Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I stopped trusting USB-C cable labels and started testing them

makeuseof.com

Post reply on HN
Print view View original

221–230 of 266 posts

Re: I stopped trusting USB-C cable labels and started testing them

#221

Post by KronisLV »

brookst wrote: post #182
dmos62 wrote: post #144

Earlier quoted context omitted.

Ehem, we already do that. USB-C is an umbrella term and everyone has a homebaked organizing system to tell which USB-C cable can do what, or a homebaked testing routine where we run through 5 cables each time we need to connect something. That's what the article is about. Calling all these cables USB-C and making them look indistinguishable is a very leaky abstraction.

So there are 4 different current-carrying levels, from 2.5A to 5A EPR. There are 6 different transfer speeds, from 480mbps to 120gbps. There’s DisplayPort yes/no. There are 4 thunderbolt genrations, each with its own marker. Some of those are dependent (480mbps cables don’t have markers), so it’s not a full permutation problem. Only about 20 different combinations are valid. Is the proposal to have 20 different physi…

USB3-5A-10Gbps-DP-TB3-C would be pretty gnarly as a label even if it’d still be a Type C at the end of the day.

Re: I stopped trusting USB-C cable labels and started testing them

#222

Post by Kirby64 »

eek2121 wrote: post #219
Kirby64 wrote: post #72

Earlier quoted context omitted.

Unless you want all cables to be thick, $60, and cap out at 1 meter long, this is not a reasonable standard. 480mbps USB-C is perfectly adequate for many applications. Likewise for 60W cables.

It’s not hard to make a functional standard, the USB folks just don’t care, IMO. If they had simply worked on better naming, things would be much more clear. (now they’d need to do something crazy like USB-C 40G PD60 as a name in order to fix things, unless the next version of USB is a clean break) Now they’ve made a mess of things, and cables AND connectors are all over the place.

How would better naming make things more clear when manufacturers don’t want to put labeling on cables, generally? Even if it was a USB-IF requirement, that doesn’t have any teeth. Ports tend to be decently well labeled, but you can only cram so much into these things.

Re: I stopped trusting USB-C cable labels and started testing them

#223

Post by Kirby64 »

itsamario wrote: post #220
Kirby64 wrote: post #72

Earlier quoted context omitted.

Unless you want all cables to be thick, $60, and cap out at 1 meter long, this is not a reasonable standard. 480mbps USB-C is perfectly adequate for many applications. Likewise for 60W cables.

Now that EU made USBC a power standard I expect the 4k/8k/16k, 3.0/3.1/3.2, and 5/10/20gbps to be replaced with 60/120/240watts. This way it's simpler to enforce a standard across the industries

Why would speed capability be replaced with power capability? They are orthogonal capabilities intentionally so you don’t have to overbuild cables in either direction if you don’t intend to.

Re: I stopped trusting USB-C cable labels and started testing them

#224

Post by advisedwang »

ryandrake wrote: post #73
Kirby64 wrote: post #72

Earlier quoted context omitted.

Unless you want all cables to be thick, $60, and cap out at 1 meter long, this is not a reasonable standard. 480mbps USB-C is perfectly adequate for many applications. Likewise for 60W cables.

So, call it USB-D or something, if one cable is significantly different (in purpose, performance, form factor) than another cable. Don't just call everything the same label, and let the non-technical buyer beware.

Right... label them differently. Now you're just arguing what the label should be

Re: I stopped trusting USB-C cable labels and started testing them

#225

Post by rtkwe »

josephg wrote: post #217
phil21 wrote: post #209

Earlier quoted context omitted.

I mean, if you care about this and it annoys you the fix is very simple and relatively cheap. Just set a minimum baseline/standardization for the type of cable/capabilities you need for your highest end device. Only keep those cables, give away the rest, and buy a half dozen new cables so no matter which you select you will always get the right one. I've rarely found I need a "special" USB-C cable outside of connecti…

> Just set a minimum baseline/standardization for the type of cable/capabilities you need for your highest end device. Only keep those cables, give away the rest I don't have a cable tester, and I don't want to buy one. The companies who make USB cables should be responsible for this testing. I don't want to do expensive homework on behalf of USB cable companies.

