USB powers up to the tune of 100W, new spec in the works

By on August 11, 2011, 10:30 AM

Apple and Intel may be pushing Light Peak (aka Thunderbolt) technology forward as a true next-gen USB replacement, but that doesn't mean the latter will die so easily. The USB 3.0 Promoter Group is working on a new specification that will have USB 2.0 and 3.0 connections delivering up to 100 watts of power.

You may ask, how is this useful?

We could virtually power anything using a USB port. We already charge devices like smartphones and tablets using USB, and that's possible because the 4.5 watts delivered through current cables is sufficient. Now think about charging your laptop using USB, or more to the point, solely relying on USB connectivity to power your laptop at all times. Transmitting data simultaneously is part of the specification as well.

Key features of USB Power Delivery include:

  • Compatibillity with existing cables and connectors
  • Enables voltage and current values to be negotiated over the USB power pins
  • Enables higher voltage and current in order to deliver power up to 100W
  • Switchable source of power delivery without changing cable direction
  • Coexists with USB Battery Charging 1.2 and works equally well with both USB 2.0 and USB 3.0

A demonstration of the technology is expected by the end of the year, with a final revision of the specification to be delivered to the USB Implementers Forum (USB-IF) for publication and establishment sometime in 2012. As ExtremeTech likes to point out, just as Apple’s exclusive rights over Thunderbolt expire in spring 2012.




User Comments: 39

Got something to say? Post a comment
MilwaukeeMike said:

Light peak and Thunderbolt? Is there no end to the relentless assault of cheesy version and product code names? I'm gonna need to start a spread sheet to keep track of all this Gulftown, Sandy Bridge, Froyo, Honeycomb, Gingerbread, Clarksdale, Haswell, Ice Cream Sandwich (?!).

What ever happened to good old fasioned 3.0.

Didn't Apple's last attempt at this fail miserably? Wasn't it called FireWire or something?

Guest said:

Next thing you see in the newspaper is people getting electrocuted by a USB cable ;)

red1776 red1776, Omnipotent Ruler of the Universe, said:

Next thing you see in the newspaper is people getting electrocuted by a USB cable

8 amps running through the USB ports? maybe so.:p

Win7Dev said:

This would be great for external hard drives. Think about all external hard drives having no power supply necessary. Yes I know there are a few usb power ones, but those usually are slow and don't work very well.

Staff
Julio Franco Julio Franco, TechSpot Editor, said:

USB speakers (the quality kind) could also benefit on a similar fashion, a single cable for everything.

madboyv1, TechSpot Paladin, said:

to win7dev's comment: yes, that would be great for full size drives, as well as full speed external burners. Another idea would be to power full size monitors (well, 19-21" monitors perhaps or perhaps piggyback two USB power connections). It'd be one less power cable and maybe video through USB as well.

SNGX1275 SNGX1275, TS Forces Special, said:

PSU manufacturers are gonna love this.

Guest said:

Just was I was thinking. But perhaps the possibilities with powering everything through USB makes up for the additional cost of a 1000+ watt PSU.

Kibaruk Kibaruk, TechSpot Paladin, said:

Next thing in the market "KIDS DONT STICK YOUR FINGERS IN THIS USB" or maybe one that says "DANGER DONT GET CLOSE TO IT IF YOU OWN A PEACEMAKER"

SNGX1275 SNGX1275, TS Forces Special, said:

lol, perhaps you meant 'pacemaker'

red1776 red1776, Omnipotent Ruler of the Universe, said:

PSU manufacturers are gonna love this.

And you jump start your car with it

dividebyzero dividebyzero, trainee n00b, said:

Just was I was thinking. But perhaps the possibilities with powering everything through USB makes up for the additional cost of a 1000+ watt PSU.

And possibly the one after that....

All the people who think this a wonnnnnnnnnderful idea...where do you think all this extra power is being drawn from?

Hint: The southbridge slash I/O hub on the motherboard.

So, a few questions...

Bearing in mind that the standard fit-out for USB on a current board usually runs to 12-14 ports, what provisions are USB-IF making to prevent over-current draw?

Are all (any?) USB/mini-USB cables specced for continuous 8.3A/100w operation ?

At least I should see some more custom eventuating due to adventures in hot-plugging. As it stands, it doesn't take much for a sub-5w connection to take down an onboard controller.

