2

I've previously used complimentary pairs (eg DMC2990UDJ) to switch ICs on and off.

I've attached a schematic which switches a potential divider for battery voltage monitor as an example. Where R7 and R8 are I'd add a sensor instead.

I'm now designing a battery powered sensor board which needs many switches. Are there any ICs suitable for this which only take a couple of nano amps when powered down which doesn't need the external resistors (R4, R5 & R6)?

Update:

VBAT = 3.4 - 3.8 V VBAT-EN = 3.3 V

batteryVoltage

josh
  • 103
  • 6
  • 2
    R4 and R5 may not be needed. Often are not needed. R6 could be 1 M or maybe even 10 M to help reduce quiescent current. Also, if the IO voltage controlling the gate is the same as the voltage you are switching, you can eliminate D1A altogether and use the high side PMOS only. In that case a low output would turn the switch on, and a high output would turn it off. – user57037 Aug 23 '21 at 02:35
  • @mkeith I've added the voltages to the questions - many thanks – josh Aug 23 '21 at 05:56

2 Answers2

2

If your measurements are for voltages between 1.1V and 3.6V then you can use a level shifter IC like the 74AUP1T45. Its not made for this purpose, but it will probably work.

https://assets.nexperia.com/documents/data-sheet/74AUP1T45.pdf

The 74xxx1T45 chip is a level translator that has one input one output and two supply voltages. Each supply voltage VCCA and VCCB consume 500nA each (for a total of 1uA per switch).

If you put a signal into the A side (powered by VCCA) that same digital value will come out of the B side (powered by VCCB). But note that, in the absence of any significant load, the output will be very very close to VCCB.

So if you connect VCCB to the voltage you are measuring then you can output the voltage to be measured when the input is 1. And you can output 0V when the input is 0.

schematic

simulate this circuit – Schematic created using CircuitLab

user4574
  • 11,816
  • 17
  • 30
0

You might find an inverter you can use instead of D1A. You would be looking for an inverter that has very low quiescent current, can tolerate 3.8 V at its VCC power rail, and will still recognize 3.3V cmos input levels when powered from 3.8V.

For example TI p/n SN74AUP1T04DCKR is close. The only problem is that 3.8 V is outside the recommended operating conditions for VCC (but still far from the absolute maximum...).

I also found this one: Nexperia p/n 74AHC1G04GV,125.

It looks like it would recognize your 3.3 V cmos input levels even when it is powered from 3.8 V. Quiescent current and input leakage current could become a bit of a concern. The quiescent current is given as 1uA maximum at 25 C, but it goes up quite a bit with temperature. Not sure if your circuit will be exposed to high temperatures.

The other option is to keep your existing structure with two FET's, but eliminate R4 and R5 (which shouldn't be needed if you drive the gate of D1A with 3.3 V gpio). And then set R6 to a much higher value such as 1 Mohm or even 10 Mohm. If you go to 10 Mohm you will have to make sure that the leakage current in D1A does not cause D1B to turn on unexpectedly. This problem will be most likely to manifest at higher temperatures when the FET's get kind of leaky.

Keep in mind that R6 only conducts current when D1B is on. If the load switched by D1B is at least 100 uA or more, then maybe you can just use 1 MOhm for R6 and not worry too much about the extra 3.8 uA. On the other hand, if the load switched is very small, then I can see how the extra 3.8 uA could be a concern.

user57037
  • 28,915
  • 1
  • 28
  • 81