I did some testing of the current-limiting USB switch we're
considering for M1rc4. In this round of tests, I supplied the M1
from a lab power supply (which means that current limits are
strictly enforced) and switched resistive loads on one USB port
with labsw. The latter simulates hot-plugging.

Details are here:
http://projects.qi-hardware.com/index.php/p/wernermisc/source/tree/master/m1rc3/usbpwr/README

Key findings:

- if my M1pre-rc4 is operating at or slightly below the nominal
  supply voltage, the USB switch does not protect the 5 V rail
  from deep drops in the case of shorts ("0 Ohm")

  
http://projects.qi-hardware.com/index.php/p/wernermisc/source/tree/master/m1rc3/usbpwr/usb-short-10us.png

- in the same configuration, a 2 Ohm load causes almost no
  visible upset:

  
http://projects.qi-hardware.com/index.php/p/wernermisc/source/tree/master/m1rc3/usbpwr/usb-2R-200us.png

- if increasing the supply voltage above the nominal 5.0 V, also
  "zero-Ohm" upsets are survivable:

  
http://projects.qi-hardware.com/index.php/p/wernermisc/source/tree/master/m1rc3/usbpwr/usb-short-5V4.png

I'm not quite sure yet why the USB voltage (channel 2) doesn't
drop to zero when shorted. Might be a grounding problem.

- Werner
_______________________________________________
http://lists.milkymist.org/listinfo.cgi/devel-milkymist.org
IRC: #milkymist@Freenode

Reply via email to