Problem with arduino analog input, not a pure sine wave signal

Hi,

The waveform may look "jerky" but probably only has 1 or 2 % distortion.

It looks like the signal is noisy. Note how it is less apparent where the signal is steep. It looks to me like the noise is at a higher frequency than your sampling frequency, which makes it hard to see the noise frequency.

What kind of signal conditioning circuit are you using?

Do you have a transformer? Is there a low-pass filter? Are there large value resistors in the circuit? Are the wire leads separated and long?

Can you post a picture of the setup?

If you take samples at a much faster rate, (or look at it with an oscilloscope) you may be able to see the frequency of the noise.

It should be fairly easy to implement a low-pass filter on the signal to attenuate the noise. You could even implement a digital filter in software.

Hello,

I do have a transformer, it's a 230V prim, and 6V sek, 1.5VA , but I don't have a filter, my circuit is just a transformer with a voltage divider, and an arduino Uno, I'll upload a picture of my setup, and a schematic which shows resistors values. As you can see I don't have an original arduino, but so far it has worked perfectly.

Thank you for the reply.

I have found a cause of my problem. The arduino that I was using wasn't an original arduino, and when I replaced it with an original one it worked perfectly. Thank you Douglas Wong for your help.

That's a strange cause. Your non-original has the same microcontroller (=the same ADC) and is also xtal clocked.

Looking at your LabVIEW screen, It looks like sampling jitter more than ADC errors (although the controller shouldn't have jitter issues with a 50-60 Hz signal).

The xtal on your clone has to be very bad to cause this type of sample artifacts.

The other thing that could cause this is the UART implementation on the clone, or different baud settings when switching between clone and genuine.

You're using serial comms to exchange data with the PC. If that doesn't go fast enough, it can explain the jitter.

As another possibility, could it be down to the Arduino A/D being ratiometric, with the supply as the reference? Poor layout and decoupling, which is a very likely difference between the two, would put the poster in the situation where they were measuring the reference variations (supply noise) as well as the signal.

I was thinking about noise too, but with the approx. 11K load between analog in and GND, the noise has to be fairly strong to impact the signal that much.

edit: if I recall my theory well, the input impedance is less than 1K and not approx. 11K, because impedance-wise, the resistance between 3V3 and GND is close to 0 ohm. It would be good to correct me if I'm off here...

edit 2: I'm not sure what the role is of the lower 10K resistor. It doesn't play a role in the voltage divider because it's a load between the fixed 3V3 and GND.

The 10k resistor between a 3.3v port, and a GND port is to prevent any current flow towards the ground. The arduino that i bought was pretty cheap so that might explain my previous problem.

The 10k resistor between a 3.3v port, and a GND port is to prevent any current flow towards the ground. The arduino that i bought was pretty cheap so that might explain my previous problem.