I'm looking for a switching device to be used for powering a sub-circuit in a battery powered low / ultra-low power circuit.

It can be a MOSFET, or something similar but it need to have some special properties. enter image description here

The sleep mode power consumption of the circuit, without the above sub-circuit + switching device, will be in the range of 25-50nA @3.3V.

The ON current for the sub-circuit is expected to be ~25-50μA, with occasional spikes of at most 10-20mA.

Because the above currents are very small, I think we need a low power switching device with very small leakage currents. This way the sub-circuit should not gets accidentally powered and the power consumption in sleep mode is kept at minimum.

As far as now, I looked at following options:

  • MOSFET-s:
    • Vishay, Si1499DH - MOSFET, source-drain leakage current 1μA/10μA, gate leakage current 100nA
    • Texas Instruments, CSD23280F3 - MOSFET, source-drain leakage current 50nA, gate leakage current 25nA
  • load switches
    • Texas Instruments, PS22860 - load switch, leakage current 10 + 2 nA
    • Vishay, SiP32431/2 - load switch, leakage current 0.001nA, with 100nA max @3.3V (note: found this after posting the question, and looks to be good enough)
  • some JFET-s, but I'm not sure if they could be used / are suitable for such an application

Now, may there may be a lot of devices / solutions I'm not aware of. What do you think are the most suitable devices or solutions for such a problem?

  • \$\begingroup\$ This looks like a shopping question, which is considered off topic (Stackexchange considers such questions to not be timeless enough). Questions about what technologies to be considered for this would be on-topic, though. \$\endgroup\$
    – TimWescott
    Commented Jan 5, 2020 at 19:44
  • \$\begingroup\$ @TimWescott, by switching device I meant anything that acts like a switch. Normally, in a non low power application I would have used MOSFET. So, before posting this question I searched for MOSFETs, but I did not found a suitable one. Then, I looked for other devices types (technologies), found some good ones, but there may be some better options. Also, there may some better suited device types, technologies, or techniques am simply I'm not aware off. \$\endgroup\$
    – bluetiger9
    Commented Jan 5, 2020 at 20:01
  • 1
    \$\begingroup\$ What do you exactly mean by "PFET"? P-channel MOSFET? I'm interested in ultra-low leakage currents (in any direction) in the OFF state. P-channel MOSFET-s, the better ones, seems to have "Gate-Source Leakage (IGSS)" of max 100nA, "Zero Gate Voltage Drain Current (IDSS)" of max 1μA. I'm looking for values around 0-5nA. \$\endgroup\$
    – bluetiger9
    Commented Jan 5, 2020 at 21:08
  • 2
    \$\begingroup\$ @bluetiger9 What about this? I note that it appears to have fairly low leakage specs. (I didn't look at much else.) The other thing to note is that even if very clean, the bulk impedance of an epoxy package is on the order of \$1\times 10^{11}\:\Omega\$. If it is in any way dirty, that goes way down and fast. Keep things clean. I had a case where I needed nearer \$1\times 10^{14}\:\Omega\$ and so I had to accept dice, instead. \$\endgroup\$
    – jonk
    Commented Jan 5, 2020 at 22:32
  • 1
    \$\begingroup\$ @jonk, the TI CSD15380 you linked looks very similar to the TI CSD23280 I looked at. This epoxy + dirt thing sound interesting. Thanks for drawing my attention to this :) \$\endgroup\$
    – bluetiger9
    Commented Jan 6, 2020 at 18:55

1 Answer 1


Because you've done some adequate research into a solution and have put together a half decent requirement I'm making an answer...

The ADG819 2:1 multiplexer operates from 1.8 volts to 5.5 volts and has a maximum D-S leakage current of +/- 10 nA across the full -40 C to +125 C temperature range. On-resistance is 0.8 ohms maximum and this would result in a volt drop of 16 mV at 20 mA current through the channel.

If you want more switching options the ADG1604 (4 channel multiplexer) is also pretty useful. On resistance is 1.6 ohms and D-S leakage is 16 nA max.

  • \$\begingroup\$ The ADG819 looks to have fairly good specks. I think, I will order one or two pieces from each of the mentioned devices so far and do some tests. Just for fun :) The only thing I need figure is how to measure nanoamps :D \$\endgroup\$
    – bluetiger9
    Commented Jan 6, 2020 at 19:02
  • 1
    \$\begingroup\$ @bluetiger9 Two choices. Lots and Lots of money directed at Keithley. (They do have a nice coulomb meter.) Or look at a few Burr Brown (bought up, as you know) parts. The DDC112, for example, takes some time to study but its pretty easy to use. (I used to maintain contact with one of the designers, Jim Todsen, whose name you'll find on some of the documentation. No idea where he's at, now. Also, Bonnie Baker [curmudgeon personality] had also been helpful for me. She has since left and gone on to other things.) \$\endgroup\$
    – jonk
    Commented Jan 6, 2020 at 19:14
  • \$\begingroup\$ This DDC112 is still a little bit expensive. One another option I have seen is to use a DMM's internal resistance as a current shunt resistor. The DMM is set into voltmeter mode (with fixed range), and connected in series with the circuit. Now, if we assume that the internal resistance of the DMM in that mode is known and constant, we from the voltage drop measured by the DMM we can calculate the current. \$\endgroup\$
    – bluetiger9
    Commented Jan 6, 2020 at 19:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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