Author Topic: USB 3.0 Power Delivery Specification  (Read 16295 times)

0 Members and 1 Guest are viewing this topic.

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: nz
Re: USB 3.0 Power Delivery Specification
« Reply #25 on: July 25, 2012, 10:41:31 pm »
They're not going to use premade switchmode modules on motherboards for usb 3 power.
For that sort of volume it would be far cheaper to design it onto the motherboard and doing so means they can design it for whatever voltages they need.

Premade PCB mount dcdc modules are for low volume production, or when you need something out-the-door in a hurry.
« Last Edit: July 25, 2012, 10:43:21 pm by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 6706
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: USB 3.0 Power Delivery Specification
« Reply #26 on: July 25, 2012, 10:44:55 pm »
They're not going to use premade switchmode modules on motherboards for usb 3 power. For that sort of volume it would be far cheaper to design it onto the motherboard. Doing so means they can design it for whatever voltages they need.

Premade PCB mount dcdc modules are for low volume production, or when you need something out-the-door in a hurry.

No, I meant current (available)DC-DC converter ICs. Not DC-DC modules, far too expensive. Anyway, this now means a PoE enabled motherboard with USB 3.0 super-duper power needs 48V and 20V rails.

I wonder if the standard will be implemented for laptops?
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: nz
Re: USB 3.0 Power Delivery Specification
« Reply #27 on: July 25, 2012, 10:48:22 pm »
They're not going to use premade switchmode modules on motherboards for usb 3 power. For that sort of volume it would be far cheaper to design it onto the motherboard. Doing so means they can design it for whatever voltages they need.

Premade PCB mount dcdc modules are for low volume production, or when you need something out-the-door in a hurry.

No, I meant current (available)DC-DC converter ICs. Not DC-DC modules, far too expensive. Anyway, this now means a PoE enabled motherboard with USB 3.0 super-duper power needs 48V and 20V rails.

I wonder if the standard will be implemented for laptops?

Are you saying there are no cheap DC-DC converter ICs that can do 20V output? because that doesn't seem right.
Lost of switchmode ICs are quite universal in there in/out specs. There's a wide range you can do just by changing the feedback and the inductor.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline ivan747

  • Super Contributor
  • ***
  • Posts: 2045
  • Country: us
Re: USB 3.0 Power Delivery Specification
« Reply #28 on: July 26, 2012, 02:20:37 pm »
I wonder if the standard will be implemented for laptops?

Most likely it won't be implemented. Maybe they will start producing some sort of adapter that plugs into the wall and injects the 100W requested while working. The computer would see it as a hub.
 

Offline Stephen Hill

  • Regular Contributor
  • *
  • Posts: 178
  • Country: gb
  • M3VXY
Re: USB 3.0 Power Delivery Specification
« Reply #29 on: July 26, 2012, 02:47:13 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #30 on: July 26, 2012, 05:12:27 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....

Don't be surprised to hear high end CPU's having a TDP of 150W ...
 

Offline Short Circuit

  • Frequent Contributor
  • **
  • Posts: 439
  • Country: nl
Re: USB 3.0 Power Delivery Specification
« Reply #31 on: July 26, 2012, 05:19:52 pm »
No, a trace can not handle 100watts (handling as in dissipating). And beyond dissipation, it's somewhat pointless to relate watts to traces.

If you're talking 100 watts in the context of this topic, then yes, sure it can. We're talking a mere 5 amps.
Enough to seriously consider when drawing the trace, but no big deal at all.
Beyond 10-15 amps things get a bit more seriously, possibly involving thicker copper, or really wide traces, but still  no problem.

Last week I got a Weidmuller sample kit, containing some PCB terminals for 1000V/76Amps, so that's good for 76kW.
http://catalog.weidmueller.com/catalog/Start.do?localeId=en&ObjectID=1851120000
 

Offline M. András

  • Super Contributor
  • ***
  • Posts: 1014
  • Country: hu
Re: USB 3.0 Power Delivery Specification
« Reply #32 on: July 28, 2012, 08:15:03 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....

