The origin of contact noise? I was trying to measure the noise of a device with metal probes. I was not sure whether I should trust the results because I was told contact noise might contribute to some degree.
I am a little confused about the notion of "contact noise". Is it because of the contact resistance (every resistor is a noise source)? Or is it something related to other factors such as probing materials or surface? Could anyone make a brief explanation? I am eager to know the origin of this "contact noise", and how I can evaluate such noise.
 A: The noise observed at a contact is known as chattering. A stronger mechanical contact should impart a steadier electrical contact resistance (ECR) [1].
However the structure and cleanliness of the surface should also be considered in the design of this setup to minimise the presence of varying passivating layers. It should be noted that time further plays a role in stabilizing contact resistance.
A: There's a few factors I am aware of.
The most obvious is contact pressure (and variable contact pressure) affecting contact resistance due to increased surface area via deformation or breaking through the oxide layer to varying degrees.
The second one is the regular noise you get from resistors. Thermal noise or Johnson noise.
A third source is related to how running high current through a "dry contact" that has been plated with precious metals to conduct "zero current" signals can no longer do so after you run a "wet" signal which is high currents through it. It produces an oxide layer on the contacts that requires high voltage or high current to punch through. In other words, you get a nonlinear resistance that varies with the voltage or current across the contact. The most obvious situation that I've run into this in (other than contacts) is trying to measure armature resistance of a generator through the output terminals. The tiny test current provided by a multi-meter or resistance meters gives wildly variable readings. You have to connect the output to a DC supply, turn up the voltage/current and use R=V/I to calculate the resistance. Even then it varies based on how much current you are running through it.
The fourth is piezo-electric effects of the materials being probed (and some unexpected materials experience this. I think aluminum is one). So as long as your pressure varies (or if you're scratching the probes along the material) noise will be generated.
