[TriLUG] why does usb card on PCIe bus need extra power to detect a disk?

[Joseph Mack NA3T via TriLUG]

On Thu, 13 Dec 2018, Stephen Bryant wrote:

hmm. Well it looks like I don't know much about usb 3.x. Until just now I 
thought it's just like 2.0 only faster. I shouldn't be surprised to find that a 
3.x host is expected to provide a fairly large amount of power to the device(s).

I didn't know the voltages on the PCIe bus. If they don't have 5V already there 
and there's amps of 5V from the computer's PS, then why not get it from there? 
In that case I should be looking for cards that require external power, since 
this is the sensible thing to do. I had thought something was wrong that the 
cards needed external 5V.

Thanks for the explanation.

Joe

> In my experience with external drives, a lot of them use bus power for the 
> controller, even if the drive itself is run on external power.
>
> My guess, given the “lowest common spec” so much hardware is built to these 
> days:
>
> External drive powers HDD off the power brick, but expects the controller to 
> get its (relatively minuscule) power from the USB power pins. This would make 
> sense, since the USB3 spec requires the host to supply 150mA for a “low-power” 
> device, and 900mA for a normal device. This allows the manufacturer to have 
> two separate (and therefore multipurpose) boards - one for drive power supply, 
> and one for USB<—>SATA.
>
> The PCI-e card is designed to have a power input, so instead of having a 
> transformer to run the USB (5v) from the PCI-e (3.3v+12v) supply, they use the 
> 5v pin from the Molex connection. Saves on complexity and BOM when you’re 
> gonna have that 5v supply anyway.
>
> That’s my perhaps-underinformed guess, anyway :)
>
> — Stephen

-- 
Joseph Mack NA3T EME(B,D), FM05lw North Carolina
jmack (at) wm7d (dot) net - azimuthal equidistant
map generator at http://www.wm7d.net/azproj.shtml
Homepage http://www.austintek.com/ It's GNU/Linux!