signal interference issue

[Crsus]

Hello, how do I eliminate the glitches in the signal generated by the encoder? There is motor interference.

[Whales]

Can you provide photographs of the motor (induction?  brushed?), motor driver (VFD?), motor wiring, encoder and encoder wiring?  Photos of all 5 items might allow us to provide specific advice.  Otherwise we can only guess or give generic advice.


Option 1: Prevent the interference from being made.  Modify or replace the motor or motor driver.

Option 2: Prevent the interference from coupling from the motor wiring to the encoder wiring.  Shielding, differential pairs, etc.

Option 3: Reject the interfering spikes with low-pass filters.  "RC circuits", ferrites, etc. 

[Crsus]

The motor uses a brushless DC motor (BLDC), which is powered through a workstation box. The encoder signals are connected to the workstation for measurement counting. I measured the power signal coming out of the box, and there is interference accompanying the power signal.

[Smokey]

That's why motor encoder signals are differential

[Crsus]

Can an RC circuit be used to improve?

[Whales]

Thankyou for the photos, they help a lot

The motor uses a brushless DC motor (BLDC)

The motor you photographed looks like a brushed DC motor, not a brushless.  I can only see two wires going to it (red and black) and this online listing for a WS-38ZYT64-R shows a brushed motor.

Brushed motors tend to create quite a bit of noise.  You may want to try:

(1) Earthing the case of the motor.  Connect a thick wire between the case and the green terminal on your power supply.

(2) Placing a small (1nF to 10nF?) film or ceramic capacitor across the two terminals on the motor, as close to the motor as possible.

The encoder signals are connected to the workstation for measurement counting.

If you have a look at the pins of your encoder circuit board you will notice that the Z, A and B signals are split into positive (+) and negative (-) versions.  This likely means they are differential signalling.  Differential signalling gives you very good noise immunity if you use it correctly. 

I suspect that the designers of this encoder knew that noise was a serious issue, so they provided differential signalling as a fix. 

You will definitely want to:

(3) Decode the + and - signals with a differential receiver circuit.  You can use off the shelf "differential receiver" ICs or make you own "differential amplifier" circuit with opamps. 

Can an RC circuit be used to improve?

EDIT: Yes, you will want to implement this at the same time as the differential decoder.  Slow opamps might (?) misbehave without RC filtering.

If you are not sure how to do this then first try some differential amplifier opamp circuits, then try adding RC filters to the + & - lines before they enter these circuits.

[eTobey]

Put a cap from signal to gnd, and then a ~ 1k resistor inline from signal to the output ( on each signal ).
Size depends on how fast the motor runs vs. how good they should be eradicated.