Talk:USB hardware

Cable differences between USB 1.x and 2.0?

The Cabling section is a little terse. Can anybody tell me if the cable standards (e.g. wires, twists, shielding) changed at all from USB 1.x to USB 2.0, other than the slight change of the maximum allowed length? I realize that was 23 years ago. Even if there was no real change (in other words, a pre-2000 USB version 1 cable would be within spec for 480 Mbit/s USB version 2 connections), I think that would be worth mentioning. — voidxor 18:00, 25 May 2023 (UTC)

Citation for "USB does not allow extensions cables"

Throughout the article there are several statements to the tune of "USB does not allow extensions cables" (specifically at https://en.wikipedia.org/wiki/USB_hardware#Connectors under the image for a non-standard extension cable and the first image in the article). However, this statement is not cited or referenced at all. Without personally going through the USB specification to see if this is true or false, the only talk about this seems to be from @Voidxor above referring to this archived discussion on the topic

It may prove beneficial to find an official source for this statement to reference, or failing that, append the good old [ citation needed ] to each mention of extension cables not being allowed. 61.69.232.98 (talk) 12:46, 1 July 2024 (UTC)

This took less time than I'd feared. In revision 2.0 of the USB standard dated April 27,2000 (available on the USB.org Web site), section 6.4.4 "Prohibited Cable Assemblies" says explicitly that extension cables are not allowed because they can create a cable that exceeds the allowed length. --Wtshymanski (talk) 17:10, 1 July 2024 (UTC)

 Done. Thanks for the legwork, Wtshymanski. — voidxor 23:04, 1 July 2024 (UTC)

A-A cables are actually permitted by the standard (SuperSpeed only, no power)

The article claims about A-to-A cables: "Existing for specific proprietary purposes, not inter-operable with USB-IF compliant equipment and possibly damaging to both devices when plugged in. [...] Other combinations of connectors are not compliant. There do exist A-to-A assemblies, referred to as cables (such as the Easy Transfer Cable); however, these have a pair of USB devices in the middle, making them more than just cables."

But the USB 3.0 standard describes a "USB 3.0 Standard-A to USB 3.0 Standard-A Cable Assembly", which works for SuperSpeed only and does not have the USB2 pins except GND (VBUS, D-, D+) connected through. (Those cables are mainly used for stuff like debugging of the operating system.)

(And those cables actually exist - for example, https://www.datapro.net/products/usb-3-0-super-speed-a-a-debugging-cable.html sells them. I've used one myself.) TheJH (talk) 00:14, 4 August 2024 (UTC)

Naming of legacy SuperSpeed connectors

In general it’s difficult to find the balance between the technical and the marketing vocabularies, not least because the USB-IF has gone out of their way to make a complete mess of this for decades. @ZH8000 changed USB connector names from e.g. 3.1 Standard‑A to 3.0 Standard‑A. I had used 3.1 because those are the official names in USB-IF materials. (They name connectors for the latest version they support, so each legacy connector is designated 2.0 or 3.1 with the exception of the Mini connectors which have no number because they only support ≤ 2.0.) Before USB 3.1 these connectors were designated 3.0 but with 3.1 their names changed to 3.1 just as the original connectors had no such designation until USB 3.0 renamed them with 2.0. I know there is a difference between the most technical vocabulary and what should often appear in articles on the topic, but shouldn’t 3.1 be the term here? Stephan Leeds (talk) 00:05, 2 April 2025 (UTC)

Stephan Leeds (talk · contribs) tldr: Theoretically you are right. But since for all three versions of the 3.x standard, the (leagacy) connectors did not change (at least not their PHY).

Every updated version of USB 3 (3.0 --> 3.1 --> 3.2) fully replaced the previous version, including the specification of the connectors. Between 2013 (USB 3.1) and 2017 (final USB 3.2), if you looked up their specification, you had to consult the USB 3.1 specification. Therefore they called them e.g. USB 3.1 Micro-B. But they were defined as before in USB 3.0. Currently, you should call it USB 3.2 Micro-B, since USB 3.2 is the only still valid specification for the third standard of USB.

