Hello, thank you for your help. I upload the file. Please note that you don’t see ILS7660, you see TMP01 because I don’t have the ICL7660 library. Please disconsider TMP01 - That is the reason I poste a picture on my first post.
It is possible to achieve 2,5 V on LM317, I did it using the informations on this site: https://www.eleccircuit.com/lm317-voltage-regulators/
I did some suicide tests with my arduino nano. Example, with an LM7805 and ICL 7660 I got +5 and -5V. I put -5 V in the GND of the arduino and + 5V in the VIN and it incredibly worked, that is, in addition to running the software, it communicated with the computer and the 5V output of the arduino was used to power the MCP4725 (here I used unipolar mode) and ADS1115. In this case, the triangular ramp generated on the MCP4725 (unipolar mode) was between -4 V and -1 V vs GND (say, GND from my 9V source that I fed the LM7805). This seems to show that the hypothesis of your last paragraph did not materialize.
I played around with the LM317 and ICL7660. I actually got 2.5 V (and -2.5 V with the ICL 7660) but it didn’t work to power the Arduino.
With +3.5 and -3.5 V it was possible to power the arduino. The triangular scan generated in the MCP4725 unipolar mode was between -2.30 and +2.42 vs GND of the 9V source when the computer was turned on so that I could see the ADS1115 output between 0 and ~ 4.8 on the serial monitor, vs Arduino GND.
Interestingly, when I disconnect the computer, the voltage range generated by the MCP4725 is between -1.2 and 2.4 vs GND from the 9V source. I think the connection with the computer’s USB cable helps to power the arduino and so I get better results.
I think this last experiment is close to what I need but I would like to learn what you gave me to use the LD1085 (since I can get 2.5 V on the LM317) but I didn’t understand the design, my knowledge in electronics is so scarce like my electrical components. Is it possible to explain better your scheme?
Thank you so much.
Potenciostato-Arduino.fzz (47.7 KB)