nRF5340 - simple circuit for measuring steady voltage and current

[Gabriele]

Hi,
I want to add some auto-test capability to my nRF5340-based project.
I thought one useful feature would be to measure voltage AND current at the hot node when all devices are active.
I needed a solution which uses few of nRF5340 pins, robust and cheap enough.
This circuit exploits one ADC channel (with 0.6V internal reference) and one GPIO.

step 1. Set HVSEL Low. Measure HVMON = V+
step 2. Set HVSEL High. Measure HVMON = V-
step 3. calculate current I = 2*(V+ - V-)/0.1

The resistors have guess values.
I'd appreciate any criticism and suggestions to improve it.

[ketiljo]

Should work ok but the resolution will be rather low as the voltage drop is only 50 mV at 500 mA current. Probably better to use a differential op-amp before the ADC instead of the analog mux.

[Gabriele]

OK. I've increased R1 and R3, getting 79 mV for 500 mA current.

Let me recap the Pros
1. I can also measure the voltage, not only the current
2. at the cost of one additional GPIO, replacing the SPDT with a SP3T, I can also measure the AA battery voltage with good resolution
3. when VDD is OFF, or mux selector are Low, the SP3T grants isolation
4. it is not expensive (1 eur)

And the Cons
1. less then 9 LSB increase in adc reading for 10 mA increase in current is not much. Would it be enough to detect a fault in the system ?
2. I think I should use oversampling to kill some noise

I don't know.

[ketiljo]

Should be easy enough to try. But you will have quite a voltage drop across the resistor at full current draw, so make sure the system can handle the ripple. Since you don't have an amplifier, there's a trade off between resolution and voltage drop.

[Gabriele]

I made your perplexities mine. I think this last step is more decent. Do you agree ?
Cost is still around 1 eur.

[ketiljo]

Definitely a much better solution!

There's an internal connection from VDD to the ADC so you can remove the analog mux as well.