Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Foreseeing noise

Status
Not open for further replies.
I was wondering what actually produced noise, I mean, I know different types of noise, but I don't know how to foresee noise.

For example, I'm building a circuit that will take an input signal (From an instrument) and then you choose a number of switches and it goes down the route you want it to.

But, how can I tell that when I bring it into the real world it wont put noise into the signal and so on?
 
By its very nature noise is random. So you can't 'foresee' exactly what it will be. You can however take steps, such as electrical screening and filtering, to reduce the likelihood of noise pick-up by circuits and to reduce the effects of any picked-up noise.
 
Your circuit is one that might send up warnings.
Signal to noise ratio: You may have a low amplitude signal coming from the instrument so any noise pick up can be significant.
High input impedance: Your signal source may have a high input impedance. As such coupling from a noise source can also have a higher impedance and cause you problems. For example a pwm trace next to your input.
Noise sources: Signals with fast switching speeds and high voltage or current couple best to high impedance circuits. They can also cause ground shifts that are amplified.
I'm only guessing at your application but it may be a low level high impedance pickup. As such it probably comes with shielded cable to try to keep noise out. So you need to be very careful to keep the exposed leads in your box as short as possible. Or if you will amplify the signal, amplify it before you switch it to improve signal to noise and lower the impedance.
 
By its very nature noise is random. So you can't 'foresee' exactly what it will be. You can however take steps, such as electrical screening and filtering, to reduce the likelihood of noise pick-up by circuits and to reduce the effects of any picked-up noise.

Ahh okay, thanks. :) Best get my hands on an oscilloscope.

Your circuit is one that might send up warnings.
Signal to noise ratio: You may have a low amplitude signal coming from the instrument so any noise pick up can be significant.
High input impedance: Your signal source may have a high input impedance. As such coupling from a noise source can also have a higher impedance and cause you problems. For example a pwm trace next to your input.
Noise sources: Signals with fast switching speeds and high voltage or current couple best to high impedance circuits. They can also cause ground shifts that are amplified.
I'm only guessing at your application but it may be a low level high impedance pickup. As such it probably comes with shielded cable to try to keep noise out. So you need to be very careful to keep the exposed leads in your box as short as possible. Or if you will amplify the signal, amplify it before you switch it to improve signal to noise and lower the impedance.

Hmm, OK, thanks a lot. I read in on an old schematic that a capacitor was placed after the input to reduce noise some how. Is this a load of "poop"?
 
a capacitor was placed after the input to reduce noise some how
That's possible, depending on the type of noise.
If your signal is passed to the input of, say, an amplifier via a resistor and the amp input has a capacitor connected to earth then 'low' frequencies get passed more or less unaffected to the amp whereas 'high' frequencies, typical of many kinds of interference, will be reduced.
The values of the resistor and capacitor will determine what is 'high' and what is 'low'.
 
Noise can come from radiation from the power line (primarily 50/60 or 100/120Hz), any other piece of electrical equipment, or even RF from radio or TV transmitters.

A series resistor and capacitor to ground (low-pass filter) on the input can reduce high frequency noise, no "poop" there You choose the filter frequency to be slightly higher than the highest signal frequency you are trying to detect. It will help reduce noise above the signal frequency, but has no effect on noise frequencies that are equal to or below the signal frequency.
 
Hmm, OK, thanks a lot. I read in on an old schematic that a capacitor was placed after the input to reduce noise some how. Is this a load of "poop"?

If the noise was a much higher frequency than the signal the cap would have the effect of lowering the impedance (to the noise) and thus reduce the noise pickup.
 
Noise can come from radiation from the power line (primarily 50/60 or 100/120Hz), any other piece of electrical equipment, or even RF from radio or TV transmitters.

A series resistor and capacitor to ground (low-pass filter) on the input can reduce high frequency noise, no "poop" there You choose the filter frequency to be slightly higher than the highest signal frequency you are trying to detect. It will help reduce noise above the signal frequency, but has no effect on noise frequencies that are equal to or below the signal frequency.

If the noise was a much higher frequency than the signal the cap would have the effect of lowering the impedance (to the noise) and thus reduce the noise pickup.

That's possible, depending on the type of noise.
If your signal is passed to the input of, say, an amplifier via a resistor and the amp input has a capacitor connected to earth then 'low' frequencies get passed more or less unaffected to the amp whereas 'high' frequencies, typical of many kinds of interference, will be reduced.
The values of the resistor and capacitor will determine what is 'high' and what is 'low'.

Mmm, thanks a lot. Some very succulent information here. :D
 
With out understanding you signal and your noise I can not do much.
I have removed "scratch and pop" from records using several methods.
Tape hiss I know what to do about.
50/60hz hum can be removed.
Random noise and all types of signal....no.
 
I understand. I'm just making it in a simulator first and I was wondering if there were any just... General procedures to cutting noise?

Thanks for your help though. :)
 
If you're just simulating it then what are you using to model the noise source?
 
If you're just simulating it then what are you using to model the noise source?

alec,

These 'noise' samples I created using Audacity Beta (Unicode) free download ,includes a simple demo asc with filter.

The Audacity program is also useful for ring tones and Chirps if you are interested.:D
 

Attachments

  • noise_samples1.zip
    403.4 KB · Views: 73
  • AAesp02.gif
    AAesp02.gif
    49.2 KB · Views: 114
Last edited:
Thanks, Eric. I was just wondering what the OP was using and would have suggested something similar. I've been using Audacity for a couple of years and suggested it a while back as a white/pink noise source to someone experimenting with FIR filters.
 
there are also different noise sources, generally external and internal noise sources. seems like everybody here has been thinking in terms of external noise sources. external noise sources are the easiest ones to deal with, using shielding, filtering, etc... there are also internal noise sources, and these are more difficult to deal with. resistors and active devices (transistors, etc) all generate a certain amount of internal noise. in fact, a zener diode (or a B-E junction in a transistor biased into it's zener conduction region) makes a very good noise source (if you want noise on purpose).
resistors create noise proportional to their resistance. transistors, FETs and vacuum tubes generate noise when in conduction. dealing with these noise sources and minimizing their effects is almost an art in itself.

some useful search terms would be:
shot noise
flicker noise
johnson noise
1/f noise
avalanche noise
 
there are also different noise sources, generally external and internal noise sources. seems like everybody here has been thinking in terms of external noise sources. external noise sources are the easiest ones to deal with, using shielding, filtering, etc... there are also internal noise sources, and these are more difficult to deal with. resistors and active devices (transistors, etc) all generate a certain amount of internal noise. in fact, a zener diode (or a B-E junction in a transistor biased into it's zener conduction region) makes a very good noise source (if you want noise on purpose).
resistors create noise proportional to their resistance. transistors, FETs and vacuum tubes generate noise when in conduction. dealing with these noise sources and minimizing their effects is almost an art in itself.

some useful search terms would be:
shot noise
flicker noise
johnson noise
1/f noise
avalanche noise

Thanks a lot buddy. Looks like I've got even more to study now. ^^ :D
 
....there's also the noise that passes as modern music!
 
that's not actually noise, just poorly assembled audio.....

that seems to be one of the constants of the history of music, one generation dislikes the music of the next..... and the process is accelerated by modern technology.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top