Hacker News Reader

Comment by notatoad | Hacker News Reader

This is really only a problem if you buy cheap cables. Pick the baseline spec you want your USB cables to meet, and throw out all the ones that don’t. Problem solved.

It’s still a minor frustration if you have to borrow a cable, but that’s solved by packing a cable, and by encouraging as many other people as possible to also throw out their cheap cables.

View on HN

Sometimes it's not the cable. Sometimes it's the sodding device. I have some IP60-something speakers for the shower. Write up says USB-C power. Brilliant says I, I've got plenty of good ones I use for charging laptops and other USB-C form factor accepting devices.

Sodding things only charge with their special USB-A to USB-C cable. They're in the bag labeled "cursed usb-c charge cables".

you'll never guess what solution i'd suggest for that specific problem...

Change the device? It works fine. The cable is labeled. I know where to find it. I'm not throwing away perfectly otherwise-good hardware because of that.

I have a water pump like that, the reason: they have omitted the 5.1k resistors…

It's because they are straight up defective. The USB-C spec is pretty clear in that the power pins have 0v until you signal for a voltage. This isn't a failure of USB-C, you should just return defective devices to the OEM.

Thankfully this defect is becoming less common outside of temu junk.

That's because they didn't implement USB C correctly. Maybe it was to save less than a penny on the BOM, maybe it was arrogance ("it works for me! send it!"), or maybe it was something else. Whatever the reason, they didn't do it right.

By standard, USB C provides no power at all unless the device being powered follows the rules. They're easy rules to follow. The minimum viable way of doing it right requires just two tiny resistors inside of the device.

I do electronics professionally. This is likely because in order to save a bit on BOM costs, the device itself doesn't have the necessary pulldown resistors to signal to turn on VBUS.

USB-A has always-on VBUS while USB-C doesn't. Because the spec allows for always-on VBUS in a USB-A to USB-C cable, some devices just assume that they're always being powered by one of those cables.

I doubt it’s even saving cost, it’s just incompetence. It works with the USB A to C cable, so job done. They’ve probably never plugged it into a PD charger, and probably just copied the schematic from the last (defective) product and will copy and paste it into the next one too.

That is likely the case; cost savings, simplicity, etc. but the end result is USB-C does not "just work" for most consumers.

It is not ideal, but you can buy C-to-C adapters with the resistors to force the charger to supply 5V.

https://www.aliexpress.com/item/1005012247371267.html

If you want to cover all use cases your only option are the 240 W, 40 Gbps cables, and they are not cheap (https://www.amazon.com/Anker-Supports-Display-Transfer-Charg...).

amazon hides some prices because i'm viewing the page from canada, but it looks like that says the cable you linked is $15?

that's what i've been paying for cables, and it seems reasonable to me.

It's a cable, it should cost $2 at most.

okay, i mean, that's exactly what my original comment said: this is only a problem if you buy cheap cables.

If some of your devices require a $30 cable and you use the same cables for everything, then you use a $30 cable for your mouse.

If you're buying a cable for a specific device in a specific circumstance than by all means get the cable that meets that. Then it's not a problem about "what is its functional support", it doesn't matter, that's your mouse cable that you use for your mouse on that desk. That's the monitor cable you use to plug things into that monitor. That's the dock cable you use to plug things into the dock.

For all the cables that are your flex carrying around places where you don't know how it'll be used tomorrow, get a good cable and you won't have any issues.

Yeah? I use a $15 cable for my mouse. And then when my mouse is done charging I put the cable back in the bag with my portable monitor, which the same cable also works for.

It's also only a problem if they overcomplicate power delivery design such that you need expensive cables.

They could have made it just DC +19V and GND over 14 gauge wire with a nice, outdoor-recreation-grade connector and called it a day.

USB power delivery doesn't need expensive cables. Basic 240W cables cost $3-4.

I have some Thunderbolt cables that are like $80 apiece because they're like 3 meters long. I have like five or six of them because each one only carries one video signal.

(I used Thunderbolt so that touch and pen inputs could go over the one cable. These days though I just use a MacBook Pro, and hook up a capture card when I need to access the Windows)

The apple usb c charging cable is like $30, it is well worth the money of you are into software defined radio I can report. Not sure what magic they pulled of, but all the tricks in the book it seems!

40gbps is an insane data rate that requires shielding, precision, and testing. There is no way you are getting that for $2. That said, outside of connecting a laptop to your monitor, you don't need a 40gbps cable.

> 40gbps is an insane data rate that requires shielding, precision, and testing. There is no way you are getting that for $2.

I dunno, Cat8 is 40gbps and pretty cheap. DOCSIS 4.0 does 10Gbps on some really ancient cables. I'm sure Cat9 will do even better, at least the cat people and the doc people are trying harder than the USB people.

Unifi even has a PoE-over-coax solution. Add some Anker GaN shit and I'm sure it can be miniaturized.

> outside of connecting a laptop to your monitor

As soon as I need a "special" cable to connect a laptop to a monitor, we're effectively back to the 1990s-2000s when I needed a special monitor cable. There is no point to connector standardization if any cable can't do any function.

The whole point of this thread was "one type of cable for everything" and "grab any USB-C cable from your personal stash and they all work for every use case".

> There is no point to connector standardization if any cable can't do any function.

I completely disagree.

When the only difference between cables is max speed, that's still a huge improvement over a nest of different cable types, half of which are custom. And it's easy to get into a position where all your USB-C cables differ only by max speed.

Which is worse because they all look sort of the same, but they are not the same.

It's even worse for non-techies, who don't understand what a gbps or a watt is, and who will leave a 1-star review, or worse, trash their cable, because their cable was "slow" but was meant for 240W PD but only supported USB 2.0. They purchased it initially because it had 5 stars.

Ideally, there should only be "5 star" USB cables, and they should all work for all purposes that they can physically plug into.

The situation in the early 2000s is I could spot the cable I needed from a mile away.

USB 2.0 cables can be thinner, more flexible, longer, and cheaper. You could decide to only ever use thunderbolt cables for everything, but you'd be paying $40-$100 for a cable as thick as a mains power cord and doesn't offer any benefit at all to charging your phone.

We need better labeling rather than making every single USB cable a 40gbit thunderbolt cable.

Just about every data cable standard has different speed capabilities on a per-cable basis. Ethernet, HDMI, and displayport have the same issue and usually with even worse labeling.

> Ideally, there should only be "5 star" USB cables, and they should all work for all purposes that they can physically plug into.

That requires they never increase the speed again. Seems like a bad starting point.

And most people will accept the tradeoff of different cable speeds when it means long charging cables can be 10x cheaper. The problem is when manufacturers choose to use bad labels.

And if you push people off USB they're going to use 27 different kinds of cable and you won't be able to spot which generic black wire you need from across the room.

Also, if we pretend they set the max speed in stone in 2013, it's easy to get yourself a full set of 240W 10Gbps-per-lane do-everything cables in every length from 0 to 2.5 meters. $25 isn't an amazing price but sticking with passive cables keeps you out of the real nasty prices.

99% of user use cases for USB are charging batteries and copying a word document on a flash drive. For those cases literally every port and cable works. For video, in every office and home, the thunderbolt cable just sits in the monitor at all times with just the USB end free. Meaning there is no confusion around which cable is the video one because the video one only ever sits connected to the monitor.

While I now no longer have to carry around a bag of chargers when I travel or fish through a bucket of black DC adapters for the right one.

There's tons of confusion when I have new usb-c devices like a portable usb-c monitor or a VR headset or some other gadget. I have cables that seem beefy because they do power, but they don't do data, and I end up having to go through a pile of USB-c cables to find one that does data. It's frustrating!

You can trade-off cable shielding for electronics power and corresponding heat to encode/decode stuff.

Ethernet is notoriously power hungry, but I bet you can use a phone cable and still get tens of Gb over a very short range.

> Cat8 is 40gbps and pretty cheap

Is it really though? People are selling cables, but 40Gbase-T hardware does not exist, and 10Gbase-T hardware is already crazy power hungry and the SFP+s heat up much you burn yourself if you touch them. Meanwhile USB-C 10 Gbps works on an iPhone.

Meanwhile, Coax is huge and stiff. Good high-speed USB-C cables actually have mini coax inside.

> throw out all the ones that don’t

> encouraging as many other people as possible to also throw out their cheap cables

One of the main advantages of a single standardized plug is reducing e-waste. This just sounds irresponsible. Having a single tool that covers every possible use case is rarely a good solution.

Device makers need to stop including these garbage 4 cm A to C cables. I have enough USB cables and bricks to last my whole life. I don't need to collect a million more spec violating ones in a drawer I will never use.

Things are shifting though. Ikea ships their USB-C stuff without cables now.

> This just sounds irresponsible.

The irresponsibility lies with the companies making the non-compliant trash in the first place.

I have two different fast charging cables, for two different devices that charge fast with those cables, on those cables, but not if you swap them around. Thankfully they have different colors, so there's some hope of keeping track of which cable to use for which device.

There's no way to know if it's a "cheap cable", not even by price.

Sure there is. Buy reputable brands from reputable marketplaces that clearly state the specs, and if you get products from them that don't comply stop buying from them.

I'd call that trial and error.

I've been doing this strategy for years and never had an issue of buying a junk cable.

Reply on news.ycombinator.com