• 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.

Noise in Passive Components

Status
Not open for further replies.

solis365

New Member
Hi everyone,

I'm having kind of a tough time finding info on google for this, wikipedia is a bit vague as well.

I'm wondering about noise sources from passive components. For example, I know resistors exhibit Johnson Noise (thermal) described as

S = 4kTR (units are W/ √Hz if I recall correctly)

are there any other noise sources in a resistor not covered by the johnson noise? even extremely low ones, I probably need to know about them. Especially flicker (1/f) noise, I cannot have anything with a high flicker noise characteristic.


I am also wondering about capacitors. They have a Johnson Noise as well, but anything else?

I just found a wiki article about kTC noise, which is the noise in an RC filter. Apparently R drops out, but how would alternative noise sources in both the cap and the resistor affect the noise in a filter?

I am working on making an extremely low-noise RC Low Pass filter using just one capacitor (roll off is not important). I need to pick components that give the proper corner frequency and the lowest noise, so I will be looking at the noise in different types of resistors and capacitors and trying to pick the right values for R and C while minimizing noise.


thanks
 

Speakerguy

Active Member
I think thin film and metal film resistors have the lowest noise. For caps you are probably looking at film (polypropylene, teflon) but I have no data to back that up.

Are you also worried about other nonlinearities besides noise? If so there's temp coefficients, voltage coefficients, and other stuff to worry about.
 

crutschow

Well-Known Member
Most Helpful Member
Carbon resistors can have a noise component proportional to the current through them and may also have a 1/f component. Metal film and wire-wound resistors generally have thermal resistive noise only, which is the theoretical minimum you can have. Wire-wound resistors should be the lowest noise.

The thermal noise associated with capacitors is actually due to resistance in the circuit even though it doesn't show in the equation. There is a gain-bandwidth effect of the resistor capacitor combination which causes the resistor value to drop out of the equation leaving √(kT/C) as the value of the noise*. The capacitor noise is essential independent of the type of capacitor although leakage, such as electrolytics have, can contribute noise. Film type capacitors are more stable, so are often used for frequency shaping circuits.

*There actually was an experiment of an RC circuit where the temperature the capacitor was varied but the resistor was maintained at a constant temperature. The noise did not vary with temperature, proving that the resistor was the source of the thermal noise. So the T in the kT/C equation is the temperature of the resistor in the circuit, not the capacitor.
 
Last edited:

solis365

New Member
I need a low corner frequency, about .6 Hz. So I will be needing something thats fairly large, either the capacitor or the resistor, or medium on both...

since the noise power density is (kT)/C, where T is the temp of the resistor, would it be better to get a large cap? the problem is the only big caps are electrolytics, which are noisy...


not really concerned about anything but noise.
 
Status
Not open for further replies.

Latest threads

EE World Online Articles

Loading
Top