Voltage follower op-amp for battery monitoring

Question

I want to read the voltage of my 12 V lead-acid battery, and I want it to be isolated from my circuit. I came across the voltage follower op-amp topology. My schematic is shown below:

I am using the NE5532 op-amp. First, the battery input is fed to a voltage divider that goes from 0-12 V to 0-3.3 V using the voltage divider network R50 and R51.

The output of the voltage follower then goes to the ESP32 ADC input.

My questions are:

1. Is there anything that does not look right in terms of how I set up the voltage follower?
2. Are there any mistakes in general on how I am doing this?

Updated version:

Answer 1

With 0-12V power rails, the NE5532 can only reliably read input voltages of a few volts above ground in that configuration. This is shown on the datasheet as so:

For the follower to work, the output voltage must also be able to cover that range, but if the input doesn't work we need look no further.

Your suggested use of zener diodes to clamp the output will likely lead to large errors near the clamping voltage.

You may be able to do better by using a rail-to-rail input and output CMOS op-amp and high impedance divider, and by powering the op-amp from 0/3.3 and using the internal protection diodes to clamp the op-amp input avoid the necessity of using zeners. Alternatively, make a better clamp using a TL431, bias resistor and a 1N4148 diode, and avoid using the top of the voltage range of the ADC (maybe up to 2.3V rather than 3.3, which saves two resistors on the TL431 since it would clamp at 2.5+ 1 diode drop). Such a clamp can be applied to the input or output.

A high impedance divider (eg. megohm) can reduce the drain from the battery to well below the self-discharge current so it is negligible. An op-amp with very low input bias current (just about any CMOS rail-to-rail op-amp) is necessary.

Answer 2

In principle the idea is okay. I have a few suggestions:

• The divider resistors will drain the battery (~2 mA at 12 V). Considering the input bias current of the op amp, the divider resistors could have been larger to decrease the consumption (10x resistance can still work here).
• I don't think you need to monitor the battery voltage every single millisecond. If you can take the measurement in, say, every second then you can place a series (preferably, high side) switch to the divider network so you can connect the divider before measurement, and disconnect after measurement. This will reduce the consumption further.

^{simulate this circuit – Schematic created using CircuitLab}

The circuit above can be useful. After turning on the switch you may want to allow some time for the filtering capacitor, C10, to charge up.

Answer 3

Generally, input protection uses a Zener ... or classical two diodes and one resistor.
Resistor limit current. Diodes used are very "low leakage".
There is some post on this site explaining a little this method.
Example.