But, since the specs did not change you can still use USB 3.0 connectors in place of USB 3.2 or USB 3.1 connectors, and vice versa. So, from user's perspective, it makes no sense to use different names for the same thing. Quite contrary, it would surely provoke misunderstandings and make things more complicated than necessary. So therefore, the consencus among USB 3.x-related articles was to use the "first" spec, namely "USB 3.0" in order to distinguish them from previous versions.

Instead of USB 3.0 Standard-A you could also use USB SuperSpeed (or: SS) Standard-A, of course.

BTW: In your new USB 3.x connector icons, would you mind to add an indication about the additionnal five/six pins in order to make them look different from a USB 2.0 connector? Actually, I like, or even prefer, the more fine-grained "original" icons of the general USB artical. Would you mind to use them instead? Update: I see now, the USB 3.0 pins are actually already there, but in small sizes you can hardly see them. Could you make them a bit brighter then, please?

-- ZH8000 (talk) 11:10, 3 April 2025 (UTC)

If the article calls what are officially named USB 3.1 connectors instead by the lowest version they support (USB 3.0), then Wikipedia will be in conflict with the official naming, and I think sticking to the official names is least confusing. Also, if the article refers to the second generation connectors as USB 3.0 then the earlier connectors must be referred to as USB 1.0, not the official USB 2.0. Calling the USB 3.1 connectors USB 3.0 requires calling the earlier ones USB 1.0. Using SuperSpeed instead of a version number is problematic because it leaves no elegant (as if that’s even a possibility given the behavior of the USB-IF) and consistent way to name the earlier connectors. Calling them non-SuperSpeed is even clunkier than using the version numbers. The version number system could be pushed a bit closer to elegance by deleting the obviously-overprecise .0 from 2.0, but that also means using unofficial terms, which is a bad idea in encyclopedic register. Original and SuperSpeed makes the distinction clear, but SuperSpeed itself is terribly problematic because it implies that the legacy connectors can do something Type‑C can’t, further perpetuating the problem of people not getting that the entire point of Type‑C is (and has been for more than a decade) that it is for everything.

Can we seek consensus on either using the official terms or the lowest applicable version numbers? Stephan Leeds (talk) 08:50, 4 April 2025 (UTC)

The answer is actually quite simple: In order to achieve the High-Speed capability USB 2.0 connectors must differ electrically (!) from previous ones (USB 1.0/1.1). For USB 3.x however, the operation mode is negotiated (by host, intermediate hubs, and peripheral devices), and not based on PHY layer (while excluding power capabilities). Therefore you cannot replace a USB 2.0 connector with a USB 1.1/1.0 connector.

Whether you use USB SuperSpeed–what corresponds with the USB Low-, Full-, and High-Speed naming convention–or USB 3.0 is not really important IMO.

What does SuperSpeed imply that Type-C can't? I don't get you. -- ZH8000 (talk) 09:09, 6 April 2025 (UTC)

Completely incorrect. USB2 type A and B connectors are identical to USB1 type A and B connectors (these were the only connectors USB1 specified though USB2 did introduce mini connectors). However: the cable specification did change. The USB3.x type A and B connectors are different in that they have 5 extra contacts. The type A USB3.x connector is backwardly compatible with USB1 type A connectors (though the type B connector is not).

It is the high speed mode for USB2 that is achieved solely by negotiation, largely because USB2 requires back terminated transceivers compared with the simply terminated type for USB1. High speed USB2 also cannot operate with the pull up resistor used to detect the connection and disconnection of USB1 devices. USB2 devices initially communicate in USB1 fast mode. Once the negotiation determines that the peripheral is a USB2 device. Both ends disconnect the USB1 transceivers plus that pull up resistor and connect the USB2 transceivers and high speed communication begins. This is why you can plug a USB1 peripheral into a USB2 port with impunity - you would not be able to do this if the connectors were different. The negotiation for USB2 fails because the USB1 peripheral does not respond to the ports attempt at negotiation.

