pcb trace, common-mode noise, and murphy

[pgib8]

I’m looking at this circuit board that has inputs from a reed switch. Everything is mounted in a panel which contains 240 VAC. The cable for the reed switch is like 6 feet in length and twisted. When I switch current through this panel, the cable I’m talking about picks up noise which I’m sure is through inductive coupling.


When I reviewed the circuit board I see two good things. One is that the input itself is isolated by an optocoupler, and the other is that the power going through the switch and optocoupler comes from an isolated supply, actually it is from the bias winding of the transformer in the switching supply, but nothing else uses it.


The bad part I see is that the supply and return basically cross over the entire circuit board. The isolated supply and voltage regulator are at one corner, the input and optocoupler for the reed switch are at the other corner.


So here is what I would like to understand better, so let’s talk about the noise.
The reed switch wires pick up differential noise in the range of +/- 5V (not too bad really, and probably fairly low because of the twisted wire). On the circuit board I assume that the differential noise creates a loop, and that the area of this loop will have a lot to do with how much of that noise is coupled into the circuit board.
Both supply and return traces are routed exactly on top of each other. The area where they are split is very small. With this in mind, am I correct to assume that the differential noise coupled into the circuit board will be minimal, despite the trace going across the whole board?

Now for the more interesting part, despite the cable being twisted, it is going to pick up common-mode noise, again through magnetic coupling (correct me if I’m wrong).
When I measured the common noise, I noticed spikes upwards of 50V!


So now I’m thinking that voltage will be applied across the board, equally on the supply and return traces. Since everything is isolated I do not imagine much of a current will flow so now I’m thinking that through capacitive coupling this is going to create noise on the whole board.


I have ideas how to fix the issue and if anybody is interested I can elaborate on those. For now I’m mostly interested in understanding how circuit boards are affected by noise and more specifically how the circuit I described will be affected. I’ve done a ton of research online and learned a lot but it seems just like with oscillator design, Murphy is hard at work when it comes to understanding noise.


If anybody is interested I can also draw up a schematic.


Thanks for any help in furthering my understanding!

 

[shrtrnd]

Are you considering 'outside' noise generators?
Where is this 'panel' mounted? In some main rack housing with other circuitry that may be inducing their own noise
into your panel? What else is your 240VAC connected to (some other motor or noise generator in another room?)
Just something for you to consider.
I'm wondering if this is one of those 'can't see the forest for the trees' moment.
I rarely get down and dirty in the nitty gritty until I eliminate all the possible outside noise factors.
If you're pretty sure your problem is in the circuitry itself, and not some outside influence, THEN, I'd entertain theory.
Whoever designed your board would assumedly try to compensate for the problem you're encountering.
(But maybe not?)
I'd look at outside issues before blaming the board. But that's just me.

 

[pgib8]

The panel is most likely mounted on a telephone pole but there are like 50 distribution lines going through it, each one with 250V. Hard to say where they go to but it might very well include motors. So inside the panel there will be PLENTY of noise. Sadly this circuit board doesn’t even have its own enclosure. I’m finding noise on virtually every wire that is connecting to the PCB, feeding noise into it. Finally I’m sure noise is also coupled into the PCB directly because it doesn’t have its own metal enclosure.
There is a 3.3V processor on the board and it is my believe it is getting beat up by voltage spikes. No doubt there are issues with the design.
In summary, it’s definitely outside issues coming into the board and that’s what I want to study and build a better understanding for.

 

[shrtrnd]

Guesswork is exactly that.
Do you have access to some kind of monitoring instruments, like the type Dranetz makes.
Companies like Dranetz make power monitoring equipment to identify line voltage problems.
Every field-use set-up is different. Most companies make their equipment to stand-up in a wide variety of
applications. They design-in circuit protection and shielding of critical circuits.
The reason it is so difficult to talk about noise, is that there are so many possible sources.
This application may not be the best learning opportunity because of all the variables involved.
But I commend you on your interest in understanding that noise CAN be an issue. A lot of techs never
even consider that as a possibility when they try to troubleshoot problems.
All I can say is that experience and bearing what you see in mind, will help you consider and then identify
noise problems.
I've seen guys used fluorescent light magnifiers to look at circuit boards under test, and not realize the noise
from the fluorescent fixture is screwing-up their readings. I've seen guys fret over a circuit breaker problem in one
room, not realizing the motor driven device on the same power-line in the next room was the culprit every time it
kicked-in at start up. The old SCR driven wall dimmers that wreak havoc with other electronic equipment in the room. The list goes on and on.
I guess I'm saying trying to figure out your present board might get you started in this learning process, but it
won't be the cure-all for everything else you run into if you stay in this field. It's something some people theorize
about, and something that other people experience in practical application.

 

[Arouse1973]

I’m looking at this circuit board that has inputs from a reed switch. Everything is mounted in a panel which contains 240 VAC. The cable for the reed switch is like 6 feet in length and twisted. When I switch current through this panel, the cable I’m talking about picks up noise which I’m sure is through inductive coupling.


When I reviewed the circuit board I see two good things. One is that the input itself is isolated by an optocoupler, and the other is that the power going through the switch and optocoupler comes from an isolated supply, actually it is from the bias winding of the transformer in the switching supply, but nothing else uses it.


The bad part I see is that the supply and return basically cross over the entire circuit board. The isolated supply and voltage regulator are at one corner, the input and optocoupler for the reed switch are at the other corner.


So here is what I would like to understand better, so let’s talk about the noise.
The reed switch wires pick up differential noise in the range of +/- 5V (not too bad really, and probably fairly low because of the twisted wire). On the circuit board I assume that the differential noise creates a loop, and that the area of this loop will have a lot to do with how much of that noise is coupled into the circuit board.
Both supply and return traces are routed exactly on top of each other. The area where they are split is very small. With this in mind, am I correct to assume that the differential noise coupled into the circuit board will be minimal, despite the trace going across the whole board?

Now for the more interesting part, despite the cable being twisted, it is going to pick up common-mode noise, again through magnetic coupling (correct me if I’m wrong).
When I measured the common noise, I noticed spikes upwards of 50V!


So now I’m thinking that voltage will be applied across the board, equally on the supply and return traces. Since everything is isolated I do not imagine much of a current will flow so now I’m thinking that through capacitive coupling this is going to create noise on the whole board.


I have ideas how to fix the issue and if anybody is interested I can elaborate on those. For now I’m mostly interested in understanding how circuit boards are affected by noise and more specifically how the circuit I described will be affected. I’ve done a ton of research online and learned a lot but it seems just like with oscillator design, Murphy is hard at work when it comes to understanding noise.


If anybody is interested I can also draw up a schematic.


Thanks for any help in furthering my understanding!

Yes please elaborate if you can and yes a schematic would be good and a few pictures, nice clear close ups of the PCB front and back.

Thanks
Adam