The ADS1115 is not a very good choice, and opamp circuits could be a pain to design.
Use a chip that's designed for shunts, like the INA226. They also come on breakout boards.
What is your plan to keep each circuit isolated fro all the others?
Can you explain why this would be necessary?
My thinking is that I'm just measuring small voltages (across the shunt) with high impedance inputs.
The "shunts" I have used picked up two spots on a copper bar that was carrying the current to be measured. There was NO isolation between the large current and the tiny measured voltage. Apparently yo have something different.
Also, I see that at least one listing refers to this module as having a 3A max current measuring range. Would it be correct to assert that with removal of the on-board 100-ohm resistor, you can connect it across any shunt with a max output voltage within the device's ~ +/-85mV range?
The shunt on modules is 0.1 ohm, not 100 ohm.
Not a big problem if you connect an external shunt with a much lower resistance parallel to that.
It just increases a 500A shunt to a 503A shunt.
Must keep the wires between module and shunt short/fat though, or remove the 0.1ohm shunt.
I should also point out that this is for a DC system, with 0 to 15V expected under normal operation.
Much easier to have the shunt high-side (in the + line). That also eliminates possible groundloop problems.
The INA226 has more family members, even one with three shunt inputs.
Yes, I have a similar species. But I still don't understand the need for isolation. There's no electrical reason for a large amount of current to be flowing into the (high impedance) device doing the measuring, nor is there any reason for a voltage above the specified range, so long as the current through the shunt is kept within range. I'm not inventing any new concepts here, just applying it to my own device.
Consider the circuit you are monitoring. IF it is not COMPLETELY isolated from ground, your Arduino will be a great path to ground for the circuit you are measuring.
A thought exercise for the OP:
In your 15 volt application, If you place the shunt in the high side, the ADS1115 will blow up, the INA226 will not.
Why is this? This is why the issue of isolation was raised.
PS: Your original statement about the input to the ADS1115 being -0.075 volts. By convention, that is incorrect, the sign is positive, not negative. The signal is above vdd/gnd, not below.
The current being monitored can flow in either direction, generating a potential anywhere between -0.075V to +0.075V across the shunt. Ordinarily, one might be concerned about feeding a voltage lower than GND into a single-supply device, but the datasheet explicitly allows as low as VDD(GND) - 0.3V. After reading the data sheet for the INA226, it has the same specification and goes into some detail on various system faults that might violate this.
The datasheet of the ADS does not specifically say the A/D can measure below ground, just that the pins can be up to 0.3volt below ground.
You dont say how accurate your shunts are, nor how precisely (ie to what degree of precision) you require to measure.
However since they are cheap I'd expect not very accurate, so this may not be an issue.
Also you dont say what is consuming that power. So eg a heater would not cause dc transients whereas a commutated device (eg motor) would.
Ignoring these concerns the problem you have is isolation - ie producing a voltage that is the difference between the + and - ends of the shunt, but ground referenced.
I have a page here that describes one way of doing this.
However since you express a preference for minimum component count, I'd recommend you use instrumentation amplifiers eg AD620
This has terminals that allow you to set the gain with 1 external resisitor.
Here is a schematic
+Vs1 needs ideally to be a little more than the 15V supply, and -Vs2 a little (say 3v) below the high current 0V.
(in fact if you are high-side sensing as shown you can use the 0V rail)
the in-amp takes the difference voltage acros the shunt terminals, amplifies it and ground-references it.
you can then use the arduino's built-in ADC to measure the resulting voltage.
Rp1,2 are just to provide protection in the event of an open circuit condition, sowhile they MAY not be needed I'd put 'em in. Say 47k. Will not affect the reading.
Your supplies at the low current (arduino) side need to be seperate from the high-current side, and you will need to plan where to connect to provide a common ground.