Talk:USB/Archive 3

Archive 1

Archive 2

Archive 3

Archive 4

Archive 5

→

Archive 9

Daisy Chaining?

The talk of "Daisy Chaining" is quite misleading. USB devices can not be daisy chained. Its either a device at the end of the wire, or it s a hub. You can incorporate hubs into a box, but you're still only connecting to the hub, not daisy chaining. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

I don't understand what Barry thinks "true daisy chaining" would be. For most users, this is a practical concept, a physical configuration. If X plugs into Y plugs into Z, then they are daisy chained, regardless of what technical innards makes this possible. 69.87.202.196 13:58, 7 December 2006 (UTC)

They probably believe that daisy chaining would be akin to how FireWire does it. In USB, you can't have USB ports on a random hard drive unless it's also got a hub, in which case it'd actually be (to the computer you're connecting it to) a hub with a hard drive behind it, i.e. two devices. —The preceding unsigned comment was added by 64.81.54.24 (talk) 07:16, 17 March 2007 (UTC).

Speed

The actual theoretical maximum data transfer rate for bulk transfers is 50MB/s (13 packets per micro frame). The most ever recorded to one endpoint is 43MB/s (11 packets per micro frame). I've personally seen transfers up to 39MB/s. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

I have read that USB reserves at least ten percent of the communications channel for control info. If this is true, it would be good to include in main article. And if you have achieved real file transfers at an overall speed of over 30 MB/s, it would be interesting to hear exactly how you did it, and what you think the key factors are in achieving such speeds. 69.87.202.196 14:02, 7 December 2006 (UTC)

Configurations

The discussion of configurations is largely theoretical, very few devices actually impliment more than one configuration. That's probably due to Windows not supporting multiple configurations in any meaningful way. To say "These configurations often" is misleading, when configurations aren't often implimented.

The iPod is probably the most common device which impliments multiple configurations. Most of them use a second configuration for a control and digital audio mode. The iPod shuffles impliment low/high power configurations, even then it needs to play games to make this work on Windows. (An iPod shuffle will disconnect if not configured and try again at low power.) Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

Host controllers

The Host controller is called the Host Controller or HC (not HCD), HCD referrs to the Host Controller Driver. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

Signalling

The signalling section just describes full/low speed signals, not high speed. High speed is a current signal which works out at about a 440mV differential signal. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

Transaction translators

"there is no standard way to determine the number of transaction translators a hub may have." The protocol of the device and the interface tell you if the hub is single or multi-tt. A multi-TT hub must have one TT for each port. Hub drivers need to know this to manage hubs properly. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

Do standard Linux utilities like USBView or dmesg yield such info? Should such tools be mentioned in the article? 69.87.202.196 14:05, 7 December 2006 (UTC)

Low power notice

The caption to the picture "Mac OS X dialog displayed when the 500mA limit is exceeded" is incorrect. That's the alert which pops up when a high powered device is plugged into a low powered port. The port can supply 100mA and the device configuration declares it needs more than 100mA (up to 500mA). If a device "overcurrents", that is it draws more than 500mA, and "fuse blows" a similar but different alert pops up. That says the device has been shut down because of the power problem. Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

Devices with 2 plugs.

Devices with 2 plugs, to allow more power, are mentioned inthe article. These are specifially dissalowed in a supplimental document to the spec. (I can' remember which one, maybe an ECN, or the OTG spec.) Barry 5Dec06 —The preceding unsigned comment was added by 208.106.30.57 (talk • contribs) 2006-12-05.

I personally have only come across one type of device which used 2 ports, this was a usb IDE adapter for laptop drives. This is wideley avaiable on ebay, and is IBM branded, however i doubt they are official IBM products, as there build quality is poor, and the IBM logo's on them slightly incorrect (in my expierence). If this is the IBM/lenovo device the article is referring too maybe this should be taken into account. - Al

This very article addresses USB as a protocol. The port is really a non-issue because as long as the peripheral doesn't communicate using USB protocol, it really isn't any USB device, but rather just a device that takes power from the USB port. The hard drive kit you mentioned is like this, where one port is used to communicate in USB, while the second port is just for power. Groink 10:28, 23 January 2007 (UTC)

Vendor ID - Product ID

Every USB device controller hardware seems to have at least one set of Vendor ID - Product ID built in. Some utility programs, esp in Linux, make this information available, along with current power levels etc. It seems like this should be mentioned in the article, along with links to the best tables of Vendor ID - Product ID information. It would also be good to have links to the best lists and overviews of the various implementation chipsets. 69.87.202.196 15:22, 7 December 2006 (UTC)

USB-to-PS/2 adapter picture

My understanding is that these things don't "adapt" anything. When used, the mouse goes into PS/2 signaling mode. Perhaps this picture should be removed. 64.171.162.77 01:54, 10 December 2006 (UTC)

where can you locate the usb port

My daughter just got a muichiz brat game. I can never figure out how to plug it in. where can it be located?

Trimming

I've removed a lot of prescriptive information from the article. It may be that it was worthy of its own article, but as-is it just looked like a bunch of prescriptive information for potential USB developers and didn't really fit with the rest of the article. Chris Cunningham 12:00, 8 January 2007 (UTC)

White Spark

Today I was plugging an external USB2 hard drive enclosure (with a hard drive inside it, obviously), into the USB1 slot on my old PC. As I was doing this, I noticed a white spark come out of the USB1 slot! Is this normal or anything close to it? Maybe I only noticed it because I have my computer under a desk and it's dark down there. Seems weird though. Doom jester 12:56, 31 January 2007 (UTC)

USB History

The design of USB is based on the upstream/downstream communication concept invented by Wynn Smith who founded Intel's Hunacre group. The upstream/downstream concept greatly simplified the user experience by allowing a computer to communicate with hardware without requiring the user to set switches or assign hardware addresses.

Verification: Considering the incredible importance and success of USB, I beg the keepers of the USB article to please dig in and find out where it came from. It's an important part of PC history that has been completely ignored. USB truly changed the world. Because of USB, it became possible for ordinary people to augment and enhace their personal computers. USB is a key element in the growth of the PC market, for example, with the common usage of imaging with scanners and digital cameras. It allowed an endless variety of peripherals to become easy to use and useful.

Ralph Smith was a member of both Intel's Hunacre team and Intel's first USB team. Jim B Johnson was General Manager of Intel's PC Enhancement Division where Hunacre was formed, upstream/downstream communication was invented, and the first prototype was built. (Ralph Smith still has the first prototype.) The original Hunacre team included Wynn Smith, Ralph Smith, and Bill Krueger. The upstream/downstream communication concept lead directly to development of USB.

You can call me at (503)650-9496, or email wds.0713@oregoncity.com

WynnSmith 09:35, 17 February 2007 (UTC)

USB Conflicts?

I have several (4?) USB connectors on my PC (2 front; 2 back), but I can only have one device plugged in at a time. If I plug more than one item in neither will work. One works fine, but plug a second in and neither will work. Why is this?

Actually, I have two desktops, one about five years old, the other about 3 years old, both run XP OS. They both behave this way: only one USB device will work at a time. Any suggestions how to resolve this problem?Raybobo 19:04, 17 February 2007 (UTC)

This is not a troubleshooting forum. It isn't normal behavior and may be a problem with the devices you are using. SchmuckyTheCat 19:26, 17 February 2007 (UTC)

It may also be the design of the USB bus on your PC. There are PCs that have many USB ports, but all the ports are on a single USB bus. This is common for built-in USB ports from the factory. My Dell PC has six on the back and two on the front, and all of the ports are on a single USB bus. What happens in this situation is that the more devices you connect to a single bus, the less power you have going to each USB device. For low-powered devices like flash drives this may not be a problem. But if you have USB devices like scanners, printers, IDE/SATA hard drives, etc. - all of which require quite a bit of power, then each of these devices may have an effect. If you read the USB article, you'll see that a USB bus provides 500 mA of power, and each devices is allowed to draw only up to 100 mA. But if even one of your USB devices doesn't fall within USB compliance for power consumption, or the combined total of power exceeds 500 mA, it may cause problems on the USB bus like what you're seeing. One work-around would be to install a PCI-based USB2 card to give the PC a totally separate USB bus. Another would be to connect a powered USB hub so that the external power provides more amps to the bus. Groink 21:14, 17 February 2007 (UTC)

Bus-powered scanners and printers would be scary for a different set of power-related reasons.