Staff
Rick Rick, TechSpot Staff, said:

Hey, I think this is great news. However, I've seen a number of computers with shorted out USB ports because of vandalistic children shoving coins, paper clips etc.. into the USB ports to purposefully damage them.

Imagine their surprise on one of these newer PCs...

dividebyzero dividebyzero, trainee n00b, said:

Hey, I think this is great news. However, I've seen a number of computers with shorted out USB ports because of vandalistic children shoving coins, paper clips etc.. into the USB ports to purposefully damage them.

Imagine their surprise on one of these newer PCs...

On the plus side, they'll only do it once. Your next PC should be juvenile vandalism free!

red1776 red1776, Omnipotent Ruler of the Universe, said:

I wasn't even crazy about my "UD7's 3xUSB power" feature. So to ask a dumb question...won't this overtax the PCB/Tracer design of the boards as well as the SB hub?

jobeard jobeard, TS Ambassador, said:

sure would kill laptop batteries - - and add a good load to OEM PSUs as well (so many are marginal)

dividebyzero dividebyzero, trainee n00b, said:

I wasn't even crazy about my "UD7's 3xUSB power" feature. So to ask a dumb question...won't this overtax the PCB/Tracer design of the boards as well as the SB hub?

Yup. That's kind of where my thoughts were.

While not every USB port would be populated by a high amp/wattage connection, I would think even a minimal number would heat up a southbridge -even if it could accomodate the pass through.

Now, bear in mind that this is supposed to be compatible with standard USB2 and 3.0 how many average users are :

1. Going to aware of maximum current draw for the system, and

2. Owning a bog-standard PSU (as jobeard pointed out), or a capable PSU but connected to a board using a 4-pin 12v auxillary power connector.

3. Going to make the differentiation between standard and high current USB devices.

Cue a whole new forum hardware section

"Hi TS, first time poster. I have a problem with my computer. Every time I start it up- it shuts down. There's a big "Whump" sound from my speakers and thats it. Help"

spydercanopus spydercanopus said:

I'm more interested in power over ethernet for cameras, printers, scanners, etc... Nothing worse for an IT guy than having to call an electrician to install an outlet for something.

red1776 red1776, Omnipotent Ruler of the Universe, said:

Yup. That's kind of where my thoughts were.

While not every USB port would be populated by a high amp/wattage connection, I would think even a minimal number would heat up a southbridge -even if it could accomodate the pass through.

Now, bear in mind that this is supposed to be compatible with standard USB2 and 3.0 how many average users are :

1. Going to aware of maximum current draw for the system, and

2. Own a bog-standard PSU (as jobeard pointed out), or a capable PSU but connected to a board using a 4-pin 12v auxillary power connector.

3. Make the differentiation between standard and high current USB devices.

Cue a whole new forum hardware section

"Hi TS, first time poster. I have a problem with my computer. Every time I start it up- it shuts down. There's a big "Whump" sound from my speakers and that it. Help"

Thats what I think. Add to this that current USB cables can't possibly be up to the task of 8A/110w current, and the additional VRM/overvolt circuitry needed is going to add how much to the cost of the MB?

Guest said:

Everyone that is quoting 8A or so, have actually read the article? It pretty clearly states that higher voltages will be supported.

"Enables voltage and current values to be negotiated over the USB power pins

Enables higher voltage and current in order to deliver power up to 100W "

So potentially it could be upwards of 12 to 48 volts.

As far as the circuitry goes, leave that up to the engineers, as most of your guys are just trolling and have no idea.

dividebyzero dividebyzero, trainee n00b, said:

@Guest 21

WTF are you on?

How many consumer computer peripherals are designed to run off 24 and 48 volts ?

While the spec more than likely calls for voltage negotiation- worst case scenario is 100w/5v = 20 Amps (!). At 12v (which would negate most consumer peripherals) 100w/12v = 8.3A (and 100w/24v = 4.2A, 100w/48v= 2.1A). So regardless of whether the source (mainboard) could supply the power, we're still left with the glaring fact that the USB-IF are saying that a standard USB cable is 20 Amp compliant....As with you post, I'll rate that "Whatever".

Current USB spec is 5v 5% output @ 500-900mA....that's 2.5 -4.5 watts just in case math isn't a strong suit.

Guest said:

Where are the USB-IF saying that a USB cable is or is going to be 20A capable?

And where did i say that a device was going to be 24 or 48 volts? The voltage can be boosted for transmission and bucked at the device to suit. That would be the only way for a two wire configuration. Since this wont be the case, the spec will most likely included mixed voltages and multiple cores for power.

