Fullscreen HTML5Med Res (??MB)HTML4

The History of the Universal Serial Bus USB Explosion! About 26 years ago, a group of companies developed the Universal Serial Bus or USB to make it easier to connect external devices to PCs, replacing the plethora of connectors and interfaces that had been used previously. It also greatly increased communications speed compared to existing serial protocols. Since then, the performance and uses of USB have grown dramatically. By Jim Rowe W hen the first generation of PCs or personal computers appeared in the 1970s – machines like the MITS Altair, the Commodore PET, the Tandy TRS-80 and the Apple ][ – they were somewhat limited in their ability to connect to peripheral devices like printers, modems and external tape or disk drives. But when IBM released their first PC (the 5150) in 1981, things started to change. The IBM 5150 PC was available with up to two built-in floppy disk drives, 16KB of RAM and a colour graphics card (for which a colour monitor was available). Importantly, it also had slots at the rear for plug-in interface cards to provide a Centronics parallel printer port and one or two RS-232C serial ports. Before long, you could also connect the PC to a 10MB hard disk. Many new PCs then started to appear, most of them offering similar features. By about 1990, just about every available PC had around 64KB of RAM, a built-in 20MB hard disk, a colour graphics card or adaptor and both a Centronics printer port and a couple of RS-232C serial ports. Many could also take a plug-in Ethernet card, so that they could be connected to a LAN (local area network). A variety of more specialised interfaces started to appear as well; for example, one to connect to the GPIB bus to control test instruments from a computer. There was also “FireWire” (IEEE1394), a high-bandwidth serial bus designed to efficiently connect peripherals like high-speed disk drives. Soon, the back of many PCs had a multitude of different interface connectors, to connect many peripherals. USB is born The development of USB began in 1994, when a group of companies that were heavily involved in the PC industry (Compaq, DEC, IBM, Intel, Microsoft, NEC and Nortel) got together and decided to make it easier to connect external devices to PCs. This would involve replacing all of the different interface connectors with a group of simpler, identical multipurpose connectors which could each be configured by software to perform a variety of interfacing tasks. So was born the Universal Serial Bus, almost immediately identified by the acronym USB. The official USB 1.0 specification was introduced in January 1996, and it defined two data rates: 1.5Mb/s (187.5KB/s), called Low Speed or Low Bandwidth (designed for peripherals like keyboards, mice and joysticks) The original USB cable for connecting peripherals like printers, with a fullsize Type-A plug at the computer end (right), and a Type-B plug at the device end (left). This appears on USB devices which the USB Implementers Forum has checked and considers to perform acceptably. 32 Silicon Chip Australia’s electronics magazine siliconchip.com.au This USB icon, or a variation of it, generally appears on or nearly all USB-compatible plugs and sockets. A variation of the USB-IF certification logo which appears on devices that are compatible with USB 2.0+ at 480Mbps. The square Type-B plug is too big for slim devices like smartphones and tablets, so the smaller mini Type-A (left) and Type-B (right) plugs were designed. These are still used by some cheaper devices, especially those which use USB 5V for charging, but have mostly been superseded by the micro Type-B or Type-C sockets now. The USB type-B micro plug and socket solved several problems with the mini type-B socket; besides being considerably slimmer, it’s also designed so that the plug wears out well before the socket, avoiding premature socket wear and ultimately, the need for replacement. The USB connectors that started it all. The rectangular Type-A appears on host devices like computers, while the square Type-B is used for peripherals like printers. This makes it difficult to accidentally plug two host or device ports together, which at best would do nothing, and at worst, cause damage. and 12Mb/s (1.5MB/s), called Full Speed (to handle higher speed peripherals like printers, disk drives etc). Intel produced the first interface ICs designed to support USB in 1995, but few USB devices or PCs equipped with USB ports appeared on the market until August 1998, when the USB 1.1 specification was released. This was soon widely adopted, leading to what Microsoft dubbed the “legacy-free PC”. The USB connectors used for these initial implementations were the flat-rectangle Type A socket (or receptacle) for the ‘downstream’ ports on a PC, and the square-with-chamfers Type B socket for the ‘upstream’ port on a peripheral like a printer (see Fig.1). Both connectors had only four pins, two for data and two for providing 