0
\$\begingroup\$

I have designed a PMOS (FDN306) based power on off switch. When I turn off the switch by connecting gate and source (via Schmitt Trigger), I see a output voltage of 50mV. This is with 3.9V Source voltage. If I increase the source voltage to 5V, the output voltage goes down to 20mV. I check the Gate and Source voltage and I see them identical.

I don't have the schematic in this computer but it is essentially a super simple switch. It sounds like, I didn't choose the right PMOS.

What type of PMOS I should choose? Why there is a significant leakage out of PMOS in this case?

\$\endgroup\$
  • 1
    \$\begingroup\$ What's the load? And is the behavior you're describing something you're seeing in a real circuit or a simulation? \$\endgroup\$ – The Photon Nov 4 '12 at 17:35
  • \$\begingroup\$ @ThePhoton the load in this case is hard to explain, the circuit drives several LDOs etc. The box consumes at 5V 0.33A, assuming all on, 12 ohm would be my best guess. This is real circuit, measured. \$\endgroup\$ – Ktc Nov 4 '12 at 23:30
  • \$\begingroup\$ So what's the problem? Your circuit is surely off at 50 mV, and the power drain will be in the uA range. \$\endgroup\$ – apalopohapa Nov 5 '12 at 3:40
  • \$\begingroup\$ (uW, in the previous comment) \$\endgroup\$ – apalopohapa Nov 5 '12 at 4:17
  • \$\begingroup\$ @apalopohapa I was hoping nA range. \$\endgroup\$ – Ktc Nov 5 '12 at 14:05
1
\$\begingroup\$

That doesn't appear to be anything unusual. From the datasheet you have a max of -1 uA @ Vds = -10 and Vgs =0. If the Ids is at the max for the off state then what the switch is seeing is a 50 mV/1uA = about 50 KOhm load from your LDO's which will be highly non-linear as the voltage increases. That seems about right.

You could decrease this leakage a bit further by driving the gate more positive but you can only do this slightly. If you drive it too far positive the leakage will increase again because of GIDL (Gate Induced Drain Leakage). Unfortunately the datasheet doesn't give these parameters.

\$\endgroup\$
  • \$\begingroup\$ Thanks.. I figured also the capacitance of the board is leaking back, if I wait sufficiently long, the leakage is really in <5uA range. \$\endgroup\$ – Ktc Nov 5 '12 at 14:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.