Hello everyone,

I'm currently working on a project that involves controlling a servo motor with an Arduino, potentiometer and displaying servo angle on an LCD screen. I have very little experience when it comes to building electrical circuits. The entire circuit operates on a 5V DC battery. I'm looking to incorporate capacitors for noise filtering to ensure smooth operation and minimize any interference.

I've come across various recommendations regarding capacitor selection, but I'm not sure about the best solution for my circuit. Some sources suggest using a single capacitor, while others advocate for using multiple capacitors in parallel. Also, I'm uncertain about what capacitance value to choose and which type of capacitor would be most suitable for the job.

Here are a few questions I have:

1. Is it better to use one or two capacitors in my circuit for noise filtering?
What would be the advantages and disadvantages of each approach (I'd be interesting to know, but not the most important thing)?

2. How do I calculate the capacitance needed for effective noise filtering in my circuit? Are there any formulas or guidelines I can follow?

3. Which type of capacitor should I use - Tantalum, Aluminum electrolytic, or Ceramic - for noise filtering in my circuit?
What are the key differences between these capacitor types, and how do they affect their performance in noise filtering applications (I'd be interesting to know, but not the most important thing)?

4. Where do I place these capacitors for noise filtering? Before servo and before LCD, right?


I would greatly appreciate any insights or any advice at all you can provide for my project. Thank you in advance!

 

What type of servo? RC or other?
Rather than treat 'noise' with caps, it is first wise to use correct wiring and earth grounding techniques in order to create a low, noise-free system .
Do you have it mapped out in the form of a DWG you can post?

 

This is a rough first sketch, so I'm open to any suggestions. Servo is not remote controlled.

 

What type of servo? RC or other?
Rather than treat 'noise' with caps, it is first wise to use correct wiring and earth grounding techniques in order to create a low, noise-free system .
Do you have it mapped out in the form of a DWG you can post?

Also assume that the wires between servo and battery will be about 4 meters long, that's why I was worried about noise and want some noise filtering too.

 

So it is a RC servo?!
Just ensure you use cable with twisted conductors, preferably shielded.

 

So it is a RC servo?!
Just ensure you use cable with twisted conductors, preferably shielded.

Sorry, I didn't know what RC meant, but I'm focused on noise reduction with capacitors. Any insights about that would be appreciated.

 

What kind of supply are you using, they usually have large caps in them, also place a 0.1uf at on the power at each end of the servo run.

 

What kind of supply are you using, they usually have large caps in them, also place a 0.1uf at on the power at each end of the servo run.

The supply is a 12 V accumulator with 12 V to 5 V step-down converter. Now about the 0.1uF capacitor: How did you come to this capacitance? Are there guidelines for this or formulas for calculation? What kind of 0.1uF capacitor? Ceramic? Tantalum?
These are the things I am wondering about and like to know about in as much detail you are willing to provide.

 

This is a rough first sketch, so I'm open to any suggestions. Servo is not remote controlled.
View attachment 318816

If it operates on a 5V battery, why do you show it operating from a 12V battery with a DC-DC converter? That makes the noise filtering an entirely different problem.

Anyway, I would suggest connecting the servo directly to the DC-DC converter, then filtering between the 5V of the servo and the rest of the circuit.
A ferrite bead followed by10uF electrolytic and 100nF multilayer ceramic in parallel might be enough. Microcontrollers these days are not badly susceptible to supply noise.
Even better if the logic worked on 3.3V, then a 3.3V LDO regulator might take care of it.

 

If it operates on a 5V battery, why do you show it operating from a 12V battery with a DC-DC converter? That makes the noise filtering an entirely different problem.

Like I've said I don't know a lot about electrical circuits and I'm open to any advice. Power supply is as I've said in my last message

The supply is a 12 V accumulator with 12 V to 5 V step-down converter. Now about the 0.1uF capacitor: How did you come to this capacitance? Are there guidelines for this or formulas for calculation? What kind of 0.1uF capacitor? Ceramic? Tantalum?
These are the things I am wondering about and like to know about in as much detail you are willing to provide.

