Hello,
I have a basic setup with arduino nano, LM2596 and delphi sensor. The signal of my Wheel Angle Sensor is unstable since I plug the usb in the arduino nano.
I tested with a multimeter and the signal is stable and as soon as I plug the usb on the nano the voltage starts oscillating. Any idea what can cause this? I don’t use the ADS1115 at the moment because mine seems defect so I connected the signal to arduino pin A0.
Video : AgOpenGPS WAS unstable - YouTube
So you have pcb setup and connect to aog via USB cable.
I think problem is unstable voltage on the Nano.
Let´s say your LM2596 give 4,8 V and the nano ´s own power supply would give 5,2 V now the power supply is all trough the USB connection (included 5v to pcb board) and the voltage regulator on the nano turns on and off because of overload/overheating.( Feel with finger)
Either cut 5 V from USB (or cut line from usb to voltage regulator on Nano) or turn up voltage on LM2596 a little
I don’t have a PCB but you’re probably right. My LM2596 is set at 5.0V, how much do you think I can increase? It seems to be the easiest way to solve my issue.
If I am right you should be able to measure a voltage shift on the 5v pin , then increase until V is stable.
max looks to be 5,5V on the Vcc but copy below is about arduino uno.
Recommended: 7~ 12V . Absolute: 6~20V. Input/Output (I/O) pins: -0.5V to +5.5V (the actual max is “Vcc + 0.5V,” which for a 5V Arduino, is +5.5V) (Note 1)22. jun. 2016
Got it, thanks ! First I will check the difference between the voltage of USB power and the output of lm2596, just to confirm.
Where do you supply the 5 V?
When using pcb nano gets power on 5 Vcc (5V pin)
I you supply on the 5 Vin you should supply minimum 7 V (then nano rcan regulate to 5V)
Here it also tells that max draw on nano is 500 mA out of 5Vcc
Below it seeems that 6 V is the max on the 5V pin (method nr 5 to destroy arduino)
Apply a voltage of 6V or higher to the 5V connector pin. Many components on the Arduino will be destroyed, and this voltage can also appear on your computer’s USB port, possibly damaging it.
I supply the 5v on the 5V pin not the Vcc pin.
Just checked the voltage when the nano is connected only to USB (4.7v) then disconnected the USB and send power through the LM2596 the voltage is 5.0v
Do you think I should increase the voltage on LM2596 ?
Not at the moment—
First clarify what pins we are talking about:
In the attached link it is named like this.
You must measure on +5v
When measuring with USB power nothing else(but your multimeter) must be on +5v
Then measure voltage on 5v out from LM2596 and nothing else connected to that.
If you supply power at Vin it must be over 7v
LM2596 alone, only the 12v input from tractor and the multimeter on output pins, voltage is 5.03v
Arduino alone, nothing connected except USB and the multimeter on pin number 27 (+5volts) and pin 29 (ground), voltage is 4.75v
I don´t know if difference in voltage that way can cause problems, as the 5v on 5V+ should give the nano 5V all over.
Do you have common GND ?
Perhaps it is time for a drawing of your wiring.
Try with power to nano ONLY from USB, no power to 5V+ or Vin , but power the rest of your stuff from LM2596 AND same GND for all.
Yes I connected all the GND together.
That’s how my setup looks like:
Note that I use D5 and D6 for PWM and DIR but this has been changed in the arduino code, no impact on this issue.
Very nice drawing, and yes that should work correct.
I just hope the test without power from LM2596 to nano, will give at good result.
Do you tray put decoupling capacitor 10n to 100n on arduino input pin from WAS.
First remove that buck converter and set some linear voltage regulator such as lm7805
100% agreed. Noisy. Personally I would keep the motor controller well away from the sensing electronics too.
Tried to remove the power between LM2596 and nano. Still the same.
I guess the buck converter is the issue and I have to use a LM7805. Anyway, I’ve ordered the PCB and parts.
Did you ever measure the voltage coming from the usb ? you can easily now that power from LM2596 to nano has been taken away.
USB might have a noisy buck converter, and you have already confirmed vith multimeter, that voltage from WAS out is stable when you only use your LM2596 (your first post at the top)
The 5V from USB is not regulated at the input to NANO (only the part that goes to 3.3 V)
That means the nano AD converter may get shifting voltage from USB voltage supply.
Voltage coming from USB is 4.75V. Didn’t catch what you mean, is my nano faulty? Hope this will work on the PCB.
Just wanted you to confirm that voltage from USB is stable an not fluctuating, and you say 4.75 V (and I suppose also staying there)
No I don´t think Nano is damaged.
When you get PCB you should not use Nano AD converter anyway, but the ADC 1115 instead. So problem solved 
Brian has always said that ADC 1115 is better than Nano AD converter, but that was mainly because of ADC 1115 posible 16 bit opposed to Nano 12 bit conversion.
Yes I know but I was not able to get the ADS1115 working, that’s why I tried with nano ADC.
Just tried again with the I2C Scanner and was unable to find the ADS1115. Guess what… I switched the SDA/SCL cables with 2 other ones and it works. 
All this time wasted just for 2 wires… Can’t wait for my PCB.