USB power depends on hardware revision

I bought a second WDTV Live recently and have reverted it down to 1.05.04V successfully.   I have kept my first Live box at 1.05.04V for several firmware functional issues I’ve discussed in other posts.  I’ve discovered an important  behavioural difference between my first (older) WDTV Live and this one with repect to USB power when the unit is shut OFF, even with the same firmware version loaded and the unit properly reset/initialized.  The older one is a model WDBAAN0000NBK-00, and the unit I just bought is an WDBAAN0000NBK-01. 

My older -00 rev Live, when shut off by power button, issues a shutdown command to the USB Hard drives, then removes power from the USB port.  The USB drives shut down completely; no spinning, no light.  The second Live, revision -01, also issues the USB shutdown command, which causes the drives to spin down, however, the light on the drives stays ON, indicating that the USB port is still powered.  If I unplug the USB cord and re-insert it, the drive spins up and sits like that indefinitely (I’ve deselected the auto-spin-down function of the drive unit). 

Using a USB breakout module in conjunction with a voltmeter confirmed my suspicions; the -01 Live puts out 1.3V on the rear port and just over 1V on the side port when shut OFF by the button.  Although the voltage levels are not at the proper USB 5VDC, they appear to be just enough to be detected as active by the USB hard drive. My  -00 Live presents a residual OFF voltage of around 0.2-0.3V, which is normal for technology like this.  This level is well below the detection threshold so the drive keeps everything off and remains in sleep.  It stays that way when I repeat the USB cord unplug-replug experiment, confirming that the drive is not detecting any USB voltage.

It’s important to note that devices that actually get their operating power from the USB, like USB flash memory sticks and WiFi adapters, will not operate since 1.3V  is well below their minimum startup level.  A separately powered USB hard disk on the other hand simply monitors the USB power line to determine whether it is plugged into an active host device.  Seems that for some drives, even 1V is enough to be detected as port ON.  My testing was done with an IOmega 1T backup drive, but I’m sure the same effect can be seen with other brands.  I should also point out that both USB ports work properly when I plug in a USB flash stick that is powered from USB, ruling out the likelihood that both ports were somehow partially damaged to a similar degree. 

Have you tried shutting down by pressing off for 5 or so seconds?

As I expected, no difference how long I hold the power button.  The -01 clearly and demonstrably behaves differently from the -00 running identical firmware version.  I am very sure this is a hardware difference; I’ve designed enough products with USB host ports on them to know that a port should present 5V when ON and less than 0.4V when OFF.