Slashdot Mirror
USB SuperSpeed Power Spec To Leap From 10W To 100W
Lucas123 writes "While news stories have focused on the upcoming jump from 5Gbps to 10Gbps for USB SuperSpeed, less talked about has been the fact that it will also increase charging capabilities from 10W to 100W, meaning you'll be able to charge your laptop, monitor, even a television using a USB cord. Along with USB, the Thunderbolt peripheral interconnect will also be doubling it throughput thanks to a new controller chip, in its case from 10Gbps to 20Gbps. As with USB SuperSpeed, Thunderbolt's bandwidth increase is considered an evolutionary step, but the power transfer increase is being considered revolutionary, according to Jeff Ravencraft, president of the USB Implementers Forum (USB-IF). 'This is going to change the way computers, peripheral devices and even HDTVs will not only consume but deliver power,' Ravencraft said. 'You can have an HDTV with a USB hub built into it where not only can you exchange data and audio/video, but you can charge all your devices from it.'"
fairly robust fibre optic solutions to date that carry data and are far more energy efficient. im confused as to why our peripherals dont use them
Good people go to bed earlier.
I have an iphone 5 and like newer samsungs and ipads these want to draw 2.1 amps from USB, which is a no-no for standard USB. THere are a number of USB hubs that pretend that they are apple/samsung compatible, promising 2.1 amps. But what they don't tell you is that you can't have 2.1 amps if the hub is connected to a computer. It will only act as a USB high current charger when it is incapable of making a serial connection. It's either a serial port or a high current charger but not both.
I'm guessing this is because a lot of devices expect their current overload regulation to come from the USB hub which is limited to 0.5 amps by spec.
Will this superspeed use the same USB plug and thus have the same limit of either being a charger or a USB port, or will it do both at the same time.
Some drink at the fountain of knowledge. Others just gargle.
I'm not a fan of a "data" cable that can kill me.
So if you have 4 USB SS ports on a motherboard that motherboard is going to have to be able to supply 400W @ 5V? You can't be serious. We'll need dedicated power connections on the motherboard just to supply this.
The example of using a TV to power multiple devices raises the same concerns. Now the TV power supply will be much more complicated. Rather than power just the 60-70W the TV draws it needs to have a power supply that could supply 100's of extra watts?
The only application I see for this is to use 100W USB SS ports on walls for a common household DC standard interface. That could be interesting, but integrating it into devices is not simple. It adds levels of complexities to the devices that will need to supply the power.
Nothing i have to say is worth saying.
100W Hot swappable. I really don't think the chinese are up to it. I'll have to double check the specs. (Will they)?
Repost here as I accidentally posted as AC. One of the main problems with FireWire was that it required expensive cables due to the high quality cables needed to carry the bulk power. With this spec change and the data model for SS USB, have we now got a high tech FireWire-- with all of the disadvantages and none of the advantages (I.e. daisy chaining. Guarantees about latency etc).
actually the power supply would need to have an extra 450 watts since you NEVER design to full rating you at the least design for the loads to be at most 90% of Full (prevents a fire hazard).
The point is if the spec says XXX watts are available then XXX+Y watts had better be available (nasal demons are allowed for drawing under spec).
Any person using FTFY or editing my postings agrees to a US$50.00 charge
the current micro-USB connector kinda sucks. if we're going O(100x) more watts, maybe we should take the opportunity to do a better connector, too.
symmetric would be nice, and less prone to jamming, misalignment and torquing.
Firewire goes to 30GB/s and 45 watts (30v @ 1.5 amps) and you can daisy chain it. Seems like a better idea than inventing a non-backward compatible serial port and pretending it is somehow related to USBs of yore.
Do you have a source on the non-backwards compatibility thing? Because the USB spec release[PDF warning] for the new USB SuperSpeed states it will be.
I should add that the newest FireWire specs only go up to 800mb/s, so also a source on that would be nice.
"None can love freedom heartily, but good men; the rest love not freedom, but license." --John Milton
I eagerly await our USB toasters, arc welders and cutting lasers.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
This "100watt USB!!!" nonsense has been floating around for a while, and it just never seems to get any better.
Uncertainty is Bad. 100watts is a lot of power. Your laptop's brick is almost certainly specced for less than that. Even a desktop PSU will likely be 250-350, outside of gamers and workstations(and often the upper end of the range is...optimistic... at best). Now, if we have this '100watt USB', what are devices going to do? is your next laptop going to ship with a 265 watt brick, so that it has the same 65 watts for itself as your current one does, and can handle both its ports being used? Is it going to ship with exactly the same brick and simply brown out the USB port at some unpredictable power level?(extra credit awarded if that unpredictable level depends on whether the battery is charging or not, and the current CPU load...) If "100watts" is actually "anywhere between ~15 watts and 100watts, largely unpredictable to the consumer", what are peripheral manufacturers going to target? What good is theoretical capacity that you can't actually use because a nontrival-but-hard-to-predict percentage of your customers can't actually deliver it?
