Usb C cable 100W rating.

[CambridgeWorks]

Seen advertised a usb c charging cable rated at 100W.
How is that achieved?
Rob

[ppppenguin]

USB power delivery is outlined here: https://en.wikipedia.org/wiki/USB_hardwa...r_Delivery

The current is kept down by both ends negotiating a higher voltage. Power over Ethernet (PoE) does a similar negotiation. Let's hope the methods are robust so you don't get anything up to 48V shoved up a poor unsuspecting device that only wants 5V.

[CambridgeWorks]

Thanks for that. I was thinking it may have been part sales gimmick. Although, I wouldn't fancy powering say a 12Vi tem at 5A through such a lead and connectors.
Learn something every day.
Rob

[ppppenguin]

If the lead and connectors aren't up to scratch things could get very warm! Fakery is everywhere so where it matters I would buy a good brand such as Anker.

[DrStrangelove]

Oh.

I thought it was A Kind of Magic.

Changing the laws of physics, Captain.

[Mike Watterson]

The USB charger can do 9, 12, 15 or 20V. The default without signalling from the gadget is 5V. It need not be USB-C on the charger, some use USB-A.

Also some USB C charging schemes need USB 3.x wiring, which is FIVE EXTRA connections on a USB-A (at front of socket or back of plug). But some 12V charging via USB can use just 4 wires. Many older PC USB 3.0 ports, (USB-A or USB-C) only supply 5V.a

Many USB-C to USB-C cables are cheap USB 2.x and only have 4 wires.

It gets worse.

There at least two main specs and versions of then as to how to have more than 5V. Even regular USB has different passive signalling methods to indicate the charger can supply more than USB 1.x currents, such as 500mA 1A, 1.5A and 2A. Usually resistors or voltages on the D- and D+ pin.

In contrast USB 3.0 (which can use other than USB-C) isn't just faster data, but a separate serial port to negotiate charging,

It gets worse.
USB C audio can be simply USB digital data on D- and D+ (Needing a DAC and ADC), or some wires can be analogue audio to directly drive earphones and possibly allow a mic. But the analog feature is optional. You can't have both at the same time.

It gets worse.
USB C video can be simply USB digital data to a USB video adaptor. Or is can be USB DP, a kind of HDMI/DP digital video that uses the extra serial wires. It's very optional, and I've only one USB device that does it. Then a monitor can use USB-C for power and digital video.

It gets worse
Some gadgets even with micro or mini USB could be a host using a special passive adaptor, so a phone or tablet could use a keyboard or USB memory stick. Very few could then charge at same time, instead providing power. USB-C also has USB2go etc, and has it more often. Also more often, but not always an adaptor cable can also charge the device. Some PC USB-C (few) allow select of client or host mode, and in theory some phones/tablets can let one end be host and the other be client.

USB has had networking via USB data for maybe decades. Few gadgets support it. Some like reMarkable eink PDF reader only do networking on USB. There might even be ethernet on a USB-C cable now.

USB 1.0 was a missed opportunity. Subsequent versions are faster, handle more power and even non-USB signals. It's now a mess. It's opaque when reviewing advert to see exactly what functions are supported.

The cables are a mess too. A USB 3.0 cable (C to C, A to C, A to B are all supported) needs extra wires and the USB A port should be blue with 5 extra connections. Most USB-C to C or A to C cables sold seem to be only USB1.x/USB2.x.

[Nick]

With USB C (the connector) & 3.2 (the protocol), you can get severe signal degradation over more than a couple of metres, even with decent cables. Many cables only have a selected number of the full range of connections available - They may not support all the Power Delivery (PD) options that the devices you plug into use - there are at least 3 different "standard" PD specs, (1.x, 2.x & 3.x) but the standards also allow some manufacturers to have custom PD signalling - e.g. the Raspberry Pi 5 is non-standard, using 5V at 5A, where the normal standard only allows 5A once you're above 20V and 3A below that, but the RPi lot don't want a high current DC-DC buck converter on the board, so needed 5V @ 5A to drive the board and to supply power to all its own USB A ports... However,  the USB PD specification R3.1 v1.8, Table 10-2 footnote 2 says "The Fixed PDOs Maximum Current field Shall advertise at least 3A, but May advertise up to RoundUp (PDP/Voltage) to the nearest 10mA. Requires a 5A cable if over 3A is advertised." R3.2 V1.0 came out in April this year but I don't have a copy to hand.

Also, USB high power charging cables generally need to be pretty short to prevent significant power loss. There are many many cables in the market with bonkers claimed specs just trying to outdo the next guy - the market is riddled with poor quality cables that look pretty but get nowhere near their claimed specs.

Over about two meters, for signalling cables, you need to switch from normal "passive" to "active" cables.

Proper active cables (from folk like Lindy) do signal shaping and have inbuilt differential line drivers etc. I have full speed (10Gb/s) USB 3.2 Gen 2 5mtr USB C active cables that work flawlessly into proper powered USB 3.2 Gen 2 hubs - passive cables had dropouts all the time.

You pay significantly for that privilege, but it works perfectly. The main cables I use are USB 3.2 Gen 2, C to C, https://www.lindy.co.uk/cables-adapters-...ble-p13340

Warning: there are a lot of fake or below spec USB C "active" cables. Buying from AliExpress, eBay or Amazon? Good luck with that! Real ones will be really quite expensive as there's a lot going on inside them. I chose Lindy as they are a highly reputable company plus they design & make their own active cables and test each one before delivery - I got to speak directly with the engineers in the UK & Germany.

[ppppenguin]

I didn't realise it was THAT bad.

We love standards, let's have plenty of them

[DrStrangelove]

Aha! So to sum up: USB C is a horse designed by a committee on mind altering drugs.

Glad we've cleared that up.

It explains a lot.

[Nick]

Aha! So to sum up: USB C is a horse designed by a committee on mind altering drugs.

Glad we've cleared that up.

It explains a lot.

USB C (or "USB A", "USB B" etc.) is a connector specification, not a protocol. USB 3.2 Gen 2x2 is the latest signalling specification (20Gb/s), however the most common current fast specification in use is USB 3.2 Gen 2 (aka "Gen 1x2) which runs at a max speed of 10Gb/s.

Just to confuse the matter further, there's also USB 3.2 Gen 2x1, previously known as USB 3.1 Gen 2, which also runs at 10GB/s but uses a different signalling protocol.