Great work on using Ohms law, very impressive. Work out the size of cable that is required for 20A and the maximum length, max 5% volt drop, and the size of the connection. How practical would this be, not very. So rather than speculating and posting rubbish, just leave it to the pros kid.

dividebyzero dividebyzero, trainee n00b, said:

And where did i say that a device was going to be 24 or 48 volts? The voltage can be boosted for transmission and bucked at the device to suit. That would be the only way for a two wire configuration. Since this wont be the case...

Quite the scientist...refuting your own argument. Sweet, saves time.

...the spec will most likely included mixed voltages and multiple cores for power.

Bearing in mind the 100 watt figure that the USB-IF have used, what "mixed voltages" were you expecting?* And using Ohm's Law, what resulting current are you likely to get?*

Now, getting away from current draw, most of the thrust of the discussion here-before it turned into the Guestroll-a-thon was centred around the 100 watt draw per port, and how it would impact on current motherboards, PSU's and systems as a whole since the USB-IF are touting "compatible with existing cables and connectors" (pdf)

* Not rhetorical.

Feel free to add your insights on where these multiples of 100 watts will be transmitted using existing consumer mainboards and I/O hubs....or carry on doing the troll thing and pick at an incidental facet of the discussion (already covered I believe in my previous post regards "voltage negotiation" while ignoring the matter at hand...whichever you feel more commensurate with your skill level...kid

Guest said:

It seems like some will just have to wait till the full spec is released to further better there understanding on this topic.

petert said:

The main danger for electrocution are the high voltages and not the high currents. There are people that can feel the shock of an applied voltage of as little as 24 V (also there are some people that can't feel the shock even at 220 V due to their physiological nature). I guess that the voltages should never exceed the 19 V value (which is required by some laptops to be charged) and to stay on the safe side.

Guest said:

Petert - I would rather get hit by a high voltage than a high current, example I would rather get hit by a 10,000 volt taser than grab both terminals on a 12 volt car battery....

Guest said:

lol... I would rather touch a 12volt car battery myself. Given that the body has a resistance of kOhms, nothing would happen.

Back to the books champ

jobeard jobeard, TS Ambassador, said:

no one has mentioned that cable compatibility means this design will attempt to put

100w thru 22 (or was it 24) gauge wire - - brain dead on arrival fellas!

this AWG table shows

that 22 guage can only handle 0.92 amps. With max current per USB connection at

500ma, that's 6 watts (@12v). To get to 100W @12v, you need to deliver 8.3A and you get to kiss the 22guage cable goodbye.

So as today is not April 1st and assuming that figures don't lie and the specification intends to actually deliver a product,

we need to await that delivery and then evaluate the product as delivered.

@Guest: The fine guys and gals that participate here have backgrounds that include actual work experience in far and highly varied fields,

including Electrical Engineering. Sit back and enjoy the group. I might be a bold step on your part to actually become a member rather than continuing to harp under the anonymity cloaking of guest. No one will notice and you get to start with a clean slate

If I had the choice of options, I would disable all posting by guests, not just this specific but on general principle.

SNGX1275 SNGX1275, TS Forces Special, said:

no one has mentioned that cable compatibility means this design will attempt to put

100w thru 22 (or was it 24) gauge wire - - brain dead on arrival fellas!

Well not in those specific terms, but DBZ and red had a few posts with that concept.

Burty117 Burty117, TechSpot Chancellor, said:

Is it me or are all the guest posts in this thread complete rubbish?

you can have 14,000 volts sent through your body as long as the current is pretty much non existant.

you cannot have 24v through your body if the current is high as it will litterally cook your insides.

Why is that so hard to grasp? And before anyone goes "you cannot have 14,000 volts sent through your body" yes you can, I can't quite remember where it was but back at school years ago we went on a science school trip and a girl in my class was chosen to come up on stage when they were doing a presentation, they made her hold these pieces of metal and then they wound some kind of device, her hair went a bit freezy and was xtremely attracted to baloons but appart from that she didn't even realise that 14,000v was travelling through her body, and the reason it all worked was because the current was pretty much not there!

jobeard jobeard, TS Ambassador, said:

Voltage doesn't go thru your body, current does. That's why a taser is not (normally) leathal.

When working on a car, have you ever been shocked by a bad spark plug cable?

That can be upwards of 20kv, but the current is minuscule.