Bus power is nice because it reduces cabling and complexity. However, if it isn't dependable, you can't rely on it, so you have to fall back on designs that pretend it isn't available. Now you have more expensive USB ports(in some devices) and wall warts or PSUs for your higher power peripherals! What a win!
This isn't to say that any increase in bus power is bad(given USB's use cases, 'enough power to spin up a 2.5 inch HDD' or 'enough power to charge a smartphone' are pretty useful things. However, you can't just keep pushing the ceiling without limit: the wider the uncertainty, the greater the costs(for devices that actually engineer to spec and include the capability to support the top of the range) and the greater the limits and confusion(for devices that target more realistic real-world output values, and for the poor bastards who think that 'USB' means 'works when plugged into my USB port').
And require 16 or 14 gauge wire, that will be nice and convenient to carry around. I can't see this adoption being too widespread, only special use cases.
USB 2.0 section 7.2.1.2.1 says 5 A max as in when you hit it the protection circuit kicks in and limits or shuts down current.
To actually pull 5A means the required protection circuit would need to trip above 5A to be useful which violates this section.
The more reality based problem is 28 gauge wire over 10-16 FT of cable carrying 5 amps is really stretching it...the voltage losses in that scenario will significantly pull down the actual watts being delivered into heating the wire.
At 10 ft the voltage drop when pulling 5 amps is ~6 volts. At 16 ft the drop is a staggering ~10 volts.
Unless there is a whole lot of intelligence to probe wire losses as part of the power specification and take the wire itself into consideration when calculating maximum current availability 100 watts over only 20 volts is really stretching it.
While I'm not impressed by USB's mutations over the years, Firewire had the major drawback that(at least in practice, not sure if the paper demanded otherwise) there was a very, very heavy emphasis on 'up to' when it came to how much power could be delivered.
A small minority of actually-well-built workstations and the like wouldn't shrug at providing full specced power. More or less ordinary PCs usually had a floppy or molex connector to supplement PCI bus power; but didn't spring for a DC-DC converter, so (since 30v isn't readily available anywhere on the DC side of an ATX PSU) you generally got 12v, albeit at a decent amperage. Laptops? In practice, "firewire" pretty much meant 'whatever Apple did on the last couple of models of ibook and powerbook; because all the PCs omitted the power pins entirely for "i.link" or similar, which usually boiled down to ~19v, if on adapter, 12-ish if on battery.
The nominal maximum was certainly fairly spacious; but a powered firewire peripheral was essentially always on the hook for a DC-DC converter, and had to deal gracefully with(or simply refuse to work with, ideally in a documented way) substantially inferior power supplies from many devices.
5v 500ma was always pitiful; but (by virtue of being so pathetic) most devices actually did as well or better than they claimed to, and lots of peripherals could get away with only the cheapest of designs for handling bus power.
That's my bet for why "100watt USB" will suck. Sure, it'll be cute and all that POS hardware vendors can now have USB printers and things that are 'standards compliant' and will actually work if purchased 100% from approved vendors and plugged in just right; but everyone else will have wildly unpredictable actual power levels.
TOSlink is sort of a weird one because it was optical; but usually over very short plastic runs, and at a data rate so low that even fairly pitiful copper has no trouble with it(which is why it is now commonly replaced by, or lives along side with, an RCA connector providing the same output in a copper flavor). I'm sure that there is some reason why optical was dragged in in the first place; but it's always a bit jarring to see.
This 100w power standard is pretty stupid, though. We're talking power levels where fires will definitely be possible from damaged USB cables.
As opposed to all of the current laptop chargers, AC power cords, DC converter bricks, etc out there now?
From TFA:
"So with this new specification, you can go from very small devices with 5 volts, 2 amps or 10 watts -- where USB starts -- up to 20 volts 5 amps and 100 watts,"
It's no worse than a current laptop charger (bit better, actually, MB chargers are only 16.5v).
The main benefit of TOSlink is avoiding ground loops in audio systems. This is especially important if you have a long run between ends of the building with a significant resistance in the building ground system between them.
My USB connector on my Samsung Droid Charge is on the wiggly-loose fritz. If I plug in a 100 watt cord and wiggle it to get the connection to work, it's not gonna burst into flames is it?
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
I want my USB controlled and powered Easy Bake oven.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
The SuperSpeed spec requires that devices specifically request the increase in power, in order to remain backwards compatible with older USB specs. In other words, you'd have to wiggle that cable pretty damn particularly in order for it to happen.
+1 IDisagreeSoHeMustBeATrollOrAnAstroturferOrAShill
Thunderbolt cables have part of the interface electronics physically in the connector body - that's why they cost so much. It also means you can swap a thunderbolt copper cable for a thunderbolt fiber cable without having to worry about the equipment at the ends having an exotic fiber interface.