Logo design

Who designed the USB trident? --24.249.108.133 22:24, 19 February 2007 (UTC)

I don't have a specific name but would imagine it came from the USB Implrmenters Forum (USB-IF), "an industry standards body incorporating leading companies from the computer and electronics industries. Notable members have included Apple Computer, Hewlett-Packard, NEC, Microsoft, Intel, and Agere." Livedream 06:27, 17 March 2007 (UTC)

Data encoding section

Data encoding section is not very well written. —Preceding unsigned comment added by 67.85.110.92 (talk • contribs) 12:48, 27 February 2007 (UTC)

USB Compatibility?

I was under the impression that there are three major versions of USB: 1.0, 1.1, and 2.0 and that they are all compatibile and that you don't need to know which computer, which cable, or which peripheral device are compatible, that they are all compatible if the cable connections fit. Is this true? I have read that USB is backwardly compatible. Does this imply that it is not forwardly-commpatible? I thought that the main differences between the three major versions of USB were speed differences and that the USB protocol would automatically adjust speeds to be compatible. Is this true? —The preceding unsigned comment was added by 166.70.81.134 (talk) 07:18, 11 March 2007 (UTC).

Yes. SchmuckyTheCat 22:56, 11 March 2007 (UTC)

USB Notched Connector

Sorry, but i couldnt find any info on this so was wondering if anyone could help. Was just wondering what the Notched connector was for. The receptical looks like a USB A-type connector but opposite the connection there is a small notch that looks like a V. I would assume that it is designed to prevent certain types of usb devices connecting to it (although 'regular' type-A connectors appear to still fit into it- albeit a tighter fit). Any info on this connection would be appreciated, Thanks

Livedream 05:50, 17 March 2007 (UTC)

Notching/keying is sometimes used to prevent a cable (like a cable with two plugs, the second being 'dead' for USB and only to draw additional power from another port) from being used on a random device, or to make (for example) an extender cable work with that manufacturer's keyboard but not with all USB devices. USB extension cables are technically against the USB spec as I understand it. 64.81.54.24 07:28, 17 March 2007 (UTC)

Thank you Livedream 15:26, 17 March 2007 (UTC)

USB/PS2 compatibility - is it PS2 protocol or USB?

I was looking for some info that I note the section on HID devices is lacking. It mentions that often keyboards and mice come with a USB connector and an adaptor to PS2. The adaptor is purely mechanical, there is no protocol translation. The question is, is the protocol being used USB or PS2? I would have thought that older PCs would not accept USB protocol into a PS2 port, so always imagined that it was PS2 protocol being sent over a USB connector.

What about high end mice like the Logitech ones though? They make full use of USB for some features. Do they fall back to PS2 mode if no USB host is found? —The preceding unsigned comment was added by 212.248.168.208 (talk) 16:41, 17 March 2007 (UTC).

The device knows what port it is plugged into and adjusts accordingly. SchmuckyTheCat 17:06, 17 March 2007 (UTC)

Thanks. I wrote the original comment while at work. I'll add it to the article. Mojo-chan 13:17, 18 March 2007 (UTC)

PAGE META DATA VANDALISED

Please check the page's meta data (as seen when found on a google search). Present data lists it as 'Universal Serial Banana'. I don't know how to change this. 217.36.120.131 22:12, 18 September 2007 (UTC)

NPOV on firewire / USB 2.0 hi-speed comparison

I've been trying to update the USB 2.0 Hi-Speed vs Firewire section to reflect the result of every test I can find which is that (a) firewire 400 is faster than USB 2.0 Hi-Speed for use with harddisks. and (b) firewire 800 is really the more appropriate standard to compare to because it's the hi speed version of the firewire standard. What is there is a long couple of paragraphs with few references that aregue USB 2.0 should be faster and then point out that it is in fact slower in tests. It makes sense to start with the key points and then move on to the less important points. I would say a and b are the important points. Theoretical arguments can follow. Pdbailey 22:36, 19 March 2007 (UTC)

Why not just remove the whole thing? It's unimportant. SchmuckyTheCat 23:44, 19 March 2007 (UTC)

The title should probablly be retitled USB 2.0 VS firewire 400 and it is an important comparison because for some years they were (and still are to some extent) the main competitors for external storage and while USB had the faster headline speed firewire won by far in the tests. It is important information to anyone deciding between the two interfaces that a supossedly (slightly) faster interface is in many cases actually much worse.

I don't think there is any controversy about usb 2.0 vs firewire 800, firewire 800 is the clear winner there both theoretically and practically. Plugwash 00:00, 20 March 2007 (UTC)

I concur with Plugwash, the information is important. It's not obvious that firewire 800 is that much better, so I would suggest that we include that information as well (it could just be adifferent name for the same stuff). Pdbailey 00:47, 20 March 2007 (UTC)

Delete: I agree with SchmuckyTheCat; this whole section should be removed. Comparisons are usually inspired by some form of agenda on the part of the creator, rather than just providing the information in an encyclopedic matter. I've yet to see any encyclopedia entry that compares two American presidents, two automobile engines, two sports, etc. I would find technological comparisons to be of use in a Consumer Reports article, and not an encyclopedia. That's basically how I see this USB vs. Firewire comparison. Groink 00:58, 28 March 2007 (UTC)

Groink, your suggesting an ad hoc policy of not including comparisons, right? Pdbailey 02:58, 28 March 2007 (UTC)

Especially in technical documentation like this, yes. I would think it is fine to post a table with verifiable "theoretical" data as published in reference manuals and leave the data it as-is, letting the readers interpret the data for themselves. But to actually interpret the data for them and give analysis within the same article, I think that is when you start running into problems, such as the methodologies used in the experiments (hardware types, operating system, manufacturer, etc.) Matter of fact, funny I brought up Consumer Reports earlier; in that article, there are numerous mistakes made on the part of Consumer Union because of the methodology used in their experiments, such as mis-calculating MPH speeds of a vehicle on a conveyor belt vs free-moving. I learned way back in Research Paper 101 that there's a difference between data and information: information is the interpretation of the data. I have my personal opinions about USB and Firewire, and I tell my clients what my findings are. But that's because they've hired me for my experiences with technology. Just because someone posts interpreted data on the Internet that does not make the findings acceptable for use on Wikipedia just because it is published. Personally I know the guy who runs Bare Feats; he's originally a Macintosh fanatic from Hawaii who was a member of our user group, and I've worked with him on a number of experiments before left Hawaii. Back then and even today, I question his methods. That's the problem with Wikipedia - that "verifiable" clause just means if you read it on the Internet, then it must be true. I don't buy that! That's why I think comparisons of any two given items should not be allowed on Wikipedia - and especially on this USB article. Groink 03:57, 28 March 2007 (UTC)

I completely agree. Comparison can be made by readers themselves, simply by reading both USB and Firewire pages! Hellgi 18:40, 29 March 2007 (UTC)

No THEY CAN'T! THATS THE WHOLE POINT! someone reading both the USB and firewire articles will see the headline speeds and some other information that doesn't seem particularlly related to performance and conclude that USB is faster when every comparitive test i have seen says the opposite. If we remove the comparison we are misleading people through ommison. Plugwash 19:05, 29 March 2007 (UTC)

I agree. The theoretical speeds are misleading as Firewire is much faster than USB2 in practice and it should be noted. Technology companies are notorious for misleading specs (Intel's P4 MHz myth vs AMD's offerings come to mind). While an encyclopedia is no place for analysis nobody can honestly dispute that Firewire is in fact faster than USB2. Robbh66 02:35, 16 April 2007 (UTC)