 

If it operates on a 5V battery, why do you show it operating from a 12V battery with a DC-DC converter? That makes the noise filtering an entirely different problem.

Anyway, I would suggest connecting the servo directly to the DC-DC converter, then filtering between the 5V of the servo and the rest of the circuit.
A ferrite bead followed by10uF electrolytic and 100nF multilayer ceramic in parallel might be enough. Microcontrollers these days are not badly susceptible to supply noise.
Even better if the logic worked on 3.3V, then a 3.3V LDO regulator might take care of it.

I'm not exactly sure what you mean by connecting directly, I'm pretty sure it IS going to be connected directly to the converter, not the Arduino. Also, what do you mean by filter between servo and the rest of the circuit, does that mean connecting capacitors before the servo or somewhere else in the circuit? And I'd love to know what is your reason for using

A ferrite bead followed by10uF electrolytic and 100nF multilayer ceramic in parallel

and how did you derive this?

 

Also see common mode/differential choke
https://www.coilcraft.com/en-us/edu...is a Common Mode,low-frequency signal to pass.

 

and how did you derive this?

from experience! Pretty much a "stock solution" - worth trying to see if it works.
The ferrite bead goes in series with the 5V supply. The servo is connected directly to the supply, the control circuitry is connected to the 5V supply via the ferrite bead. Then the capacitors go across the power supply at the control board.

 

Then the capacitors go across the power supply at the control board.

Where would that be exactly?

 

It's not always easy to predict. Sometime you discover the nature of the interference by the methods that work to remove it.
Basically, you need something in series with the supply that gives impedance at high frequencies, and something in parallel which presents a short circuit at high frequencies.
Sometimes a resistor will work for the series impedance, sometimes a ferrite bead works better, sometimes it requires an inductor, but as the inductor gets bigger its self capacitance stops it working at really high frequencies.
Similarly, large values of capacitance have too much self-inductance to be effective.

 

It's not always easy to predict. Sometime you discover the nature of the interference by the methods that work to remove it.
Basically, you need something in series with the supply that gives impedance at high frequencies, and something in parallel which presents a short circuit at high frequencies.
Sometimes a resistor will work for the series impedance, sometimes a ferrite bead works better, sometimes it requires an inductor, but as the inductor gets bigger its self capacitance stops it working at really high frequencies.
Similarly, large values of capacitance have too much self-inductance to be effective.

Got it, thanks. You've been of great help!

 

Got it, thanks. You've been of great help!

The core of the inductor also makes a difference. A ferrite core is less lossy, high frequency interference can reflect and just go somewhere else to cause trouble. An iron core inductor becomes lossy at high frequency, so it can turn the interference into heat, which prevents it from causing problems elsewhere. If there is too much high frequency on the supply you might end up with a hot inductor.
An iron powder toroid with a single layer winding should be the best, because of its low self-capacitance. It might even work better than if it had another layer of turns which would give it four times a much inductance.

 

 

Noise filters cover a wide range of possibilities. In order to cover them all, one would have to consider the application, frequency and power consumption of the entire circuit. However, here are some general guidelines, based on the circuit that you have provided

1) A linear power supply should have a bulk reservoir capacitor, typically about 100μF/16V electrolytic at the +5VDC output.

2) Every digital IC should have a 100nF ceramic capacitor between power (VDD, VCC, or +5V) and GND, as close as possible to the IC power pins.

3) Pay attention to wire connections. +5V and GND to the servo motor should go directly from the capacitor in (1) to the motor on a thick wire (22AWG - 16AWG).

4) If noise is a problem, you can use an RC or LC filter to the rest of the digital circuit as suggested by Ian. Try R = 33Ω and C = 10μF/16V electrolytic. Measure the voltage after the resistor (or across the resistor) to make sure that the resistor does not drop more that 0.5V.

 

That's a uselful video, but it doesn't deal with ESR. Everyone seems to think that low ESR is best, but the resultant filter has a high Q and therefore rings. Adding resistance in the guise of an electrolytic with high ESR can damp the ringing.
This is from Bob Pease's book (which you can download, probably illegally, depending on which country you are in, and what your country's government deems "fair use", from Pearl Hi-Fi).