[circuit_flash]

Hi, I made a chain positioner powered by a stepper motor. One arduino mega feeds a pulse train to advance the chain a certain speed and distance. Another mode of operation is scrolling or jogging the chain to its starting position. When in pulse ready mode, it allows a second arduino uno to tell it to start the pulse train; when the pulse train is finished, it hands control back to the arduino uno.

Everything works great, except occasionally it starts on its own or does a second unwanted chain advancement. I put a scope on the pins and found 5 volt peak to peak oscillation at about 16KHz interval on the arduino power rails as well as all input and output pins. The stepper controller, when enabled, not even moving, emits a 16KHz square wave detected by a scope probe brought near the stepper controller. I have disconnected all wiring from the stepper controller to the arduinos, the connections are through opto-isolators anyway, grounded the machine and all power supplies, changed power supplies, etc. Also tried 0.1 uF disc capacitors everywhere. The microcontroller arduinos are powered separately from the stepper controller. There is no wiring connection between the arduinos and the stepper controller at this time but the noise gets into the arduinos.

Any other ideas to suppress this noise? The project would run great except for this interference. Thanks for any input about this problem.

[Vlod]

are you using pull up/down resistors? I think arduino has pull up option on all digital inputs/outputs. If you are using positive impulse to control then put 10k resistors on each control wire to the ground. It would pull it close to zero.
If you are using negative pulse to control then activate pull up resistors in you program

[MakingGears]

16kHz sounds like the PWM frequency of the stepper driver. In other words, even when stopped in one position, the stepper is supplied with power so it won't move. That power is current regulated by a switching regulator (aka PWM). If it weren't current regulated, then you could get too much current and overheat the motor. If it weren't PWM regulated, the electronics would get too hot.

If the 16kHz noise is false triggering the Arduino, then it could be that your grounds are coupling. The ground for the motor power driver (switching regulator, PWM) should be a separate wire back to the power supply. If you use the same long ground wire for the Arduino and for the stepper driver, then every time the stepper draws current, it will put a spike of current in ground, and that will cause the ground voltage to bounce up.

Another way to reduce this problem is to use bypass capacitors. You need a capacitor with large enough value to supply the surge of current and not have the voltage sag. It also needs to be a fast enough capacitor to handle the sharp edge of the current waveform. Often, this is done with an aluminum electrolytic capacitor (good for high current) in parallel with a small ceramic capacitor (good for fast edges). It is important to put the capacitors in the right place - where the power and ground branch to the Arduino and to the driver. The capacitors will keep that location relatively noise free.

Please forgive me if this is too basic for you or if you need more explanation.