Don't be surprised to hear high end CPU's having a TDP of 150W ...
thats more than 120amps approx for a cpu btw :) im lazy to do the math, remember its running around 1.1-1.5 volts
 

Offline Stephen Hill

  • Regular Contributor
  • *
  • Posts: 178
  • Country: gb
  • M3VXY
Re: USB 3.0 Power Delivery Specification
« Reply #33 on: July 28, 2012, 08:25:15 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....

Don't be surprised to hear high end CPU's having a TDP of 150W ...
thats more than 120amps approx for a cpu btw :) im lazy to do the math, remember its running around 1.1-1.5 volts

But a typical processor will have well over 100 Vdd, so the power is spread out over multiple pins and traces. For example, an Intel i7 has over 200 Vdd pins and can take a total of 145Amps. So that's no more than 725ma per pin/trace.
 

Offline markus_b

  • Regular Contributor
  • *
  • Posts: 115
  • Country: ch
Re: USB 3.0 Power Delivery Specification
« Reply #34 on: July 28, 2012, 09:06:42 pm »
I find that a very useful development. Now I'm waiting impatiently for interesting chips to handle the protocol and DC switching, for me to use this for my homegrown devices.

On my shopping/watch list right now figures a new USB power supply with a couple of outlets. This is the ideal device to feed my lab power supply project too.
Markus

A good scientist is a person with original ideas. A good engineer is a person who makes a design that works with as few original ideas as possible.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #35 on: July 28, 2012, 09:26:56 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....

Don't be surprised to hear high end CPU's having a TDP of 150W ...
thats more than 120amps approx for a cpu btw :) im lazy to do the math, remember its running around 1.1-1.5 volts

But a typical processor will have well over 100 Vdd, so the power is spread out over multiple pins and traces. For example, an Intel i7 has over 200 Vdd pins and can take a total of 145Amps. So that's no more than 725ma per pin/trace.

On a trace not much bigger than the bga pad ... But i really begin to wonder where do they dump all the 150W of heat without getting over 70C ... I worry for that. Because Sandy Bridges and Ivy Bridges max temps have gone down to 65 ...
« Last Edit: July 28, 2012, 09:44:02 pm by DaveXRT »
 

Offline tom66

  • Super Contributor
  • ***
  • Posts: 6706
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: USB 3.0 Power Delivery Specification
« Reply #36 on: July 28, 2012, 09:37:29 pm »
Just waiting for a USB soldering iron...
 

Offline M. András

  • Super Contributor
  • ***
  • Posts: 1014
  • Country: hu
Re: USB 3.0 Power Delivery Specification
« Reply #37 on: July 29, 2012, 03:59:23 pm »
Can a motherboard PCB trace handle up to 100W? I'm not sure it can. Could be wrong....

Don't be surprised to hear high end CPU's having a TDP of 150W ...
thats more than 120amps approx for a cpu btw :) im lazy to do the math, remember its running around 1.1-1.5 volts

But a typical processor will have well over 100 Vdd, so the power is spread out over multiple pins and traces. For example, an Intel i7 has over 200 Vdd pins and can take a total of 145Amps. So that's no more than 725ma per pin/trace.

On a trace not much bigger than the bga pad ... But i really begin to wonder where do they dump all the 150W of heat without getting over 70C ... I worry for that. Because Sandy Bridges and Ivy Bridges max temps have gone down to 65 ...
to an almost 1kg weighted copper+aluminum block with heatpipes? then to heat up the room in winter
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #38 on: July 29, 2012, 05:09:54 pm »
They stopped giving you aircooled heatsinks ... AMD i last recall gives you a Corsair H80-type heatsinks for the FX8150
 

Offline M. András

  • Super Contributor
  • ***
  • Posts: 1014
  • Country: hu
Re: USB 3.0 Power Delivery Specification
« Reply #39 on: July 29, 2012, 07:25:23 pm »
good to know that :) but if watercooling. a big 3x12cm fan radiator with high flowrate a laing d5 pump split connectors at the gpu and at the cpu to give both of them cool water. then feed them back again to 1 with the same connector after the outlets. much quiter then those corsair stuffs
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #40 on: July 29, 2012, 07:38:32 pm »
good to know that :) but if watercooling. a big 3x12cm fan radiator with high flowrate a laing d5 pump split connectors at the gpu and at the cpu to give both of them cool water. then feed them back again to 1 with the same connector after the outlets. much quiter then those corsair stuffs

