Hi folks,
Why would the magic smoke escape when I connected up the solar panels first.
The voltage on the IDS1115 should be around 4.8V. It's measuring differential.
The solar panels do not share ground with the supply.
When I connected up the PV cells there was an audible pop and that smell. 
I thought this would have been an OK way to measure the PV voltage.
Comments welcome.
Thanks.
Crispin:
The solar panels do not share ground with the supply.
There is your problem. Without a reference the voltage can be anything. But it should not exceed Vcc, without reference, how can you be sure? Give them a shared GND
And to be sure, also check the voltage to after the divider with a DMM.
And note, the ADC might not be very happy it there is a voltage on it's input (from PV) but no Vcc.
Why would the magic smoke escape when I connected up the solar panels first.
What septillion said plus connecting a voltage input to an unpowered chip is a great way to blow up the chip. It’s tried and tested.
Thanks for the replies.
So how can I give it a common ground? Would that not create a ground loop? I intend doing this on a larger array so want to understand it correctly.
PV powers PV controller
Controller charges battery
Battery powers Arduino
If I start mucking about with grounds I could create more trouble?
Noted on wiring up the DAC before the power. Will be more careful 
How does this guy's example differ what what I am doing?
No common ground (although in his case you could easily have done that)
How does this guy's example differ what what I am doing?
He is using lower voltages, but he still needs a common ground to work correctly. Basically what he is doing is rubbish as well. He doesn’t understand much.
Would that not create a ground loop?
I am not sure what you understand by a ground loop and why it should be bad but no.
how can I give it a common ground?
Connect the Negitave of the 20V charger to the ground of the Arduino.
Ahh, that is new info. All we see is a separate PV and a separate battery 
If something is connected between the two you need to know how And that really depends on the kind of charger Some have shared negative, some share positive etc. So unlike you're schematic, there is some sort of connection between the two
Hi,
Is there a reason to measure open circuit PV volts?
When your charger connects the PV to the battery, the PV voltage will probaby be only about 0.3V different to the battery volts.
As @Septillion has said, there are many types of charge controllers and some switch positive others switch negative.
Tom...
I double checked the common ground - there is none. The positive though seems to be common between the PV input and the battery charger.
So, updated diagram with the charge controller in between.
Am i going down the wrong path? I want to measure PV voltage (which I believe is quite higher than the battery) as well as battery voltage and current drain (phase two)
thanks.
Hi,
What is your magical PV charger, link to specs/data please.
The only instance that I know of that the PV voltage will be much higher than battery volts while the charger is in charge mode is if the charger is a MPPT type.
Please on your circuit LABEL your pinouts, not just numbers, they have names like, GND, Vcc, Batt+, P1, P2 etc.
How big is the PV, link to specs/data.
How big is the Battery, link to specs/data.
OPs new circuit!!!!
Thanks.. Tom..
Hi,
Have you looked at the specs of the ADS1115, can it measure negative voltages?
I double checked the common ground - there is none. The positive though seems to be common between the PV input and the battery charger.
Because you are effectively measuring the two differential voltages with respect to the positive supply.
When the charger drops out, the negative terminal of the PV will go to full open circuit volts which will be more than the system volts.
This means the ADS1115 inputs, or at lest one of them will go negative with respect to the ADS1115 gnd.
With a DMM, and the ADS1115 out of circuit, and the charger in the OFF condition, measure the open circuit volts of your PV, then measure the voltages where the ADS1115 is connected, ALL of the connections on the ADS with respect to gnd.
Thanks.. Tom...
Its quite common for Solar MPPT chargers to have a common positive ground.
OP needs to check this , and if this is the case the circuit as is will not work.
Most likley 20V has been fed into the ADCs input with no ground referance, which has let the smoke out.
