Servo V2 pulls warm_reset up to 2.5V |
|||
Issue descriptionIf you probe the voltage on J7 pin 5 (DUT_WARM_RESET_L), you'll find that it's pulled up to 2.5V even without any flex attached. This leaks into DUTs and can backpower things slightly. R135 is definitely DNS'd. My best guess is that the FTDI has a default pull-up to 3.3V on that pin (ADBUS5), and that the IO buffer is clamping the voltage to its own 1.8V supply. Todd, can you try configuring the FTDI part into a mode where it doesn't pull up on ADBUS5 and see if the voltage goes away (i.e., it floats to another voltage, or if you short it to ground and release it doesn't move much)?
,
May 25 2016
When I do a dut-control, I get warm_reset:on and I see the leakage from the warm_reset being asserted high. When I do a dut-control warm_reset:off, the leakage does not go away However, when I do a dut-control warm_reset:on, the leakage DOES go away. Not sure what state we are initially reporting...
,
May 26 2016
Nothing obvious I can find to fix this in SW. There's no references to pull-ups in datasheet nor the eeprom that I can find and anytime I make the pin and input it floats to 2.5V
,
May 26 2016
Should have mentioned to test outside of servo can use small utility ftdi_gpio For example, sudo fgpio_test -v 0x18d1 -p 0x5002 -i a -g 0x20:0x00 Would force warm_reset to ground after launch it defaults to interactive to continue setting the direction & value register of the gpios. 'fgpio_test -h' for usage.
,
Jul 8 2016
Wontfix based on #c3. Will need to make sure isolation ckt on dut is included. |
|||
►
Sign in to add a comment |
|||
Comment 1 by nsanders@chromium.org
, May 25 2016