5V DC power to the peripheral. Most external devices were connected to a PC using a USB cable fitted with a Type A plug at the PC end and a Type B plug at the device end. The exterior of both types of plug were identified with a distinctive USB logo known as the “trident” (see above). This was the situation until April 2000, when a revised USB 2.0 specification was released. This added a third signalling rate of 480Mb/s (60MB/s), named High Speed or High Bandwidth, in addition to Low Speed and Full Speed. It also allowed for Mini-A and Mini-B connectors and cables, and before long, Micro-USB connectors and cables as well. Both the Mini-USB and Micro-USB connectors were provided with five pins, with the additional pin used for identifying the type of device, either peripheral or host, they were plugged into (‘device ID’). The USB 2.0 specification also allowed two peripheral devices to communicate directly, instead of only via a PC host – a feature called USB On-The-Go or USB-OTG. The specification was also expanded with support for dedicated battery chargers, as well as allowing increased current flow in a PC-to-peripheral USB cable compared with the original limit of 500mA (or 100mA for unconfigured devices). These days, when a device has a mix of USB 1.1 & USB 2.0 ports, the USB 1.1 ports tend to be colour coded with white plastic, while the USB 2.0 ports use black plastic. released in November 2008, when the overall management of USB was transferred from the USB 3.0 Promoter Group to the USB Implementers Forum (USB-IF). USB 3.0 added yet another transfer mode: SuperSpeed, providing for a nominal data rate of 5.0Gb/s (625MB/s) in addition to the three existing transfer rates. Communication in SuperSpeed mode is in full-duplex, whereas in the three earlier modes, it is in halfduplex. USB 3.0 also introduced the UASP protocol, which provides generally faster transfer speeds than the Bulk-Only-Transfer (BOT) protocol provided by USB 1.X and USB 2.0. The USB 3.0 specification added a range of backward-compatible plugs, sockets and cables. The SuperSpeed plugs and sockets have a total of nine pins (4 + 5) and are identified with a distinct logo and an internal insulation layer coloured in blue, in contrast with the black or white used for USB 1.1 and USB 2.0 connectors. Low-speed and high-speed devices remain operational in USB 3.0, but SuperSpeed devices can take advantage of the increase in available current to between 150mA and 900mA. We’ll siliconchip.com.au USB 3.0 arrives The USB 3.0 specification was Australia’s electronics magazine June 2021  33 You can clearly see the five new contacts and blue colour of this USB 3.0 socket. The USB-IF certification logo for devices compatible with USB 3.0 or later specifications at 5Gbps. Thunderbolt 1/2 uses the same connector as Mini DisplayPort and replaced FireWire on Apple computers. It merges PCI Express and DisplayPort signals and provides DC power. Source: https://w.wiki/o26 Thunderbolt 3 uses cables with the USB Type-C plug but provides more functions than just carrying data; it also offers more power than USB (eg, for charging laptops) and can carry video and even PCI Express lanes. USB 4.0 essentially merges Thunderbolt’s features into the USB lineup. Source: https://w.wiki/o27 34 Silicon Chip describe the various USB connectors and cables in detail a bit later. In July 2013, the USB 3.1 specification was released, providing two variations on the USB 3.0 SuperSpeed mode: USB 3.1 Gen1, much the same as the original USB 3.0 specification, and USB 3.1 Gen2, which introduces a new SuperSpeed+ transfer mode. SuperSpeed+ doubles the maximum data signalling rate to 10Gb/s (1.25GB/s), while reducing the line encoding overhead to just 3% by changing the encoding protocol to 128b/132b. Then in September 2017, the USB 3.2 specification was released. This introduced two more SuperSpeed+ transfer modes, designed to take advantage of the 24 pins (2 x 12) on the newly released USB-C connectors. Although the two rows of 12 pins had been provided initially to allow the Type C connectors to be inserted either way around, the USB 3.2 specification uses them to provide multilane operation (using additional wires in the cable) to allow data transfer rates of 10Gb/s or 20Gb/s (2.5GB/s). When computers have both USB 2.0/3.0 and USB 3.1/3.2 ports, typically the USB 3.1/3.2 ports will be colour-coded teal or yellow, with the USB 3.0 ports remaining blue and the older ports having black (USB 2.0) or white (USB 1.1) plastic. Thunderbolt 1, 2 & 3 Back in late 2008, Apple introduced a miniaturised version of the DisplayPort audio-visual digital interface, dubbed Mini DisplayPort or MiniDP. This replaced the DVI port on most of Apple’s models like the MacBook, MacBook Air, MacBook Pro, iMac, Mac Mini and Mac Pro. The MiniDP port also started to appear on notebooks from Asus, Microsoft, MSI, Lenovo, Toshiba, HP, Dell and other makers. Then in early 2011, Intel and Apple announced their Thunderbolt hardware interface, which combined the functions of PCI Express and MiniDP and superseded FireWire (IEEE1394). Thunderbolt cables combine copper and fibre-optic transmission, with the copper wires generally used to convey power while the optical fibres convey high-speed data. They use the 20-pin MiniDP connector. In June 2013, Intel announced Thunderbolt 2, which used the same connectors as Thunderbolt 1 but doubled Australia’s electronics magazine the data rate to 20Gb/s (2.5GB/s) by combining the two 10Gb/s channels. The first consumer products featuring Thunderbolt 2 were the Asus Z87-Deluxe/Quad motherboard and Apple’s Retina MacBook Pro, both released in the latter half of 2013. Then in 2015, Intel announced Thunderbolt 3, which doubled the maximum data rate again to 40Gb/s (5GB/s) while also halving power consumption. Using copper cables and the 24-pin USB-C connectors which had been introduced in 2014, Thunderbolt 3 can incorporate USB Power Delivery and transfer up to 100W of power along with the high-speed data. Devices with Thunderbolt 3 ports became available in November 2015, including notebooks from Acer, Asus, Clevo, HP, Dell, Dell Alienware, Lenovo, MSI, Razer and Sony running Microsoft Windows, as well as motherboards from Lintes Technology. Then in October 2016, Apple announced the updated MacBook Pro, featuring two or four Thunderbolt 3 ports depending on the model. USB-C The USB Type-C or USB-C specification was finalised by the USB-IF in August 2014, and is primarily associated with the miniature 24-pin (2 x 12-pin) USB-C connectors. Initially, these connectors were used with USB 3.1 interfaces so they could be inserted into the sockets either way around (they are also significantly more robust than Type-B mini or micro plugs). But when Thunderbolt 3 arrived in 2015, they were used for that as well. And when USB 3.2 arrived in late 2017, they also gained SuperSpeed+ capability. But note that a device fitted with a USB-C connector does not necessarily implement USB, USB Power Delivery or any of the defined Alternate Modes. An Alternate Mode dedicates some of the physical wires in a USB-C 3.1 cable being used for direct device-to-host transmission of other data protocols, such as DisplayPort. The four high-speed lanes, two sideband pins and (for docked, detachable device and permanent cable applications only) two USB 2.0 data pins and one configuration pin can be used for Alternate Mode transmission. The modes are configured using vendordefined messages (VDMs) through the configuration channel. siliconchip.com.au USB4 A further use for the USB-C connectors was defined in August 2019, when the USB-IF released the USB4 specification. USB4 is based on the Thunderbolt 3 protocol. It supports 40Gb/s (5GB/s) data throughput, is compatible with Thunderbolt 3 and backwards-compatible with USB 3.2 and USB 2.0. The architecture defines a method to share a single high-speed link with multiple end devices dynamically, designed to optimise the transfer of data by type and application. Thunderbolt 4 Thunderbolt 4 was announced in January 2020 at CES (the Consumer Electronics Show), and the final specification was released in July 2020. The main improvements are support for USB4 protocols and data rates, a minimum bandwidth requirement of 32Gb/s for PCIe link, support for dual 4K (or one 8K) displays, and Intel VT-d-based direct memory access (DMA) protection to prevent physical DMA attacks. The maximum bandwidth remains at 40Gb/s, the same as Thunderbolt 3 and four times faster than USB 3.2 Gen2x1. Still, the minimum that vendors are required to implement has been doubled from the 16Gb/s previously allowed in the Thunderbolt 3 specification. USB Power Delivery (USB PD) In July 2012, the USB Promoter Group finalised a USB Power Delivery specification (USB PD rev.1), to permit uniformly powering or charging laptops, tablets, USB-powered disk drives and similarly higher-powered consumer electronics. It is a logical extension of existing European and Chinese mobile telephone charging standards. The USB PD rev.1 extension specifies using certified “PD aware” USB cables with standard USB Type A and Type B connectors, to deliver increased power (more than 7.5W) to devices with greater power demands. Devices can request higher currents and voltages from compliant hosts – up to 2A at 5V (10W), and optionally up to 3A or 5A at either 12V (36W or 60W) or 20V (60W or 100W). In all cases, both host-to-device and device-to-host configurations are supported. The power configuration siliconchip.com.au protocol uses a 24MHz Binary FSK (frequency-shift keying) transmission channel on the Vbus line. Revision 2.0 of the USB PD specification (USB PD Rev.2.0) was released in August 2014 as part of the USB 3.1 specification. It covers the use of USB-C cables and connectors with four power/ ground pairs and a separate configuration channel, using DC-coupled low-frequency BMC (Biphase Mark Code or Differential Manchester) data encoding to reduce the possibility of RF interference. Since then, there have been further revisions of USB PD Rev.2.0. In March 2016, version 1.2 was released, creating new USB PD Power Rules which define four nominal voltage levels (5V, 9V, 12V and 20V) and output power levels ranging from 0.5W to 100W. Then in January 2017, the USB-IF released USB PD revision 3.0, which defines a programmable power supply (PPS) protocol that allows control of Vbus power in 20mV steps, to facilitate both constant current (CC) and constant voltage (CV) battery charging. This was followed up in January 2018 with the release of a “Certified USB Fast Charger” logo, for chargers that use the USB PD 3.0 programmable power supply protocol. USB connectors & cables There are now so many different USB connectors in use that it isn’t feasible to discuss them all in detail. But we have prepared some information to help you recognise the most common types of connectors and cables. As mentioned earlier, Fig.1 shows the ones you’re probably most familiar with: the Type-A socket and plug and the Type-B socket and plug. These are the original four-pin USB connectors, with the Type-A connectors intended to be used at the host/PC end, and the Type-B connectors at the peripheral/ external device end of the USB cable linking the two. The table below them shows the names usually given to the four pins, the nominated colour of the insulation for each wire, and the description of its function. Note that in a USB cable, the D- and D+ wires are a ‘twisted pair’, to reduce the risk of electromagnetic interference (EMI) – both in terms of reducing emissions and avoiding problems with EMI pickup. Australia’s electronics magazine The logo used on the fastest USB devices available today. USB Type-C plugs have more contacts, and they are arranged symmetrically so that the plug can be inserted either way around and it will still work. A closer view of the USB Type-C plug clearly showing all 24 contacts. This design appears on plugs and near sockets that support the fastest 40Gbps speed of USB4. A USB-IF certification logo for a device which supports USB-PD at up to 45W. This involves supplying both higher voltages and currents than the normal 5V/500mA at the request of the device. June 2021  35 These miniaturised versions of the original Type-B connector are much more suitable for smaller devices like mobile phones and tablets. They add a fifth Device ID pin and importantly, the micro Type-B plug is designed to wear out rather than the socket, so you just have to replace the cable if the plug wears out, instead of the socket or the whole device. The new plugs and sockets of USB 3.0/4.0 add five new contacts to carry higher bandwidth signals. They are designed so that USB 1.0-2.0 devices can still be plugged in and operate normally over the same four pins they have always used. USB 3.0/4.0 devices can plug into an earlier style Type-A socket, but the extra pins will not make contact, so communications occur at a slower speed. 36 Silicon Chip Fig.2 shows the Mini-USB and Micro-USB connectors, still used for connections to many compact devices like tablets and PDAs (personal digital assistants), smartphone and digital cameras. Although there were Mini Type-A plugs and sockets when USB 2.0 and Mini-USB were introduced in 2000, they were officially ‘deprecated’ in 2007 along with a Mini Type-AB socket. That is why you won’t come across many of them nowadays, and as a result, we haven’t shown them. The same applies to Micro Type-A plugs and sockets. It’s worth noting that although the functional part of Micro-USB plugs is similar in width to Mini-USB plugs, they are approximately half their thickness. Despite this, they are rated for at least 10,000 connect-disconnect cycles, which is significantly more than the Mini-USB plugs. Fig.3 shows the details of the USB 3.0 SuperSpeed connectors that were introduced in 2008. They are essentially a modified version of the original Type-A and Type-B connectors, with five pins added to cope with the SuperSpeed requirements, while keeping backwards compatibility with USB 1.X and USB 2.X. Perhaps the most obvious difference at first sight (especially with the Type-A connectors) is the blue colour of the plastic insulation inside the connectors, compared with the white or black insulation inside the earlier connectors. Inside the Type-A connectors, the additional five contacts are located a short distance away from the first four, parallel with them and spaced slightly further apart. On the other hand, in the Type-B connectors, the top of the functional part of the connector is extended upwards by about 3mm, with all of the additional five contacts mounted closely together in the narrower upper section. This allows the Type-B socket to accept older Type-B plugs, but of course, the SuperSpeed Type-B plug is not compatible with the older Type-B socket. As before, the table below the connector diagrams shows