USB-C: the definitive thread

wco81 said:

That's where I found myself. Have a bunch of name brand cables such as Anker cables rated for 65 or 100W PD.

But they all are only good for 480 Mbps speeds. Ridiculous.

I got two short cables with two Samsung SSDs. When I use them on my M1 Pro MBP, the speed in noticeably better. Replace one of them with the Anker and it chokes, takes minutes for 5-10 GB of data transfers, which is not even a full day of shooting RAW files.

It seems most of them are being marketed now, at least on Amazon, for charging specs, not data transfer speeds.

Click to expand...

This is exactly the anticipated behavior, so not exactly ridiculous. As mentioned many times, the USB-C specification allows for two types of cables in terms of wires present: charging and full-function. All cables are required to have the USB2 wires, but only full-function cables are required to have the USB3 wires.

Given that so many devices with USB-C plugs use the connection only for charging/power (whether that means it doesn't transfer data, or people almost never transfer data using the connection, e.g. smartphones), it isn't unreasonable to have a cable type that can support that usage properly without requiring the extra cost of those wires that are never used. The only issue is that cables aren't as clearly marked as they could be - aesthetics (e.g. cloth-wrapped cables) take precedence over clarity unfortunately.

Before USB-C -- and before HDMI 2.1 -- a plug delivered all the capabilities.

There wasn't this splitting of charging cables and data transfer cables.

Most people aren't paying attention, they see a plug and know that it charges and transfers data.

But then splitting up into different data transmission speeds and different charging capabilities, they must have allowed that to allow some cheap products to flood the market.

Again, this clashes vs. consumer expectations and it's new with this plug.

So there are many people who are going to find it ridiculous.

OK the other issue with these small hubs, besides making sure they have the right data transfer speeds, you have to see whether they can power the right combination of storage devices.

Some of them will let you plug in say two SSDs but not one HDD and some other storage device (well maybe flash keys).

Some of them say all you have to do is plug in a power source to the USB-C input port, which is often a PD pass through, so that you can charge your laptop through the hub, which is taking up one of the USB-C ports on the laptop.

Unfortunately, these hubs which are sized to be convenient for travel, are often going to be used with portable SSDs or HDDs, which do not have their own power supplies.

So it sounds like you can't reliably connect more than one storage device into these hubs at a time. Or ideally, avoid plugging any into it, just use it for peripherals and things like memory card readers but plug SSDs and HDDs directly into the laptop.

Which could work, since even my 14-inch MBP has 3 USB-C ports. However, that could mean running it on battery while you have two storage devices, one of which could be an HDD, plugged into the MBP running on battery.

May be fine for short stints, like under an hour or under half an hour.

Yes I could use the MagSafe but it's easier to pack a USB-C cable than a MagSafe cable, trying to minimize weight and bulk in your travel kit.

wco81 said:

So do I need to buy a more expensive Thunderbolt hub? Problem isn't just the cost but the bulk and weight isn't great for travel.

Click to expand...

The problem with high speed hubs is that high speed interfaces burn power, more than a 5V 3A standard port can supply.

For travel I would get something like a https://www.anker.com/products/a8356?variant=42474958028950

It has PD passthrough so modern laptops that charge on USB-C can use a single port for data and charging. And that also powers the hub, and includes HDMI out if you want to watch Netflix/whatever on a bigger (hotel) TV.

5 gbps is still 500+ MB/s so transferring a few GB wouldn't take forever.

wco81 said:

Before USB-C -- and before HDMI 2.1 -- a plug delivered all the capabilities.

Click to expand...

False. A few counter-examples:

USB3 with an A connector: same exact shape and backwards compatible with USB2 A connectors, but had an extra 5 pins buried in the back of the connector which you'd probably never notice if you weren't looking for them.

DVI: DVI-D vs DVI-A vs DVI-I. Single link vs dual link for -D and -I versions. Presence of pins didn't necessarily mean they were actually connected.

HDMI: Ethernet was added in 1.4; before then the pin was "reserved" and wouldn't need to be connected. Further, HDMI has increased the signalling rate three times, not just once, so even back when 1.3 was the best available, you could have an older cable that only worked properly with 1.0-1.2 signals.

VGA: pins 4 and 11 used to do nothing, then they were part of DDC, then they were reverted back to unused once E-DDC came out and used an I²C signal over 2 instead of 4 pins.

Ethernet: 10BaseT and 100BaseTX used only two pairs, while everything since 1000BaseT (excepting specific stuff like SPE) has used 4 pairs (or optical once you get fast or long enough). All of these use the standard 8p8c connector.

SCART: The original spec defined five different cables using the exact same connector, with which pins were wired through varying between them.

Is USB-C worse than prior examples? Yes, but that's because there are so may different combinations of port capability, not so much that the number of types of cables have increased.

I have less issues with HDMI and DP as cable spec is usually printed on the side repeatedly. USB 3 is pretty easy to spot too with its blue connectors.

USB C is just a mess, if you're lucky you see a SS5/10/20 or mark. If not, just assume slow data and no e-marker for PD.

Economic incentives led to this situation. USB organization wanted there to be plenty of cheap cables and accessories so they let anything go.

I presume that they offer a certification program but it's optional and is so costly that few companies participate.

Then you have even name brands like Anker playing the same game, mostly offering different PD cables but they're all limited to USB 2 data transfers.

wco81 said:

Economic incentives led to this situation. USB organization wanted there to be plenty of cheap cables and accessories so they let anything go.

Click to expand...

Based on what @Semi On has said in the past, it's more that the USBIF is pretty dysfunctional. Too many different companies wanting too many different conflicting things, so you get standards like "USB-C" that are more of a chinese menu than a single specification.

wco81 said:

I presume that they offer a certification program but it's optional and is so costly that few companies participate.

Click to expand...

They do, but they don't really have any way to police it. They don't even own the copyrights to USB or USB-PD. If I recall correctly, USB-C(R) was the first term they registered. They do have some of the logos trademarked, but they don't have the resources to search out offenders and request pull-downs, which is why Benson took it upon himself in the beginning.

cogwheel said:

Based on what @Semi On has said in the past, it's more that the USBIF is pretty dysfunctional. Too many different companies wanting too many different conflicting things, so you get standards like "USB-C" that are more of a chinese menu than a single specification.

Click to expand...

The good news is that the USB4 and PD3.2 specs remove a lot of that variability.

iljitsch said:

Well then, only buy cables from vendors that properly mark them.

I agree that the annoyance exists, but it's really a failure of commerce, not of technology. Makers and sellers of affordable cables apparently see no upside to clearly listing the limitations of their cables, and we are too cheap to buy the much more expensive cables that do list their capabilities fully (at least on the packaging and in the online description).

What would also help tremendously is for our computers to tell us the capabilities of a cable.

Click to expand...

It wouldn't work for power, but a bandwidth loopback test should be pretty simple to code for any device with two USB-C or Thunderbolt ports.

I'll add it to my list of "projects to do someday (or never)".

iljitsch said:

<snip>
Looping back USB-C to another port on the same machine is of course great because then it will charge itself.

Click to expand...

Sheer bloody brilliance!

Why didn't I think of that before?