I2C on mismatched voltages - should I have any concerns?

My master (or at least for sake of this post I'll call it master) is an esp32 with the standard 3.3v I2C voltage. My slave is an MCP4725 similar to this. I have it connected to the servo where I removed the internal feedback pot and put this DAC in it's place, v++, OUT and GND.

The issue I have is that I cannot connect the DAC voltage to the esp32 voltage, nor their grounds, without getting sputtering results from the servo that the DAC is connected to. I have to connect the DAC to the voltage regulator of the servo internal mcu (measuring 3.0V) only. Only the SDA and SCL wires are interconnected between the 2. Then Everything works great.

Is there any reason to be concerned about the DAC reference voltage being separate from the ESP32? And the voltages being different? I know I can use one of those level shifters but I'm trying to keep component count low and this seems to be working great.

One way or another you have to have common GND. Are they powered from same supply?
You could post a scheme of your wiring.

Yes, sorry I'm not sure how to draw it all out easily, but I guess they are currently interconnected at the Wall AC.

The Servo is on its own 7.4V Power supply plugged into my wall AC, and the ESP32 is just being powered from the PC USB.

In the end I will be using a 7.4V lipo battery to the servo, and have the ESP32 powered from it (with a 3.3V or maybe 5V voltage regulator)

You use that thing at the end of your arm to do the drawing and labelling, then take a picture (make sure the picture is legible)

SA responses need not apply

I don't consider some AC earth connection as proper common ground. Also, going with battery would change the situation. There's too much to guess here without illustration of your wiring.

ps. are you aware of dac that Esp32 has?

Try this:

Unlikely, and a guess doesn't help, they need to be connected.

Without a schematic and a definite description of what is actually connected, without guesses, it is difficult for anyone to help. Clear photos are useful too.

Yes agreed, and those are 2 of my concerns that brought me here.

I'm working on drawing it out. It's taking me a bit but I'm taking his advice though he could have worded it less condescendingly. Ahh the internet ....

anyways, brb with a sketch

Your description is not so clear we forum folks can understand what you have nor why you have it.

The DC common (aka ground) of the ESP32, DAC and servo supply MUST all be connected together.

It must be wired so when the motor its current path does not cause and electrical noise in the DAC to servo wiring.

Ignore the AS5600 for now, its also connected to the same power as the MSP4725

Thank you for all the input, I'm not sure how it worked so well before based on Google and your responses, but I wired in a ground coming off the main 7.4V PSU output, and connected it to the esp32 ground.

My main concern was the servo jitter issues would return. So far so good, no servo issues (yay!), and now I have a common ground.

Anyway This is now the most precise (and powerful) servo I own, and it also doesnt jump into position like normal servos when first powered up.