the name and significance of each of the nine contacts. Note that contacts 5, 6, 8 and 9 have a different name for the A connector and the B connector. Australia’s electronics magazine Fig.4 shows the details of the distinctive USB 3.0 SuperSpeed Micro-B connectors, in which the additional five contacts are alongside the original five contacts in their ‘chamfered rectangle’ and inline with them, but in a separate group. The table below shows the names and significance of all ten contacts. Finally, we come to the 24-pin USB-C connectors. Fig.5 shows the details of the USB-C socket and plug, at twice actual size for clarity. The contacts are in two rows of 12 and are labelled A1-A12 and B1-B12. These were originally just duplicates of each other, to allow the plug to be introduced to the socket either way around. But nowadays, to cope with the many expanded applications for USB-C, most of the contacts have different functions, shown in the table below the plug and socket. Contacts A1, B12, B1 and A12 are now all devoted to ‘power ground’, while A4, B9, B4 and A9 are all devoted to Vbus power, to provide the added power capability for USB PD. B5 is also dedicated to Vconn, to supply power for powered cables. The other thing to note about the USB-C connectors is that in addition to the original USB differential data pair (allocated to contacts A6 and A7), they also provide for four pairs of shielded differential pairs for SuperSpeed, SuperSpeed+, USB4, Thunderbolt 3 and Thunderbolt 4 high-bandwidth data transmission. These are allocated to contacts A2 and A3, B11 and B10, B2 and B3, A11 and A10. There is also a configuration line assigned to A5, and finally, two ‘sideband’ lines allocated to contacts A8 and B8. How it has grown... Clearly, USB has grown dramatically over the past 26 years, both in terms of performance and functions. It has changed from a system intended to simplify the connection of basic devices like keyboards and mouses, to a system with at least nine different types of connector – some with as many as 24 contacts – and nine different data transfer speeds, ranging from the original 1.5Mb/s right up to the 40Gb/s of SuperSpeed+ and Thunderbolt 3. The ability to provide power to devices via a USB cable has also grown significantly. From the modest 100mA siliconchip.com.au Rather than add extra contacts internally, as was done with the Type-A plugs and sockets for USB 3.0, the USB 3.0 full-size Type-B expands the plug shroud. But as the lower square section is identical to the earlier USB 1.0/1.1 & USB 2.0 Type-B plug, older cables can still be used in devices with sockets that accept this newer plug. Similarly, the USB 3.0 micro Type-B plug adds a whole new section to one side with the five extra contacts. Once again, sockets are compatible with older (4-contact) plugs, but not the other way around. RUN LONGER GO FURTHER – Upgrade your dead or dying batteries EBIKE? SEGWAY? MOBILity BUGGY? GOLF CART? ESCOOTER? The SuperSpeed micro connectors are wider, adding a separate section with the five added pins alongside the original socket. Therefore, these sockets are also backwardscompatible with older cables and hosts. The USB Type-C plug and socket is similar in size to the micro-B plug and socket, but is capable of much higher data speeds and greater power delivery. It’s also reversible and considerably more robust than the micro-B. or 500mA at 5V available via USB 1.X and USB 2.X, USB PD now allows devices to request one of four different voltage levels (5V, 9V, 12V or 20V), at current levels up to 5A. This opens up the ability to run much higher-powered peripherals, as well as allowing many more battery-powered devices like laptops, tablets and mobile phones to have their batteries fast charged via a USB cable. These days, you can run a single USB Type-C cable between a portable computer and a monitor, and not only will it charge the computer (from the monitor’s internal power supply), it will also carry high-resolution video signals and even connect a keyboard, mouse, printer, fast storage devices and more. We wonder whether the originators of USB would have even considered that possible back in 1994, when they set it all in motion! siliconchip.com.au Further reading • • • • • • • USB standard: https://w.wiki/Usb USB 3.0: https://w.wiki/ntm USB4: https://w.wiki/ntn USB Type-C: https://w.wiki/nto Thunderbolt: https://w.wiki/ntp USB PD: https://w.wiki/ntq USB Implementers Forum: www. usb.org/ SC Australia’s electronics magazine Premier Batteries can recell and/or custom manufacture Lithium Ion batteries for Segways, Ebikes, Electric Golf Carts, Scooters and Mobility Buggies –– often with increased capacity and range etc. Quality cells: Sanyo, Samsung or LG and batteries are Fully Guaranteed PREMIER BATTERIES High quality batteries for all professional applications SUPPLIERS OF QUALITY BATTERIES FOR OVER 30 YEARS email: info.premierbatteries.com.au Web: www.premierbatteries.com.au June 2021  37