LTC2954-2 is partially enabling power when it should be completely OFF

We are following the simple application note for the LTC2954-2 but notice that when the button is not pressed we are measuring about 500 - 650 mV output from the pfet going to the Vin of the regulator.

Measurements before button pressed:

1. LTC2954-2 Vin - 5.12V
2. PB - Measured 1.7V at PB
3. EN - 4.66V

4. Output from pfet: 0.88V

Below is our actual schematic for the design that is causing issues

added voltage measurement notes
[edited by: at 5:16 PM (GMT 0) on 11 Mar 2019]
  • 0
    •  Analog Employees 
    on Mar 27, 2019 3:28 PM


    Form the information provided, seems like PFET is slightly ON. Since Vgs is = Ven - Vin = 4.66V – 5.12V = -0.46V. (FDMA291P Vth is -0.4V min, -0.7V typ, -1V max.) EN being at 4.66V implies there is 4.6uA ( = 0.46V / 100kOhm) of current flowing down through the 100k resistor (R11) between EN# and VIN. (Note that measuring EN# voltage with a 10MegOhm meter input impedance contributes 0.466uA to this leakage. Be wary of hooking up any 1MegOhm scope probes to EN#.) Worst-case leakage (over temperature) of LTC2954 EN# pin is 0.1uA at 1V and 1uA at 26.4V (datasheet page 3). Therefore, please investigate where this excess leakage is coming from: is it something on the board or could it be that the FET (try a new or different FET) or LTC2954 has suffered some damage. Try disconnecting the downstream DC/DC while doing their investigation.