I don't know if you can even get a thunderbolt fiber cable yet. They don't go any faster than copper, but they do go longer, which could be handy in a few niche applications. I'm thinking supercomputer and cluster interconnects. Could be cheaper than infiniband, and lower latency than ten-gig ethernet.
Apple fucked Firewire which is why its dead now. For anybody that wondered why Firewire never went anywhere, even at a time when it could just curbstomp USB, the reason why was simple in that Apple charged a buck a port to the OEMs and peripheral manufacturers while Intel practically gave away USB. When you are talking about razor thin margins every dime counts so they went with the "good enough and cheap" over the "fast and expensive" solution.
So I'm sorry but Firewire is as dead as floppies, USB won ages ago and isn't going anywhere.
ACs don't waste your time replying, your posts are never seen by me.
This 100w power standard is pretty stupid, though. We're talking power levels where fires will definitely be possible from damaged USB cables.
As opposed to all of the current laptop chargers, AC power cords, DC converter bricks, etc out there now?
Those cables are built for that power, do you want to carry around usb cables that thick for every device that uses usb? Unless there is a way for the chipset to identify the cable (no high power if the cable is a type that's thinner than a certain size) it could be a risk.
You can't trust users to decide sensibly if the thin cable would be safe to charge your laptop if the connector is the same as the thick cable that came with the laptop. Lots of people reason that if it fits, it should work.
The better question to ask, is how are manufacturers ever going to be able to offer 100w from a USB port. Do we really want our computer PSU's to have to be able to handle another 20A on the 5v rail just to be able to offer a single 100w capable port?
What about your TV or anything else that ghas a USB port? One of the reasons USB is so popular is because you didn't need to re-engineer your entire power supply and all it's rails.
Ever tried putting 5v @ 20A though a PCB trace?
Normal people worry me!
Ever tried putting 5v @ 20A though a PCB trace?
Yes, but I still pretend I didn't.
Never speak of it again.
"I've got more toys than Teruhisa Kitahara."
Infiniband costs a fortune. That's fine if you've big money behind you. Otherwise, you go with ethernet. Thunderbolt could fill the space between: Faster than ethernet (and lower latency), yet still cheaper than infiniband. It could be just the thing for medium-scale computing clusters, the things used by smallish universities and companies that can't afford to spend millions of dollars on a top-of-the-range supercomputer. It'll easily match ten-gig ethernet for performance - all you need is a switch, and if demand is there someone will manufacture one. The protocol is basically just an external PCIe lane, there's nothing to stop you from sending frames over it with an address. Or you could just use an unswitched topology - loop, hypercube, torus. Something that runs entirely on point-to-point links.
When will this myth ever die? The USB standard appeared in 1996, the USB iMac appeared in 1998 and it wasn't even the first with USB 1.1. Not to mention that the Mac marketshare was below 5%, how exactly did they manage to "push" USB. You want to talk about a standard that Apple really pushed and the PC largely ignored, talk about Firewire. How well did that go, exactly?
TFA mentions that it will switch up to 20V, which is similar to what most laptop chargers use to reduce the amps required down to 5. Still fairly hefty but workable.
The device will have to request 100W of power, and the host is then free to refuse if it can't handle it. I expect hosts will come with a sticker that says something like "50W USB power" so you know that your 35W laptop will be fine with it, but your 100W... er... pipe soldering iron won't.
As an aside it's a shame USB didn't start out using 3.3V for power instead of 5V. The signaling is all 3.3V so it pretty much mandates some kind of voltage regulator or level shifting hardware in the device. Particularly in the early days that added cost. I guess there was still a lot of 5V stuff they wanted to be compatible with back then.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Do you remember the computing landscape then? I had a PC back in 1996 (possibly 1997) with two USB ports. Windows NT 4 and Windows 95 didn't support them, there were no peripherals available for them, and they were unused. Windows 95 OEM Service Release 2.1 was released in August 1997 and added support for USB. Every PC came with a PS/2 keyboard and mouse (occasionally an RS-232 mouse, but they were mostly phased out by then, 5-pin DIN keyboards were still common though). Every Mac came with ADB input devices. PC printers were all parallel. USB mice were rare and expensive. Ditto keyboards. I didn't even see a USB printer
Then, in 1998, Apple introduced the iMac with no ADB ports. In fact, no legacy interfaces at all. Any peripheral maker that wanted to sell to Apple customers (and ride on the back of the massive ad campaign for the iMac) sold USB versions of their products. If you walked around a computer store around 1999 / 2000, then every USB peripheral was using translucent plastic to go with the iMac. When I bought a new computer in 2000, it still came with PS/2 ports, and it still came with a PS/2 keyboard and mouse, but when I bought a replacement it came with both PS/2 and USB support.
Without the critical mass of people who could only use USB devices, there would have been far less incentive for manufacturers to start shipping USB peripherals. Why add USB circuitry when everyone has PS/2 anyway? Because that gives you access to the 5% of the market that's buying a Mac for a relatively small extra cost.
I am TheRaven on Soylent News