Hey, hey, okay, we got it. All right, so I am Giggs. I am the founder of Pound Electronics on Freenode.
It's Pound Pound Electronics, but it'll afford you.
So we're going to talk about what is ground, and we're also going to talk about some transmission line theory stuff toward the end there.
So is this ground? I would say not.
So I found this picture on the internet. Clearly there's some confusion about what grounding is, what grounding does.
And this is obviously not accomplishing anything.
So what? Not legit, not legit at all.
So ground is a concept by which we measure our voltage.
Voltage only exists as defined between two points.
There's no such thing as voltage that exists by itself.
So you have zero volts, and that's just whatever you say zero volts is.
And anything else is not zero volts if it's got a different voltage.
Voltage is also known as potential, and that's a good word,
because it's the potential of the voltage.
It's the potential between two different objects.
Just like, you know, you have like gravitational potential.
Gravity only really makes sense between two objects.
Voltage is the same way.
So here's some symbols and names that you might see for ground.
The first one's pretty common.
You might see the second one, the third one on some electronics things.
These words are not very good words.
I mean, obviously chassis ground is the chassis.
But these are not very good words to use for ground.
Sometimes you also see this little.
It's upside down T for ground.
So better terminology for ground is to just split it into safety grounds,
which do not normally carry current,
and working grounds, which do carry current.
And really, those are the two meaningful distinctions in ground.
So here's how you blow $1,000 in five seconds.
Hooking up your oscilloscope ground clip to something that has a potential
with relation to ground.
And if you hook it into the wall, that will blow up your oscilloscope ground lead,
or possibly worse.
So you see in this here, we got the ground clip hooked up to a 100-ohm resistor.
We're reading the signal across that resistor.
But in effect here, you're clipping your ground,
because the signal ground in the oscilloscope is attached to the chassis ground,
which is attached to the wall plug ground.
So you're grounding out through that 100-ohm resistor, 200 or 2.3 amps or so is going to go
through.
That fuse, in this case, the fuse might save you, but this is not a good idea.
So the solution to this, or this is a little diagram here, it's crudely drawn.
But you can see that the scope signal ground goes right to the breaker box,
and the neutral and ground in your wall plug come together at the breaker box.
So it's the same as connecting live and neutral, same as making a short circuit.
So one solution to this is to just use a battery powered oscilloscope.
That way, it's not referencing.
It's referencing your wall ground.
You're isolated, and you can get a little portable oscilloscope that's on a battery.
Then it doesn't matter where you clip to, because you're just measuring the voltage
between the probe and the probe ground, and your scope itself isn't referenced.
Another way to do it is to just break the ground pin off your scope.
That's not a good idea, because then if you clip your ground to something hot,
the chassis of the scope is going to be hot.
The battery powered ones are generally insulated on their chassis,
so you're not going to get a zap from touching them.
But yeah, so battery, breaking the ground pin.
Third option is isolation transformer.
This only works if you are using a low powered thing under test here,
because an isolation transformer bigger than about 200 watts is huge.
So generally, for 200 watts or less, you can use an isolation transformer for the thing you're testing,
and that isolates it as well.
It breaks that path between the live and the ground.
Oh, whoop.
Oh, my gosh.
Oh, my.
All right, or you can use a fed isolated probe.
These are kind of expensive, but they have an amplifier in them,
and they're measuring just that signal between the probe,
none of the rest of the circuits involved, and you can get those for your scope.
So what does this do?
I mean, going back to that first picture, what does this ground rod in the ground actually do?
Most of you all have probably seen this.
So I've got a demo to help explain that.
If the demo works, this is the iffy part here, and I also can't see my mouse down here.
All right, so this is called Faustad, I'm sorry, this is called Faustad Circuit Simulator.
It's a pretty neat little simulator.
You can download it in Java.
Oh, crap.
Josh, what's wrong with your mouse?
Anyway.
All right, so let me just go over the basics here.
So the dots show the flow of the current, and the red is negative voltage compared to the arbitrary ground that we've picked,
and green is positive voltage.
So this is a little LC tank oscillator.
I'm going to try, I'm having to use the web version of this because the Java version did not run.
I hope you don't have any porn-like anything.
You can't see it anywhere obvious, right?
All right.
Not obvious.
Yeah.
Y'all want to do find?
Find slash?
All right, so this is a little simulation of your actual home wiring.
Oh, it loaded.
Okay.
So we're just going to have to do it like this.
All right, so this is sort of what's going on in your house.
You've got this transformer up here in the upper left.
Is this?
the transformer on your pole or buried in your yard in some places. Then on your side
of the transformer, you've got 240 volts between the two lives. And then the neutral is in
the middle. So in the U.S., that gives you 240 between the sides, 120 between the middle
and either one of the sides. And that little interleave pattern, if you look over there
on the, let's see if I can point. Ooh. I don't know why I'm clicking. Why am I clicking?
