat ends up transferring into the air where the thicker insulator ends, through the bare contacts, and into the plastic of the wall-socket faceplate.

3. The rest of the house, through the thermal conductive path of the house wiring — the house wiring, being lower-gauge and so needing relatively thinner insulation, leaks heat more readily than the sort of "super-insulated" cables made for high-amperage appliances, and so the heat is gradually sloughed off all over your walls.

4. Your neighbourhood electrical distribution box. Despite heat leaking through your house, modern residential wiring's electrical insulation is still thick enough, that if you're using enough household load to just be barely below tripping your mains breaker, you're probably heating up your neighbourhood distribution box. Ever noticed one humming really loudly when you walk by? Just like your computer's PSU, these have active cooling that goes harder when houses are dumping heat at them.

Note also that this is among the reasons that industrial power routing (for e.g. factories, mills, etc) doesn't use insulated NM in-wall wiring (Romex), but instead routes individual [lightly-insulated, essentially just for electrical separation] conductors through metal conduit, with as little in-wall routing of that metal conduit as possible. The metal conduit is a heat sink for the conductors — such installations are trying to take on the responsibility of dumping electrical waste heat locally, so that they don't overheat their nearby electrical distribution station, or produce line sag on the poles connecting them to it!

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by CWuestefeld on Thursday September 01, 2022 @09:23PM (#32685529)
There are lots of charge-only USB-C cables, without the data lines connected.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by Dylan16807 on Thursday September 01, 2022 @09:46PM (#32685660)
Anyone making those doesn't care about the spec. The single slow data pair is really easy to put in.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by Gigachad on Friday September 02, 2022 @12:17AM (#32686529)
Those are usually recognizable because they are about 10cm long.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by Dylan16807 on Thursday September 01, 2022 @09:46PM (#32685664)
I don't think that works. Plenty of thunderbolt cables are passive.

Also there are cables that bulk up on insulation to pretend to be more premium.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by stephen_g on Thursday September 01, 2022 @11:05PM (#32686130)
I think all Thunderbolt cables are actually active, it’s just the chips at each end are pretty tiny so for a lot of cables you wouldn’t know it’s hiding inside the connector.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by Dylan16807 on Friday September 02, 2022 @12:02AM (#32686456)
They require a marker chip, but that's pretty different from being an active cable. Pennies vs. dollars.

Passive ones are available in multiple lengths up to 2 meters, but only go half speed above .5 meters.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by bombcar on Thursday September 01, 2022 @09:28PM (#32685556)
I’d be interested in a cable tester that could determine what’s what - probably couldn’t figure out max power however.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by kimburgess on Thursday September 01, 2022 @10:47PM (#32686013)
Yes it could, once.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by bloopernova on Friday September 02, 2022 @07:26AM (#32689347)
I have one of these: https://www.kickstarter.com/projects/electr/c2c-caberqu-usb-...

It works great, although I'd like a microcontroller added that can display the standard rather than looking up the connections myself.

Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by isatty on Friday September 02, 2022 @01:03AM (#32686792)
It would be very very expensive. Even proper Ethernet testers (not just continuity and pin tests) are expensive and it’s been around forever.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by capableweb on Friday September 02, 2022 @07:04AM (#32689202)
Searching around, I can find USB testers (some with 3-1 functionality, including USB-C + PD) in the range of $20 - $100 on Amazon and bunch of other websites, that doesn't seem "very very expensive". Would anything be wrong with such devices?
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by bombcar on Friday September 02, 2022 @08:55AM (#32690102)
Many of those will just check continuity on pins (which might be enough for what you want) - a full on tester will run actual data at faster and faster speeds until it fails; report error rates, etc.
Re:USB 4 version 2 promises 80 Gbps speeds over existing cables by ender341341 on Friday September 02, 2022 @02:05AM (#32687146)
I've taken to using a label maker to version my cables cause I got sick of throwing away HDMI/DisplayPort cables when I couldn't determine what they actually supported.