# Very low leakage enhancement mode mosfets

I have a 1.8V logic signal that I would like to use to control an enhancement N-channel MOSFET to switch current through a resistor from the drain to a 3V power rail.

It's in a very power sensitive design. All of the enhancement-mode mosfets I looked at seem to specify max "Zero Gate Voltage Drain Current" Idss as 1 μA, which means if I would want the resistor/drain junction to go close to the 3V rail - say within 0.27V - the resistor would need to be a 270KΩ pull-up, and a 10μA on current, all of which are significant currents in a small battery powered application.

Update: I have limited board space, and need about 6 off devices. I have looked at some TI FemtoFET devices (bit of a pain to place as very small) and TPS22860 ultra low leakage load switches which might work.

Any suggestions - should I be using a different type of device?

• Include the schematic! Yes even if it is not working/wrong/whatever. You mention "enhancement mode" a lot, almost no-one uses the opposite (depletion mode) as they're very rare and usually not needed. My guess is that you're overcomplicating things but to see that I need your schematic. – Bimpelrekkie Jan 27 at 15:00
• Usually the 1uA spec is at 25°C Tj, so you need to consider temperature. 2N7000/2 has a maximum leakage of 0.5 or 1mA at maximum temperature! That said, the typical leakage is much, much lower, in the nA range. Do you feel lucky? – Spehro Pefhany Jan 27 at 15:08
• What MOSFET on resistance are you considering and, what value of resistor are you considering and what's the worst case volt drop across said resistor when the MOSFET is off. – Andy aka Jan 27 at 15:12
• Thanks for comments guys. I am an old physicist, hence the mention of enhancement mode mosfets. The 1uA is worst case across full temp range for the devices. I will add a schematic – Giles Jan 27 at 15:49
• If the current through that pull-up resistor is an issue, use a levelshifter IC instead of this circuit. For example: ti.com/lit/ds/symlink/… – Bimpelrekkie Jan 27 at 16:15

The ADG601/602 analogue switch has very low leakage: -

And the leakage is lower on a lower supply voltage such as 3.3 volts and 0 volts.

It also has a 2.5 Ω on resistance if that helps?

The supply rail is: -

Dual ±2.7 V to ±5.5 V or single +2.7 V to +5.5 V supplies


So, just regard it as a load switch.

• Thank you for the reply - I have a 0 to 1.8V digital input, so cant meet Vinh – Giles Jan 27 at 15:56
• @Giles then the problem has reduced to how can I make my signal greater than 2.4 volts surely. – Andy aka Jan 27 at 16:03
• Thanks for the suggestion - I have limited board space, and need about 6 off devices. I have looked at some TI FemtoFET devices (bit of a pain to place as very small) and TPS22860 ultra low leakage load switches which might work. – Giles Jan 27 at 16:19
• You should mention these constraints and endeavours in your question. – Andy aka Jan 27 at 16:20
• Indeed - apologies Andy. It is my first post; I will try and include more information next time. – Giles Jan 27 at 16:38