Buy nice cables from a trusted brand and bin anything else. You're going to be overpaying though because there's not a lot of reason to be powering your laptop with a 480 Gbps TB 4 USB-C cable.

Re: I stopped trusting USB-C cable labels and started testing them

#226

Post by rtkwe »

roaringrocky wrote: post #145
ChrisMarshallNY wrote: post #80

Earlier quoted context omitted.

Think about it. You need to generate a precise, calibrated, 40GHz signal, transmit it, then precisely measure the returned signal for phase, amplitude, and waveform integrity -at that speed. Oscilloscopes and logic analyzers that run at that speed, are damn expensive.

But for bandwidth can't you just have two simple computers at either end that can send a 100GB file and if it arrives in 20 seconds, you know you have 40gbps of bandwidth?

You have to first identify and ensure the two devices and their ports are capable of reliably creating and absorbing that data rate too or you'll never know if it's the cable or the devices that causes the cables to fail. What if your devices can but only intermittently or a software/firmware update messes with it's capabilities?

Dedicated test equipment that can reliably and verifiably test these high speeds is niche and therefor expensive. There's not really a way around that issue.

Re: I stopped trusting USB-C cable labels and started testing them

#227

Post by selicos »

dmos62 wrote: post #144
stouset wrote: post #84

Earlier quoted context omitted.

Okay, great, it’s called USB-D. Is this a new connector? Or do we now need to resort to labeling cables and checking the label? Do we do this for every new version and combination of versions for USB? USB-PD? Thunderbolt? DisplayPort?

Ehem, we already do that. USB-C is an umbrella term and everyone has a homebaked organizing system to tell which USB-C cable can do what, or a homebaked testing routine where we run through 5 cables each time we need to connect something. That's what the article is about. Calling all these cables USB-C and making them look indistinguishable is a very leaky abstraction.

They are all USB C form factor port cables. The actual functionality being different is a USB spec/labelling failure not one of naming.

Thunderbolt is also a USB C form factor/port, just has better naming and private control over usage.

I have a similar USB tester as the author. It's rather useful and inexpensive. Good bit of homelab kit.

Re: I stopped trusting USB-C cable labels and started testing them

#229

Post by CamperBob2 »

ChrisMarshallNY wrote: post #167
CamperBob2 wrote: post #110

Earlier quoted context omitted.

Someone could do an equivalent-time sampling instrument for not much money, if the market were there. It's probably not, though.

Time sampling isn’t enough. You need to at least look at waveform integrity (which can be approximated, by doing things like looking for harmonics). For example, you may send in a 40 GHz square wave, and get back a wave full of ringing, or a sine or sawtooth wave. It’s damn difficult to maintain clean square waves, at that frequency. The cable could also double as a microwave antenna. Since it’s digital information,…

That's not really what I mean. You don't have to diagnose the signal, just the medium, which can tell you what will happen to the real signal. It requires a fast edge (or a very broadband sweep), but nothing much beyond that.

E.g., here's what a bad connector at the far end of an SFF-8654 cable looks like on a TDR: https://i.imgur.com/AiLnSuC.jpeg . This type of plot can be interpreted as impedance versus distance along the line. It could be made with a small handheld instrument, but currently a human is needed to read the tea leaves. While I was making that measurement, I couldn't help thinking that it would be nice if it didn't require either a 50 GHz VNA or equally-expensive training to read.

Equivalent-time sampling gets you out of needing an expensive network analyzer, but it doesn't help with the diagnosis itself. And of course when checking a USB cable you are most likely just after a go/no-go result, not the gory details. That further calls the market economics of such a gadget into question. Still, if somebody offered one at a reasonable price, they'd sell at least one to me.

Post reply on HN