microcontroller - Measuring an analogue voltage with isolated ground

I am working with a solar EV (car), and I need to measure the throttle pedal signal. The throttle peddle is essentially a 10k ohm pot between 5v and gnd on the motor controller. (0-5V signal).

As per the rules, the high voltage and low voltage circuits in the EV must be isolated. The exact spec is 100 V ohm resistance isolation from both terminals of the high voltage battery bank to the low voltage circuit, where V is the nominal voltage of the bank. The bank nominal is about 110V so afaik I need to have at least 11k ohm resistance to the HV side. The low voltage circuit (which runs my electronics) is powered from an isolated DC-DC converter.

As such, the problem as it stands is that the HV gnd and LV gnd are isolated, so I cannot feed the throttle signal directly to my MCU. I can see a couple of solutions but I'm not sure what is the best/easiest/quickest(in terms of implementation and signal delay)/most reliable solution. I think I would prefer an optically isolated solution if possible. My ideas:

• Isolated op-amp (TI ISO124 http://www.ti.com/lit/ds/symlink/iso124.pdf ? Problem: expensive, don't have part on hand, never used before.. simple solution though)


• Linear opto-coupler (IL300 , no part on hand, complicated, never used before)


• ADC - opto - DAC (use small MCU on HV side to do ADC and send over opto-coupler serially. Easy but complex, potential signal delay?, will take time..)


• Volt to Freq -> opto -> MCU(LM331 : Dont have part on hand, never done before, have to write freq conversion code for MCU. Simple however)

Or non isolated:

• Differential instrumentation amp, tie HV gnd to LV gnd at amp with high value resistor (INA128 : Have part on hand, simple solution with minimal parts. Disadvantage of no isolation but maybe not a problem because it complies with the rules AFAIK. Potential noise issues? Potential for fault through high value resistor? Potential for damage to amp if HV gnd rises with respect to LV gnd?)


• Option B ?

Not sure which path to take from here. Short on time and need a reliable solution. Feel like there is a super simple solution I'm just missing? Perhaps the diff amp solution is fine and I'm just being paranoid? Any suggestions or resources would be appreciated!