What's so special about USB

[zdas04]

I built computer systems in the early days of personal computing (IBM PC and PC/XT days). During that time manufacturers of accessories using Serial or parallel connections kept messing with the pin-out on the cables. Every cable had to be labeled for the device it worked with.

Today everything is USB and seems to really be universal. I've got a dozen mini- and micro-USB cables floating around and any device that one of them fits in works fine both for connecting to my computer or for charging. What is it about USB that has prevented device manufacturers from screwing with the pin-outs? Whatever it is, I am certainly glad for it, but I find it surprising.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[IRstuff]

Actually, there appears to be some subtle differences for charging applications, but I'm not clear what those are.

USB was developed specifically by a wide consensus industry group to avoid some of the issues that arose with RS232 and RS422/RS485. It appears to be a clear confluence of industry acceptance and adherence, and deviators know that they will be potentially stranded if they deviate. Even the micro USB-B connector was not widely used until Android devices standardized on them.

The area where there was the greatest variety in connectorization was in cell phones, and eventually, the manufacturers recognized that having non-standard connectors was detrimental to their ability to maintain their market share.

But, note that like other standards, the ubiquitousness of USB was an evolutionary process; Firewire was, at one point in time, a serious competitor to USB, but has essentially fallen by the wayside, due to IP issues. USB does not have such issues, and even proprietary connectors tend to terminate in USB on the other end for comptibility with PCs.

TTFN
faq731-376

 

[MikeHalloran]

ISTR there are only four pins in USB, so pinout variations would likely result in smoke. The only serious differentiation has been limited to the physical size of the connector, and the cost of tooling for quantity production has limited that to a few sorta-standard sizes, few enough for Radio Shack to cover with relatively few part numbers.

One borderline magical feature that is not apparent from the physical manifestation of USB is that the various connected devices exchange digital information with each other, about what they do and what they expect from the bus host, including allocation of electrical current.



Mike Halloran
Pembroke Pines, FL, USA

 

[BigInch]

The old connections actually only used 4 wires anyway.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[Skogsgurra]

The "special thing" is that it IS a standard with almost no exceptions. And that IS amazing. Especially when you consider all the wall outlets and plugs that are used for houshold appliances and voltages. Even in limited regions like Scandinavia, we cannot always be sure that the connector does its job if plugged into an outlet - especially across Swedish/Norwegian border. And in Europe, there must be at least ten variations. With the 13A "Claw" being the most deviating one that I have seen so far. But, then again, I never travelled Russia.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

We like our deviant plugs.

 

[zdas04]

I get that there is a published USB standard and that there never was an agreed-upon standard for Serial and Parallel, but why in the hell would someone decide to wire their device so that they have to build a cable connecting pin 1 to pin 7? I had several devices that required their own circuit board so they were in complete control of both ends of the wire and they still had unique pin-outs. What could they possibly have been thinking?

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[BigInch]

SELLING THEIR PROPRITORY CABLE FOR $49.00 OF COURSE.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek

 

[MikeHalloran]

Blame it on the Bell System.

The interface between a teletype and a modem used all 25 pins of a 'D-subminiature' connector.
Each pin had one function. Some of them became redundant when the interface was adopted for computers, but not all. The redundant ones were omitted when IBM adopted the 9 pin D-sub for serial connections.

Throttling, in particular, was necessary. When you connect devices of disparate capability, it becomes an issue. If a fast transmitter just sends a big lump of data over the wire to a slow receiver, it will probably overrun the receiver's data buffer. So the receiver asserts a signal when it is ready to receive data, and releases it when it is not. One extra, dedicated, throttling wire for each direction. There are actually two levels of throttling in RS-232, one for packets and one for individual bytes. ... and I think another signal just to say the modem is plugged in. One signal per wire.

USB does not have or need handshaking/throttling signals; every piece of information travels in a standardized packet, and the data rate is settled at startup.

Back to RS232; Pins 1 and 7 are different kinds of ground, but they have to connect somewhere.

On the other, other hand, count your blessings. Before IBM 'standardized' the cabling by gaining huge market share, serial printer manufacturers all mis-read the 232 standard differently, so every one needed a custom cable.

Mike Halloran
Pembroke Pines, FL, USA

 

[zdas04]

That was where I started, with every device needing a special cable (usually not provided with the device).

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[ScottyUK]

And way back in the dark ages there was the mighty Digital Equipment Corporation - DEC - whose mission was to make every single one of their products directly compatible with nothing else on the planet. That they turned out some pretty decent minicomputers was secondary to the aim of eliminating compatibility with non-DEC hardware.

 

[IRstuff]

And yet, Apple contains its own lunacy with a $50 charging cable for their latest products. Of course, it's brilliant on their financial level, given the absurdly high markup that there must be on this cable, and the fact that cables tend to be almost consumables.

TTFN
faq731-376

 

[VEBill]

It's not all flowers and sunshine in the land of USB.

Many USB enabled gadgets require unique cables to accommodate the unique connector at the gadget end. Latest example is the Apple iPhone 5 'Lightning' connector that is far more complicated than just USB. But many gadgets have combined other signals onto a common port resulting in unique connectors, and thus unique "USB" cables.

Next problem is the seemingly endlessly expanding variety of 'Standard' USB connectors. A, B, Mini, Micro, USB 3, etc.,....will it ever end?

Next problem is the ever changing standard for current output limit handshaking. It used to be as simple as 500mA The End. Then they've invented protocols to up the current. Not all devices play together nicely. There are certain combinations of proprietary chargers where the charger will refuse to provide more than 500mA to a device by another manufacturer. Shouldn't happen, and yet it does.

At the software interface layer, some devices are locked down and will only allow limited functionality except through their proprietary software. Apple product connected to a Linux PC for example. Maybe you'd get access to the photo folder - period.

All in all it is far superior to serial. Ever transfer a multi GB file by RS232 serial?

 

[GregLocock]

"Ever transfer a multi GB file by RS232 serial?"

ever backed an FEA model up on paper tape?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[VEBill]

No, but I've punched Fortran onto cards.

 

[IRstuff]

I've punched paper tape programs for a MacroData MD104 memory tester ;-) using an ASR-33 teletype... Those were the days

TTFN
faq731-376

 

[BigInch]

Scotty, we've got those rhino plugs in the UAE too. Thanks for that.

"People will work for you with blood and sweat and tears if they work for what they believe in......" - Simon Sinek