For the avoidance of doubt: USB1.1 changed virtually nothing from USB1.0. It merely cleared up a minor ambiguity over the operation of hubs. 2A00:23C8:9883:A001:B170:3E92:31C6:AE41 (talk) 14:11, 6 April 2025 (UTC)

The previous poster does obviously not know what he speaks about. Chapter 7 (pp. 119ff) of the USB 2.0 specification (27 April 2000) clearly shows a different (from a USB1.0/1.1, only capable for low- or full-speed) electrical diagram for a high-speed capable connector; c.f. figures 7-20 up to 7-24. Nothing else to add. - I suspect the previous poster as a very, very old socketpuppet reappearing year after year like a cockroach; see Wikipedia:Sockpuppet investigations/I B Wright/Archive; do not take him seriously and do not respond to his posts anymore. I will not, he's only intention is disruption. Do not feed the animal! -- ZH8000 (talk) 18:43, 6 April 2025 (UTC)

The connector diagrams are literal, from the specs, so the contact sizes are literal or, where the spec is imprecise, something reasonable within the allowed range. I’ll look at making them brighter or larger, or both, but the reason they’re dark is do distinguish the gold contacts that are barely visible deep inside the real connectors from contacts that are plainly visible to the user. It may be that an entire separate set of less realistic illustrations would be helpful. Stephan Leeds (talk) 09:05, 4 April 2025 (UTC)

It is just about the brightness / visibility of all pins in small icons. -- ZH8000 (talk) 09:11, 6 April 2025 (UTC)

Unimportant BTW: I my opinion the replacement of a single & nbsp ; with the nowrap template makes no sense, since the template makes nothing else but the opposite, re-adding & nbsp ; again. So, besides using useless additional computing resources, it does nothing useful. -- ZH8000 (talk) 11:30, 3 April 2025 (UTC)

The issue is that it was the HTML entity inside a {{nowrap}} template, which is redundant. Either is fine. Stephan Leeds (talk) 08:46, 4 April 2025 (UTC)

Well, the template needs additional computation resources (mainly time). -- ZH8000 (talk) 09:13, 6 April 2025 (UTC)

Some connectors are missing

Come here to identify a USB port of one of my devices and this page doesn't contain it. Found this image with captions on Wikimedia though.

USB-Steckerformen

Interestingly it's a newly bought device and it has the connector to the very right "Typ Miniatur-B 4-polig (Aiptek)". Anyone know more about this connector? Is it something for e.g. governmental or regulatory environments?

For reference:

• The device with that port: cyberJack® RFID komfort
• Documentation only mentions it as "USB 2.0"

I also took this picture of the receptacle for wikimedia:

Proprietary Aiptek USB Mini-B receptacle from a "cyberJack RFID komfort" smart card reader from Reiner SCT newly bought in 2025

Agowa (talk) 05:02, 15 April 2025 (UTC)

WP:NOTFORUM — ZH8000 (talk) 12:43, 16 April 2025 (UTC)

@Agowa: It's proprietary as you say, and therefore doesn't deserve coverage in this article. Unfortunately there are far too many manufacturers that don't follow the USB standards. ZH8000 is correct that we're not here to help you identify it—only to discuss improving the article. — voidxor 14:07, 16 April 2025 (UTC)

Ok, but do we have another page covering the proprietary ones then? Not covering them at all also doesn't seem right. Agowa (talk) 21:57, 16 April 2025 (UTC)

No, because that wouldn't meet our notability standard. I also know from experience that there are, unfortunately, an endless number of the proprietary ones. We cannot and should not attempt to cover things that are so ephemeral in a high-level encyclopedia. — voidxor 22:01, 16 April 2025 (UTC)

Force colors to not change in dark mode

In the tables regarding USB colour coding, the black and white backgrounds are reversed when using Dark Mode. The text remains correct but the colours are flipped. Is there a way to force the colours to not change when dark mode is toggled? OneToRuleThemAll (talk) 11:06, 2 May 2025 (UTC)