High Impedance node noise pick up

Hi Hari,

The above is HPF, see the simulation below. 

Noises can be divided into intrinsic and extrinsic noise categories. In the TI Op Amp training video below, it talked about intrinsic noises affect op amp extensively. 

https://training.ti.com/ti-precision-labs-op-amps-noise-1?context=1139747-1139745-14685-1138803-13232

If you are talking about extrinsic noises coupling onto high impedance nodes, we are still able to summarize into 3 noises category,  namely induced voltage noise, induced current noise and 1/f noises. 

In the circuit above, Kai demonstrated how capacitive coupling through high impedance node can be significant. It is a part of CdV/dt coupling, when changes in dV/dt through conductor's insulation material to a high impedance node. 

In inductive noise coupling, the induced voltage is Ldi/dt, when external environment changes in di/dt. This is likely happened near high current or ground loop test environment.   

If we consider the noise couplings in terms of noise power, when a small percentage of external power is coupled or "absorb" onto an EUT.  For a given amount of noise power, high impedance node will generate larger voltage noise, Vnoise = sqrt(Pnoise*Z) than lower impedance node. 

Best,

Raymond

Hi Hari,

the circuit shown in my simulation is no low pass filter, it's a high pass filter. See also Raymonds simulation.

Inductive noise coupling happens whenever conductors form a loop and a magnetic field (from a mains transformer, e.g.) is penetrating this loop area. Then, a voltage is induced in this loop and a current will flow in this loop which causes voltage drops across impedances in this loop.

The best cure against inductive noise coupling is to minimize the loop area formed by any signal or supply current. Because of this, it's always necessary to keep the conductor in which the ground return current flows close nearby. No signal or supply conductor must be routed without having the conductor of ground return current close nearby.

Kai 

Hi Hari,

Do you still have questions? I noticed that you have three "? " in your last paragraph. 

Kai is implying Faraday's law. V=Nd(phi)/dt. If magnetic lines of force cut the conductor (EUT), a voltage will be induced in a conductor (even N=1, if d(phi)/dt changes is large). 

Best,

Raymond

Hi Hari,

The noise coupling is fairly complicated topics. Even you understand Faraday's law and Ampere's Law in magnetic design, it is still difficult to have full understanding in how noise is coupled into a circuit. 

I checked some videos in youtube that may guide you in these learning topics. I would suggest to look up Power Input and EMI-EMC sections of RTCA DO-160 document. DO-160 is aerospace airworthiness compliance and test requirements. It will give you ideas how aerospace is testing these issues. From there, you may try to learn other relevant topics. 

Automotive has similar test requirements and compliance, but it is more difficult to find these in a single document. 

https://www.youtube.com/watch?v=PACur_GcTJ0

https://www.youtube.com/watch?v=J6gPzMYb5Z0

https://www.youtube.com/watch?v=gBCOeYfRL5g

Best,

Raymond

Hi Hari,

yes, even this parasitic high pass filter is forming a loop. But shielding and grounding breaks this loop. The induced current is now flowing in the shield and no longer through R1.

Kai

You are welcome :-)

Kai