I don't do arduino, but I suspect any ADC you use should be referenced to ground, not a pin-to-pin differential.

But wouldn't that give a different reading altogether ? Using a DM its would be on pins 1 and 2 but how is it done with an ADC ? is it difference in voltage ?

 

Based on the schematic would this simulation be correct ? i am trying to determine the current flowing through the weld cable to the nickel strip. The Rds(ON) value is based off the MOSFET's datasheet.

 

I do not think you can rely on the MOSFET Rds(ON) to be a fixed value. You will do better to have a known resistor.

 

Ignoring the constant current and the gain amplifier, is it how it would be simulated to show the amp draw ?

 

Here is one with the actual Rsense resistor. The heat dissipated is a crazy 500W, how bad would it be considering that the welds only last for less than 50ms.