I don't care for the fan volume, those can be easily modded but corsair was the first one they went to and readily agreed to stamp their names on the water cooling systems
those GPU's gone entirely mad, the 7970 goes 250W TDP on full load! ...
And they are TBA on the 7990 ... aka the 7970 X2 ... 500W, want that? 2 of them makes for a very expensive 1kW room warmer in the winter
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16283
  • Country: za
Re: USB 3.0 Power Delivery Specification
« Reply #41 on: July 29, 2012, 07:41:12 pm »
My watercooling at work is a cooling condenser from a display cabinet ( new spare part, the prototype was a used one) with integrated fan. Keeps the water at 24C with whatever heat it collects from 6kW of heating from the packing machine it is on. It needed a bit more cooling than it had from the 50l copper tank it has as the first improvement i added over the oem method of running tap water to waste.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9016
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: USB 3.0 Power Delivery Specification
« Reply #42 on: July 30, 2012, 03:58:09 am »
On a trace not much bigger than the bga pad ... But i really begin to wonder where do they dump all the 150W of heat without getting over 70C ... I worry for that. Because Sandy Bridges and Ivy Bridges max temps have gone down to 65 ...
Just a large heatsink and some powerful fans. My i7 will easily hold Turbo Boost with full load on all 6 cores indefinitely. I just bought a cheap but good Cooler Master heatsink, replaced the weak bundled fan with a more powerful one, and then stuck another powerful fan in the side panel.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #43 on: July 30, 2012, 10:28:03 am »
On a trace not much bigger than the bga pad ... But i really begin to wonder where do they dump all the 150W of heat without getting over 70C ... I worry for that. Because Sandy Bridges and Ivy Bridges max temps have gone down to 65 ...
Just a large heatsink and some powerful fans. My i7 will easily hold Turbo Boost with full load on all 6 cores indefinitely. I just bought a cheap but good Cooler Master heatsink, replaced the weak bundled fan with a more powerful one, and then stuck another powerful fan in the side panel.

Yeah, i know... Basically any cheap CM ones are a heck lot better than the stock cooler
and the HSF supplied with the 980X isn't helping either ...
But then again for those people who bought the 980X ... they would use the 980X "stock" cooler for something else
 

Offline M. András

  • Super Contributor
  • ***
  • Posts: 1014
  • Country: hu
Re: USB 3.0 Power Delivery Specification
« Reply #44 on: July 30, 2012, 06:58:18 pm »
good to know that :) but if watercooling. a big 3x12cm fan radiator with high flowrate a laing d5 pump split connectors at the gpu and at the cpu to give both of them cool water. then feed them back again to 1 with the same connector after the outlets. much quiter then those corsair stuffs

I don't care for the fan volume, those can be easily modded but corsair was the first one they went to and readily agreed to stamp their names on the water cooling systems
those GPU's gone entirely mad, the 7970 goes 250W TDP on full load! ...
And they are TBA on the 7990 ... aka the 7970 X2 ... 500W, want that? 2 of them makes for a very expensive 1kW room warmer in the winter

i have an asus g73 with the ati card in it mobility 5970, the desktop has an 6970+1090t amd 4 core processor. playing mmorpg(not world of warcraft if anyone familiar with these) on both of them at the same time at winter my room was over 25celsius with the heating turned off :)
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: USB 3.0 Power Delivery Specification
« Reply #45 on: July 30, 2012, 07:20:10 pm »
I see what you mean. What kind of mmorpg have you been playing that made the 6970 went max ::)
 

Offline M. András

  • Super Contributor
  • ***
  • Posts: 1014
  • Country: hu
Re: USB 3.0 Power Delivery Specification
« Reply #46 on: July 31, 2012, 04:14:21 pm »
I see what you mean. What kind of mmorpg have you been playing that made the 6970 went max ::)
"eve online" 2 client/computer, hungry little game. it has the reference cooler on it so its loud and hot. perfect room heater, the laptop consumes over 150watts when fully loaded dual fan cooled each ducting blow approx 70celsius air with a pretty good flowrate
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf