capacitive coupling due to noise

noise can get coupled to the circuit via capacitive coupling. now suppose if we take an example of normal power lines(50/60 Hz) . now if we place our board(which has wires ~1m) can we get an approximate value of the coupling capacitor between the 50hz and the wire. the distance between the the board wires and the power lines ~3m.

i think it should be inversely proportional to distance between them(power line and board wire), and proportional to the current in the board wire, board wire length.but how to get an approximated value of the coupling capacitor.

Hi,

Capacitive coupling has nothing to do with the current flowing through the line.
--> this is inductive coupling

Capacitance.
It surely is below picofarads with the given distance. Don't expect noticable capacitive coupling effects at 50Hz.

Klaus

1. so in general the coupling is almost every time inductive coupling, because everywhere instruments will be in vicinity of other instruments.but

2. normally if we have two conducting wire we assume there can be a stray capacitance between them. please see page no 16 of the link.
http://www.analog.com/media/en/analo...16-number3.pdf

3. also why the capacitive coupling effects are negligible is it because of the separation between the two.

4. so suppose if i want to have capacitive coupling then instead of current sourec i need high charge storing source. power lines (in kv) can be a good example for such source. am i right

Hi,

For capacitive coupling you need two electrically conductive items with an AC voltage between them. There need to be an AC current return path between both.

Inductive coupling is caused by a magnetic field which is caused by an AC current.

Your link...is a 28 page scan of an 35 years old article.
It tells about "adjacent" wires and a "GND plane". If I understand correct, none is relevant for your question.

Let's do some calculations.
I assume the coupling capacitance is far below 1pF. But let's calculate with 1pF.
Then the impedance at 50Hz is about 3GOhms this is 3000MOhms. This is probably more than the isolation of wires...
With 230V...the coupling current is limited to 0.07uA.
Then where is the feedback path for this coupling current? If there is none, then the current can not flow..

And if there is a feedback path... what voltage can be generated? There need to be an impedance between the "receiving" wire and the return path...where the flowing current causes a voltage drop..
I don't see this in your description.

Klaus

Thanks Klaus,

can you refer me something(any books or paper) for noise reduction techniques.

can you tell how to be sure about type of coupling? also in home case scenario typically the coupling will be inductive?

Hi,

I referred to your headline....you are talking only about "capacitive coupling"
Therefore I kept discussion about inductive coupling short.

My opinion is....that 50Hz capacitve coupling across 3 meters will have about no effect.

There are other possible problems:
* maybe I misunderstood your explanation
* maybe we should discuss inductive coupling
* maybe we should talk about higher frequencies
* maybe we should talk about electromagnetic waves (also higher frequencies. But I'm no expert with HF)

In short:
* for capacitive coupling you need AC voltage and two electrodes (of any shape)
* for inductive coupling you need AC current and a somehow e lenghty wires

Klaus

Hi KlausST

basically i am struggling with the noise due to 50Hz pickup now looks like from what you explained and from my experiments till now, it may be because of inductive coupling.
i will explain a little about my setup basically i'm working on wearable ecg patch. now for simulation and measurement purpose we connect it to ecg simulator (source) and see the behavior. now the problem is that i'm seeing noise on the waveforms which are due to 50hz (what i feel from the experiments till now).so now what are the ways i can reduce it the inductive coupled noise?
so i'm going to try out following things

1. twist the cables connected from simulator to device(ecg patch), use shorter cable.
2. use some high permeability material to shield the wires and connect to ground from only one side.

can you suggest more.the wires are approx 1m in length between the simulator and the device.

also what are the commonly used shielding material which can be economical and readily available i may not get mumetal easily.

Hi,

Your picture and the latest description makes things clearer.

One is the simulator, one is the patch.
What I wanted to know:
* is this a single cable with five wires?
* is it shielded?
* what's the source and load impedance of these wires? Only brief description.

I assume you have a low impedance GND (maybe shield) and a signal wire wih high impedance load..
Then a coupling in these wires will have different influence.
Then if you want to measure a signal with GND as reference you may see a lot of coulped noise errors.

Therefore my recommendation:
* use a shielded cable
* even if the shield is GND use a separate wire as GND reference
* make this GND reference signal the same impedance as the signal wire. At source side as well as on load side.
* now use a difference amplifier to measure the difference between signal and reference wire.

Because both wires have the same impedance...the influence to bothe wires will be about the same...and the difference amplifiers cancels those errors out.

Klaus

Yes, differential signals with twisted wires are useful to reduce external coupling.