0
\$\begingroup\$

This is for a specific scenario.

Sometimes when a high current electrical device like an AC motor is turned on /off; or another crude example sometimes you have an inductive load sourced directly from AC mains controlled by a a manual bad switch and you turn the switch on/off. Sometimes one can even see some electric arc.

What type of noise is sent to the mains grid/wiring in these cases? I wonder how this kind of inductive switch effects other sensitive devices like TVs or radios or measurement systems for a short period of time.

Lets say there is a measurement system(transducer like an analog temperature sensor) and by using a coax cable it is send the measured voltage to an ADC.

And lets say that spiky inductive switching device and this transducer is powered from the same mains power plug or they are using the same line neutral.

How does the inductive spike route and effect the transducer or its coax line? Would that happen as a conductive or radiate way? How does the conductive way routes all the way to the transducer system?

The reason I'm asking is I'm wondering whether this type of noise through mains as conductive a common mode or differential noise. Would the spike/noise route through line or both line or neutral? How would it make it all the way to the coax through conduction?

Since it is impossible to track and see I cannot find an answer.

\$\endgroup\$
  • 2
    \$\begingroup\$ "effect" and "affect" are different words and "effect" makes absolutely no sense here. \$\endgroup\$ – Brian Drummond Sep 18 '17 at 10:35
  • \$\begingroup\$ The best place to deal with inductive spikes is at the source. That usually involves clamps and/or snubbers, but that would be a different question. \$\endgroup\$ – Tut Sep 18 '17 at 12:10
1
\$\begingroup\$

How does the inductive spike route and effect the transducer or its coax line? Would that happen as a conductive or radiate way? How does the conductive way routes all the way to the transducer system?

It's a pulse of HF with a wide spectrum, so it will be a mess.

It begins as differential-mode across the switch.

Consider the electrical path from the arcing switch to the "receiving" device: this isn't exactly shielded controlled impedance coax cable. Let's say it runs in differential mode along wires inside conduits.

Every time one of the wires gets close to something conductive (wall studs, your fridge, anything) then the coupling between both wires and said conductive object will be different, thus some differential to common mode conversion will occur.

By the time it reaches your device, some of it will be common mode.

Now it reaches the transformer: let's say it's an EI core, in that case one "side" of the primary will be a lot closer to the core than the other side. At this point, via capacitive coupling, more differential to common mode conversion will occur.

It it's a toroid, then the boards and other stuff inside the enclosure will have more coupling towards certain parts of the primary coil than others, so same effect.

Now say you have a CD player connected to an amp... both have different transformers wired in different ways, so the differential to common mode conversion will create different levels of common mode noise for both. This creates a common mode current across the unbalanced connections, and a loud crack is heard in the speakers every time the fridge turns off.

Fix is either at the source (add a capacitor to prevent arcing) or at the receiving end (add IEC mains filter, or plug everything in a filtered outlet).

| improve this answer | |
\$\endgroup\$
0
\$\begingroup\$

Sparks are wide-band frequency generators. This is one reason RF transmitters relying on sparks to generate the carrier frequency are specifically illegal in many jurisdictions. There is just too much off-band noise.

A spark across a switch that is in series with the power line will add some noise to that power line. In theory, well behaved devices filter out most noise they make before it reaches the power line, but you can't count on all devices sharing your power line to be well behaved.

If you have sensitive circuitry powered by the power line, it needs to be defensive. It must have enough filtering so that the inevitable noise on the power line doesn't affect its operation.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Lets say the spike made its way to AC power line. And now it is routing to another devices transformer's primary. Would that spike noise appear both on line and neural? Because if it appears on both I will call it common mode noise otherwise I will not call it common mode noise. \$\endgroup\$ – floppy380 Sep 18 '17 at 11:09
0
\$\begingroup\$

Consider the opening switch as a capacitor, the wiring as inductor. As switch opens, the energy of the wiring (inductively stored energy) causes an ARC to form across the two metal pieces of the switch. The ringing frequency is set by the capacitor value (constantly changing, as the metal pieces move further apart) and the wiring inductance.

Assume metal piece spacing is 100micron (1e-4 meters). Assume 2mm by 2mm area to the metal pieces. The capacitance is

C = 9e-12 * er=1 * 2mm * 2mm / 1e-4m = 9e-12 * 4e-6 / 1e-4

C = 9e-12 * 4e-2 = 36e-14 = 0.36pF

We'll assume 10 meters (33 feet) of wiring, thus approximately 10uH inductance.

Fring is sqrt(25,330 / (LuH * Cpf)) = sqrt(25,330 / (10uH * 0.36pF))

Fring is sqrt(25,330 / 3.6) = sqrt(9,000) or approximately 95MHz.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.