I got a k-type thermocouple installed which I measure with a MAX31856 and in combination a DAC (MCP4725) i want to make a PID-controller. When I only connect the MAX31856 the measurement works fine, but when i also connect the MCP4725, which is hooked up to a dc power supply (6261B), i receive a way too high measurement. I figured out that the problem may be, that the Grounds are not on the same voltage. If i measure only the grounds of the MAX31856 (powered by an Arduino Nano every) and the Ground of the MCP4725 (= Ground of the 6261B) i get a reading of 1V. Now when I connect these two together on the same Ground i get the inaccurate readings.
Is there a way to isolate these 2 Signals, or to get rid of the difference in the ground voltage? Or better how to isolate the MAX31856 completely from the MCP4725 lines?
It sounds like a bad connection, per your schematic all grounds are connected together. Did you place the reference lead (probably black) of your meter to ground?
Hi,
yes i when i connect all the grounds together it creates this problem. But from my understanding this should be connected together to get a common ground.
The 1V difference is measured when from the disconnected state, so the arduino ground vs. the power supply ground (power supply ground is then 1V below arduino ground). when connected together it is the same but then my MAX31856 measures wrong! (I suppose the difference of the Ground voltage is responsible)
Power off the circuit.
With a DMM, Check for continuity between GND pins.
The DMM beeps so there is continuity given between the GND pins, if these are connected and powered off
What is the actual resistance between GND pins.
All measured GND values must be 0 Ω
It was rather low, about 10 Ohm if i remember correctly
All measured GND values must be 0 Ω
Ok i make a quick double check and measure it again, just to be sure it was 10 Ohm
Seems like i remembered it falsely, I measured 0.1 Ohm (lowest resolution my DMM can measure)
Show us good images of your actual wiring.
Sadly I am not in the lab anymore since it was already 8pm here. I can make a picture at Tueday. But i also got a Fritzing wiring plan which should help a bit, it is wired EXACTLY like this.
Thermocouple is a shielded 2 wire and the wire to the DC supply is a shielded 1 wire. The shielded part is also the ground for the cable (not the best solution but was already installed before i started to work on this project). At the DC Power Supply part it is wired like this:
the Voltage source is the DAC (MCP4725), and the R_L the heating element. Also the power draw is stated in the data sheet and the MCP4725 can provide sufficient.
If they are not connected together they are not common grounds. Deliberately not connecting them together is a big mistake.
That is why you connect them together. Without connecting them together nothing will ever work correctly.
No it is not responsible for your problems. Your problem is somewhere else.
How wrong? What is the difference between what you measure with a meter and what reading you get?
It is good that you used a schematic, but you did not show the separate power sources you used. Can you amend this to show that.
There are many things that can cause a fault like you are getting, the most common cause is bad layout of your wiring. Where ground currents between devices cause what is called a "ground lift". This can be fixed by using star wiring, that is making all the grounds from each device meet in one place. The Arduino with its multiple ground connections is a good place to wire each ground to, instead of chaining them.
This is why we asked for a photograph of your wiring.
I also notice that you have NO pull up resistors on the I2C lines. That is just asking for trouble.
I know this is what i meant with "this should be connected together"
It measures a constant value (about 320°C) and this does not fluctuate. With a DMM I get a reading off 0V when i measure the 2 wires (thermocouple voltage is too low to be measured with a DMM typically). If i measure without the connection to the 6261B I get a reading of 24.xx°C +- ~0.02°C, so it fluactuates a bit which is a normal behaviour espacially in that order of magnitude.
One power source is the USB connection from a computer, (a grounded PC). The 6261B is also hooked up to a power source.
Yes but i sadly not in the lab anymore and it is closed over the weekend, that was the best i can give you at the moment. I double checked each wire so there are no lose wires and the plan is 1:1 how i hooked up all the wires. (Only USB from PC missing, which powers the board)
Pullup resistors are included in the MCP4725 board! 
Also the DAC works fine when i tested just this board
That sounds like something is saturating.
Yes, but as I said without connecting the grounds measurements are meaningless.
This worries me because that is shown without a ground. Sure there is a GND connection shown, but from the diagram this looks like a "mains power" ground. This is a different thing to the "common signal" ground. Both are called ground but they are different things.
I would be inclined to treat the -out pin (-5V) as the signal ground, and ignore the mains ground. I notice that the voltage source, your MCP4725 treats this point as ground. Which is a good thing otherwise it would be subject to over voltage, if you actually had -5V at this point.
Do you need a bidirectional supply here? Indeed have you actually got one? Otherwise what is the point of a negative voltage here?
Hi, @ardu_user
Can you please post a schematic?
Not a Fritzy.
Reverse engineer your project, post a copy of your circuit, a picture of a hand drawn circuit in jpg, png?
Hand drawn and photographed is perfectly acceptable.
Please include ALL hardware, power supplies, component names and pin labels.
Thanks.. Tom.. 
Sounds to me like a grounded thermocouple with significant current flowing in the sheath.
It is connected to the Ground from the arduino (USB) then. It ist Just without the mcp4725 Board.
Ok how can i treat them properly? I guess since these are not the same i need to seperate them somehow.
It is not bidirectional the output of the 6261B is voltage controlled through the voltage given by the DAC
Yes i will draw a complete circuit diagramm. I am gonna to need some time for this will post it later when i find time for it
Checked this before there is No current flow in the thermocouple