(PeteCresswell) said:
On my little USB wall warts, I see two different output voltages.
Some claim to put out 5v, others say they put out 12v.
http://www.usbmadesimple.co.uk/ums_2.htm says that "The fourth
wire is called VBUS, and carries a nominal 5v supply...".
How does this relate to devices that get power from USB cables?
Is using a 12v wall wart going to damage a device that is
expecting 5v?
Or is there some standard where most devices can deal with either
voltage?
You are referring to a wall mounted power supply, which is
used to power USB 3.5" disk enclosures.
On the older ones, the cable would have four wires, and a
connector similar to a mini-DIN on the end with four pins.
http://en.wikipedia.org/wiki/File:Mini-DIN_male_4-pin_power_connector.jpg
The signals in that case would be GND, GND, +5V, +12V.
Those two voltages are the voltages that 3.5" hard drives use.
Later enclosures switched to a two pin wall mounted power supply.
The connector in that case, may be a "barrel" connector, and the
power would be +12V. The +12V can be used directly by the hard
drive (powers the motor). A 7805 style regulator inside the hard
drive enclosure, converts the +12V to +5V. All in the name of
not having to use that four pin connector for powering things.
(A barrel connector...)
http://rsk.imageg.net/graphics/product_images/pRS1-6717502w345.jpg
*******
A typical scheme for such a 3.5" USB hard drive enclosure, might
look like this.
+-------------+ +------------+ (Power)
(+5VSB) VBUS ----| USB to IDE | | Hard Drive |-------- +12V,+5V
D+ ----| or SATA | | | GND,GND
D- ----| chip |===|(Data Cable | from wall
GND ----| | | | adapter
+-------------+ +------------+
The only reason I'm drawing the diagram, is to show there is
not necessarily a path between "VBUS" and the wall adapter.
The +5V from the wall adapter, should not "leak" to the VBUS
side. I have one enclosure here, a 3.5" hard drive enclosure
with external adapter, and it's as well behaved as can be.
A model citizen, so to speak.
*******
Not all peripheral devices are as well designed. Some self-powered
USB hubs, attempt to push +5V (from wall adapter), into the +5VSB
(computer power) rail when they're connected. This can produce a
variety of symptoms, which disappear when the hub is disconnected
from the computer. What they're supposed to do there, is use a
small relay, to disconnect VBUS, when the adapter is present
and supplying +5V.
A difference between Firewire and USB, is Firewire uses an
"unregulated" VBUS, while USB uses a "regulated" VBUS. The
intention in the USB case, would be to save product cost, by
not having power regulation circuitry in the peripheral. The
"unregulated" distribution scheme has an advantage, that
precision control of where the current flows, is possible.
In the Firewire case, they can use simple diodes, to "wire OR"
all the potential power source(s). The diode allows current
to only flow in one direction. Devices can even provide
different magnitudes of voltage, like the one on the left
could make +12V and the one on the right could make +25V,
and it would still work. When the voltages aren't exactly
the same, all that happens is the one on the right (the 25V one)
"wins" and gets to supply all of the bus current flow. It's a
great scheme. Once the power gets inside an enclosure,
it must be converted (switching converter = $$$), to
a voltage like 5V or 3.3V or 1.8V or whatever. But at least
in terms of how the VBUS is handled, this is relatively
worry free. Just don't short VBUS to a data pin on the
Firewire bus... (if connector fits improperly, stop using it).
+VBUS +VBUS
| |
--- ---
\ / \ /
- -
| |
---+---------------+ (to anything on Firewire
needing power)
On USB, the VBUS is set to 5V, and the voltage distribution
budget on the USB "wire tree" is such, that it is not practical
to use the "diode steering" approach. Then, idiots are tempted
to connect power sources together, to save a few pennies on
using a relay or some other kind of appropriate solid state
switching solution (MOSFET with boosted gate voltage). So
depending on the pricing pressure ($10 or less for a USB hub),
the powering side of the product may be less than adequate.
If your USB device seems to be misbehaving, well, disconnect it :-(
The advice would mainly apply to USB devices that have
external +5V adapters. Those would include USB self-powered
hubs, or perhaps USB 2.5" hard drive enclosures with the
optional +5V barrel power input. Those are the ones that
come to mind. And the self-powered hubs are the ones
with the bad reputation. I haven't heard of problems
with 2.5" drives. Not all self-powered USB hubs with
wall adapters are bad designs, just some of them.
HTH,
Paul