I don't think so. In your comparisons you sort of implying that USB specs LIE about actual speed. And that statement IS misleading. Are you questioning the ability of USB2 connection to actually move bits with declared speed? Probably, no. So what are you talking about? In fact, you're saying that in the chain of electronic components and software drivers between RAM and harddrive there is a bottleneck that limits bandwidth below theoretically possible for for USB2. But that doesn't mean at all that anything is wrong with USB2 itself or that its specs are incorrect. The most likely culprit is an HD controller (ATA/SATA to USB2 adapter or whatever it's called) but ths particular articl is NOT about controllers. If/when somebody makes a better controller (are you absolutely sure it isn't here yet?) it would be still the same USB2. So why say that USB2 is slow while in fact that slowness should be attributed to a device connected rather than the USB2 itself?

Now, I'm not saying that comparison data should be never mentioned anywhere. Just not in this particular article. Those comparisons are NOT the comparisons of USB2 vs. FireWire, they're comparisons between certain models of EXTERNAL HARDDRIVES utilizing aforementioned interfaces. So in an article discussing such harddives the comparisons would be welcome but not here. And even there it should be clearly stated that such unfavorable data are due to the flaws of the controller design rather than communication protocol and hardware. 74.69.159.177 02:20, 27 June 2007 (UTC)

Misleading??? We're not Consumer Reports. We're an encyclopedia. The role of an encyclopedia is NOT to make sure the interpretation of the theoretical data given is fair or accurate to real-life use. Whether or not people are mislead because of information that is left out is not Wikipedia's responsibility. Wikipedia is not the source that should be telling people what technology is faster/slower/more superior. Leave that up to the pundits on the blogs and benchmark sites. Not just technology - this covers ALL aspects, such as political ideologies, governing models, commerce models, etc. In all of those examples, it is NOT the responsibility of Wikipedia to inform people what's the better model/technology/ideology. Wikipedia should cover only the blueprint and not real-life use/experiences. If the "theoretical" FACTUAL data for USB 2.0 shows it being slower or faster than IEEE 1394, and you're saying that you've experienced the opposite, then IMHO it is the implementation of the technology, and NOT the fault of the blueprint. Groink 21:22, 29 March 2007 (UTC)

I don't get it. There is a real transfer rate for hard drives, every measurement agrees, the technical specs are misleading, but you think that saying this is somehow spin central because it implies analysis? What's the analysis, its just a cited source. Pdbailey 04:35, 30 March 2007 (UTC)

I decided you guys need some clarity, I took a look at the section and saw the whole technical discussion as being redundant and extraneous, summarized it in two sentences that were already in the article. The information in the section was just a useless rehash of stuff thats on the firewire/usb pages and written as though it was an argument, honestly thats not the place for an argument. --69.111.56.122 23:17, 29 July 2007 (UTC)

This is Bobby B. The information showing the real life speeds of USB and of Firewire are very useful and the backup showing how it was tested allows one to interpret the data. Not allowing one to show data of similar products side by side is too limiting. Knowledge comes more from comparisons and correlations than by raw data that can not be connected to anything else.69.228.83.182 03:09, 11 October 2007 (UTC)

Facts not mentioned

One important fact is not mentioned. In USB1.x specification, the maximum speed a single device can use is 6Mbit/s and total speed for all device cannot exceed 9Mbits/s. I do not know if there are similar limitation for USB2.0 but from comparasion of USB/Fireware, it implies there are.NYCDA 19:27, 2 November 2007 (UTC)

Masses of detail, but too technical?

An absolute bucketload of technical information is included in this article, all the way down to describing wire protocols. Is this really a necessary level of detail? Is there a sensible way of splitting out the most technical stuff to its own article? Chris Cunningham 13:44, 12 April 2007 (UTC)

IMHO, techies have no idea how to convey things like USB in a simplistic manner. They believe that by spitting out pages and pages of technical data, a novice will get something out of it. There is a time and place for knowing the maximum voltage to go across a piece of wire. For Wikipedia, it should serve the common man who is puzzled over how he can use the funky looking port on his new PC. LOL, maybe the Wikipedia folks should develop a new system called "How Sh*t Works Wikipedia", and move 98-percent of this article to it. Groink 10:47, 13 April 2007 (UTC)

Yes, we, "techies", truly have no idea about what you, "novice" or "common man", do not know about USB. There are probably many levels of "common men", therefore you would need to define a "level of dumbness" if them, so we might try to help you and structure the article properly. But let me ask you this: If you do not want details, then why did you come to this page in the first place? Just plug it in, and it will work for a "common men", what could be so funky or puzzling here? But if you need details, you go here. But do not complain that the details appear to be quite above your head. Alexei123 17:28, 1 September 2007 (UTC)

For one thing, note the date of my last comment. Since then, this article was developed, and much of the technical information was moved there. Second, I was speaking for the other novices and not myself; I myself am a very technical person. But, as an educator, I KNOW how to present technical information to novices, AND know where the "draw the line". Groink 21:20, 1 September 2007 (UTC)

That isn't constructive. I'm well aware that the article is too technical just now, which is why I'm trying to raise discussion on doing something about it. Care to help? Chris Cunningham 11:12, 13 April 2007 (UTC)

I agree with this and will come back and make some suggestions. I'm working on several other articles right now. SchmuckyTheCat 16:17, 17 April 2007 (UTC)

I'd have to disagree. I'm an IT Technician/avid home user and popped onto Wikipedia due to curiosity about the plethora of plugs, pins and sockets for USB and the DVI connectors. The detail about different standards and things like mini and micro USB were a great help, and likewise the pinout diagrams for DVI are useful too. Maybe there is too much data, yes, but I don't think Wikipedia is concerned about keeping file sizes down! I'd much rather pop onto Wikipedia and have all the data accessible to me there, and find the bits I want, rather than trawling through numerous websites trying to piece together the information I need. Thepineapplehead 12:13, 18 April 2007 (UTC)

The point is that Wikipedia is not an indiscriminate collection of information. Stuff which is better off in a technical manual would be best put in a technical manual on Wikibooks, where it's only one click away. As-is, the article contains so much technical info that for people who know nothing about USB and want an overview it's very, very hard going. Chris Cunningham 12:38, 18 April 2007 (UTC)

I was not aware of the existence of WikiBooks (I'm fairly new to being a wp contributor) but yes, I now agree that should a technical guide about USB (or data transfer in general) be created as a wikibook, then much of this information should go into it, and this article condensed and simplified into what the "average user" needs to know. Many thanks for you response. Thepineapplehead 13:22, 18 April 2007 (UTC)

No problem :) I've moved a lot of the more technical stuff over to Wikibooks for the time being (although it needs a lot of TLC) and tried to make this article a little friendlier. Still some work to do of course. Chris Cunningham 14:09, 18 April 2007 (UTC)

Would that be this article here? http://en.wikibooks.org/wiki/Serial_Programming:USB_Technical_Manual There's a lot of stuff in there, and a lot of the stuff from the Wiki article should be moved over to it as well. A lot of it is probably already in there so there's no need to duplicate info, but maybe use what we have her to pad out the stuff at WB, and add missing details? It's also missing Micro-USB - like Grich asks for in the discussion page. Perhaps (if you're not already) liaise with him/her and see if they have any more plans for it? Thepineapplehead 14:33, 18 April 2007 (UTC)

That's the one, yeah. With any luck someone will take up the cause of completing it. Chris Cunningham 14:42, 18 April 2007 (UTC)

I'm just an user. I was interested just on on the voltage (then I read all of this article because interesting). Please don't move power specifications. Dario Martini 00:01, 10 June 2007 (UTC)

LUC> I am a bit puzzled by this article. The USB standard is very well structured with a very nice introduction chapter, then followed by "climbing" chapters (mechanical, electrical, Signaling, Framework...). Why don't we see that structure in this article? I also wonder why there is no reference to packets/transactions/transfers which is the heart of USB and also simple to understand. Ex: A packet is the smallest piece of data on the bus, starting with a SYNC electrical signal and finishing with an End Of Packet signal (also purely electrical). A transaction is up to three packets and always initiated by the host. They are composed of specific packets (IN, OUT, SETUP, DATA) that are called in specific order to transmit commands/data to or from the device. Transaction are unsplittable: once started, they must finished before any other bus transactions. Transfers are a series of transactions dedicated to an unique endpoint. It is the highest level of communication in USB. It exist four transfers: Isochronous, Bulk, Interrupt and control. Etc... Once we know that we can even explain in simple words what's the enumeration, the classes... And there are no real technical details in saying that. We just respect the standard. My references: I developed a complete embedded USB stack for certain televisions and I am also a USB trainer for other engineers. - Lfourestier 08:20, 14 June 2007 (UTC)

Better to have at least rudimentary technical details in the article. And put the advanced stuff in wikibooks. I want however be able to find out pinout, impedance, volt levels etc.. Directly. The main problem is usually lack of a simple explenation, not too much information. Electron9 04:05, 28 August 2007 (UTC)

History?

Is there any history on USB available? When was it introduced? When was it widely available? The two paragraphs in the History section seem pretty recent. j-beda 15:08, 25 June 2007 (UTC)

I don't remember any exact dates but i remember most motherboards having it but hardly any perhipherals using it. Then the IMac came out which forced those who wanted to continue to serve the mac community to make USB versions of thier perhipherals. Plugwash 01:08, 26 June 2007 (UTC)

I seem to remember seeing strange ports on my computer back in 1998, not knowing what they were, I ignored them. They turned out to be USB ports, but at that time very little support for them existed. In fact, the P-166 that had them crashed and lost all BIOS programming the first time I plugged a scanner into the USB port. None of the early development seems to be included in the history section.Landroo 13:18, 1 September 2007 (UTC)

I found an article at IBM that expanded a little bit on the history although this section still needs a lot of work. Floorwalker 04:30, 10 October 2007 (UTC)

Generic hub limit in Mac OS

I see the article says that you can't have more than 5 generic hubs in the same line... is that a limitation of the USB specification or only Windows' implementation of same?

This is a limitation of specifications. Many USB transactions require responses from/to devices, and USB imposes a certain limit for the response time. Each hub has certain internal propagation delay, and the longest 5m cable has a delay too. Five hubs plus five 5-m cables makes up the limit for response.--Alexei123 07:32, 28 July 2007 (UTC)

Maximum Cable Length is ~3 meters with out a repeater?

(Sorry, my first post to Wiki... forgive any errors)

I'm not a USB expert by any means, but I understand from other engineers here at my company that you need a USB repeater after about 9 feet. But according to the article, you don't need a repeater under 5 meters.—Preceding unsigned comment added by ProsperousOne (talk • contribs)

5 meters, ~15 feet. SchmuckyTheCat

This is not mentioned in the article. If noone else does, I will add it. --Xerces8 08:53, 21 August 2007 (UTC)

The image on top only shows one USB type

I think that this image in #Types of USB connectors should replace the image on top of the article. Or we need (a table of) images that show the different types clearly on top of the article. Also we could need a table that lists the differences of USB types somewhere in the article.

We need a clearer distinction between the USB types for people who see an USB stick for the first time. I am not an expert, who knows them all, to do that.--Ollj 10:59, 10 August 2007 (UTC)

The image shows the USB plug which is by far the most common - if I'm not mistaken, all USB cables have this connector at the computer end. The other side is always different, but that can be explained later on. Note that this image also shows the logo quite nicely. On the other hand, I do agree that a second picture with many of the other connectors might be placed somewhere near the top. Classical geographer 13:43, 10 August 2007 (UTC)

OTG ?

In the article there is a picture of an OTG logo, but nowhere is explained what OTG is. —The preceding unsigned comment was added by Xerces8 (talk • contribs) 10:45:15, August 19, 2007 (UTC).

OTG is USB On-The-Go. There's a separate article for it. Groink 23:40, 19 August 2007 (UTC)

= Yet another plug type

There's another USB "B" connector that is almost square shaped, has no flanges on the sides like the micro or mini B, and has two corners slightly angled. The contacts are in a single row. This connector was used on devices like the RIO PMP500 MP3 player and many digital cameras. In the RIO and some cameras, the data lines are reversed but electrically it's identical to USB. The cables with reversed data lines do not have the USB "trident" logo (they have an arrow) while cables using this connector without reversed data lines do have the USB "trident" logo.

I'd put up a picture of the connector but I gave away my RIO PMP500. Radio Shack sells a (tremendously overpriced) "Digital Camera Cable" with this connector and reversed data lines. (The red power wire must be cut to use it with RIO MP3 players.) This connector is also included on an adaptor in some universal USB cable kits. —Preceding unsigned comment added by Bizzybody (talk • contribs) 06:46, August 27, 2007 (UTC)

dozens of non-standard B connectors exist. Yes. SchmuckyTheCat

Overhead?

USB suffers from overhead and other bad design practices. Would be nice with some accurate information on this in the article. This is particularly obvious when comparing to other interfaces. Electron9 04:09, 28 August 2007 (UTC)

10% is dedicated to the system interface. That isn't a "bad design practice"; that is .08% per device for a host controlled bus. I believe that is in the article. What accurate information is missing? SchmuckyTheCat

On second look a large series of mostly improvements in April deleted a critical section defining transport types. The section on control, interrupt, bulk and isoch should be restored. If someone else doesn't do it soon I'll do it tomorrow. SchmuckyTheCat

That is not the original USB logo

The sidebar/box has a caption indicating that the pic shown is the "original USB logo". But the logo shown isn't the original logo. The original logo included the three-branched tree with triangles and squares and so forth that is often embossed on USB connectors. The one shown looks much more like the logos developed when USB 2.0 was made.

Either the caption should be removed, or the picture changed. I would guess the caption, though I think it would also be worthwhile to show the older USB logo somewhere else in the article. —Preceding unsigned comment added by Jhf (talk • contribs) 06:49, 31 August 2007 (UTC)

Meta data vandalised

Please correct ASAP TDN 22:15, 18 September 2007 (UTC)

Prohibited cable assemblies (photos)

There are several prohibited cable assemblies pictured in in this article. Titled "Series "A" plug and receptacle", "USB extension cord" and "Different types of USB plugs". The first two are pictures of passive USB cables with a type a plug and receptacle. The third picture shows a type A receptacle with the other end not pictured.

These cable types are not allowed under the USB specification, section 6.4 "Cable Assembly" which states that all cables must have a series A plug and be terminated on the opposite end with a series B plug. "Any other cable assemblies are prohibited". Revisions to the specification allow for different types of A and B plugs and receptacles (mini, micro), but nowhere is a cable with a series A plug and Series A receptacle allowed.
These pictures should be removed or replaced. The use of 'passive' usb extension cords encourages violation of the maximum USB cable length which can lead to data and power problems, not to mention the fact that these cables cannot even be called USB as they do not comply with the spec. 70.75.14.108 03:19, 19 September 2007 (UTC)Dave

Why should we remove pictures of a commonly used item just because some beurocrats would rather it didn't exist? Plugwash 08:23, 8 October 2007 (UTC)

For two excellent reasons. First the very specification that defines USB outlaws these cables. Second as I had already mentioned, "The use of 'passive' usb extension cords encourages violation of the maximum USB cable length which can lead to data and power problems". I have personally seen this happen many times. The USB Implementer's forum FAQ http://www.usb.org/developers/usbfaq/#cab2specifically states that extending a USB cable by as little as 5cm can cause problems. At the very least we should mention the potential problems caused by using such a cable, and we certianly shoudln't be showing it as an example of a 'typical' USB cable, since it isn't either typical, or a USB cable. —Preceding unsigned comment added by 70.75.14.108 (talk) 03:40, 14 December 2007 (UTC)

I reckon it should be all means be pointed out in the article that some unapproved cables/configurations violate the USB standard and the problems this can cause. However burying our heads in the sand and pretending they dont exist isint the way to go. Far better to forewarn users of these issues. Leave the photos but have them clearly marked as unapproved cables/devices and include a section explaining why 80.229.222.48 (talk) 19:51, 14 December 2007 (UTC)