Fatal shocks are as little as 5ma; see this wiki

@SNGX1275; with some 30+ posts, I didn't see those comments :o

Guest said:

@ guest under my post of "Petert - I would rather get hit by a high voltage than a high current, example I would rather get hit by a 10,000 volt taser than grab both terminals on a 12 volt car battery...."

I have a background in electronics an will swear by my statement, and I will back it up with a rather graphic mental image of a less than careful co-worker that managed to accidentally complete the circuit on a car battery by means of an arm, and have his wedding ring fused into his skin. Give me a fork and a mains socket any day!

Oh and registered members calling all guest posts pointless, wrong and useless; superiority complex much? Stick it up your arses and your fake forum names.

jobeard jobeard, TS Ambassador, said:

From a 12v source 5ma IS possible e=ir ; e/i = ? 12/0.005 = 2400 ohms.

entirely possible as the conductivity of skin varies inversely to wetness

(ie the wetter you get, the lower the resistance )

PS: jobeard is real; J. O. Beard (if you read the signature).

Using real names reminds one to be carefull with what you say online.

Nom de plume and ghost writers (imo) denigrate themselves due to some

ulterior motive or some political necessity. As the kids are starting to say "MAN UP"

dividebyzero dividebyzero, trainee n00b, said:

Oh and registered members calling all guest posts pointless, wrong and useless; superiority complex much? Stick it up your arses and your fake forum names.

Now, now...don't be going all butthurt on us.

While there may be some members here who pin every Guest post with the stupid tag, I think you'll find that the prevailing sentiment here (much like the ever diminishing number of other sites that allow non-registration posting) is that for the occasional reasoned statement, it really isn't worth wading through the dross of "Guests" in the form of company shills, PR shills, drive-by kiddies, rampant fanboys, forum members from other sites spamming the threads, and generic internet trolls.

Think that's an overstatement ? Be my Guest and use the search function to look for posts using the keywords (in alphabetical order) AMD, Apple, ATI, Intel, Microsoft, Nvidia, Sony, Windows etc... and look at the quality of the Guest posting...then ask yourself whether you think it was worthwhile spending the few minutes of your life reading them.

spydercanopus spydercanopus said:

Someone get Bill Nye in here..

dividebyzero dividebyzero, trainee n00b, said:

back at school years ago we went on a science school trip and a girl in my class was chosen to come up on stage when they were doing a presentation, they made her hold these pieces of metal and then they wound some kind of device, her hair went a bit freezy and was xtremely attracted to baloons but appart from that she didn't even realise that 14,000v was travelling through her body, and the reason it all worked was because the current was pretty much not there!

Van de Graaff generator

Guest said:

I take my, as devidebyzero so eloquently put "butthurt" comment back.

After taking a scan at the comments I can completely agree 90% of the guest comments are moronic. I can understand why registered users dismiss these posts almost immediately, but consider this, people fly by these websites so often as there are so many, and want a small amount of instant input without going through the trouble of signing up.

Would you like to sign up to every single website that you visit? No I hear you say.

For the most part it seems registered users are the ones getting "butthurt" over guests comments, most of which are obviously "trolls"

Guest said:

I left on a fairly long trip down the interstate and when I went to plug in the GPS an hour down the road it turns out one of the children had put a coin in the cigarette lighter/power connector. Had to ask for directions.

Guest said:

I completely understand what registered users are talking about.

anyways my input is this. I dont know ANYTHING about electricity other then how manny watts are needed for computer components, so i will not argue with anyone.

my point is this = if they say it can be done im sure it can be, as for limitations im sure they know what they are and the problems that would arise if they dont take precautions.

all in all i say everyone stop acting like children fighting over pitiful things with people that you wont ever see, and mean nothing to you or your life. Even if you do win the debate it doesnt mean anything. you get some satisfaction yes.

but at the end of the day your input means nothing as USB will work out the bugs and issues and finally release the update.... just stop fighting and making a big deal out of nothing. Just be damn happy that we have a new update coming to our computers! -____-

1-4 people will die yearly from shock. and no one will be able to over current their motherboard..

USB has done very well so far, doubt they would be morons and not let concerns like all of yours fall on def ears....

Load all comments...

Add New Comment

TechSpot Members
Login or sign up for free,
it takes about 30 seconds.
You may also...
Get complete access to the TechSpot community. Join thousands of technology enthusiasts that contribute and share knowledge in our forum. Get a private inbox, upload your own photo gallery and more.