All right. This little interleave here is the breaker box. So they actually do interleave
like that. And the reason for that is that they can make a breaker twice as wide, and
that makes it 240. Or they can make a breaker skinny, and that makes it 120, because it's
either connecting to one side or to both sides. And then the neutral ground bar is
the center of the transformer. But if you notice there, that ground, why is it moving?
All right. It shouldn't be moving. So the ground rod isn't really doing much of anything.
And in the Java version, it doesn't even show any current moving on it. But either way,
the ground rod is not doing much of anything. Now, remember we talked about safety grounds
and working grounds. The, we do have some leakage. I'm not sure. In this Java version,
it doesn't show any movement on the safety grounds. But these boxes that go around everything
here, those would be the safety grounds, like the chassis. The working grounds are these
ones that are carrying a lot of current here.
So if we have a ground fault, how much does the thing completely negate my point? So now
we have a ground fault here.
Now, the safety ground is carrying current, and that's the only time that does. But if
you notice the ground rod is not doing anything. It's not doing anything under fault conditions.
It's not doing anything under normal conditions.
If you disconnected it, not much would happen.
It doesn't do a whole lot because that fault is completely contained within your house's
transformer.
That neutral ground bar where your grounds come together to the neutral is where you're
short-circuiting.
If you have a washing machine with a metal chassis, a hot wire comes unattached, it touches
the chassis, it's a short-circuit through the neutral ground bar through the transformer
in your house.
Nothing to do with the ground rod.
What does the ground rod actually do?
Why is it there?
We have a little lightning simulator here.
This is 10,000 volts and it's just out there in relation to ground and we'll add that in
maybe.
There it goes.
Now we see the ground rod doing something.
We've got an external static charge.
That is in relation to ground.
It's not self-contained within your house's transformer.
The lightning or any static buildup in your house does filter out through the ground rod.
It gives it a potential reference to keep the ground near your house at the same reference
voltage as the zero volts that we defined as the center of the transformer.
But it really does not do much most of the time other than exist to make sure that your
house doesn't float in relation to the ground and to shunt off any kind of lightning or
static charges.
Does that make sense to everybody?
Good.
The ground rod is basically the least important part of your grounding system.
This neutral ground bar is what handles fault currents.
It's what's handling all of that sort of thing.
Now we're going to shift gears a little and talk about transmission lines.
Oh, crap.
What did I do?
Josh?
I closed what?
You minimized it.
Did I?
Oh, I minimized it.
Yeah.
Where is it?
You closed it.
You totally closed it.
Well, then what is this?
That's that.
Well, that's what I want.
Oh.
For now.
Maybe.
Is it going to work?
Yeah.
Yeah.
Yeah.
Yeah.
Is it working?
All right.
All right.
So now let's talk a little bit about transmission lines.
So there's two different ways that transmission lines generally work.
They're single-ended, in which case one side of the transmission line is defined as zero
volts or ground.
And then there's balanced or differential signaling in which there isn't really a ground.
The pulses are symmetrical in relation to each other.
And the reason that that's done is because if there is common mode noise in the bottom
picture, you see the noise, it gets subtracted from itself.
The input pulse is one positive pulse, one negative pulse.
You subtract a negative from a positive, you get a stronger positive because you flip
the signs.
But when you have noise on both sides, you subtract them, they cancel out.
So differential signaling can cancel out common mode noise.
Am I losing anyone?
All right.
So the way that you help make sure that the noise is common mode is that the wires are
generally twisted together.
They aren't necessarily shielded, which is why you see unshielded twisted pair in things
like Ethernet.
In single-ended, you're more likely to see something like coax, where the shield is
sort of acting as a Faraday cage.
Both of these are generally built as controlled impedance lines, which is something we'll
get to in a minute.
But that just means that they carefully make the wires so that it's consistent along the
whole length, that it's not varying a lot.
This is a little block diagram of something that's handling a balanced signal.
You see there is two grounds here.
Then they've got two different ground symbols, a signal ground and a chassis ground.
They do tie together right there.
But you see that first op amp, that triangle symbol, that's subtracting them.
It does something to them.
And then they've got two op amps pushing and pulling the signal like this.
To output it.
So it's a little block diagram of how a differential or balanced signal would work.
So now we're going to talk about getting high frequency.
I'm not going to read it.
You can read it.
When you start to get a wavelength that is short compared to your wire, or when your
wire starts to get long compared to your wavelength, then transmission line effects start to matter
more.
So if your frequency is low, which means your wavelength is long, then transmission
line effects don't matter as much.
And you'll see what I mean here in a minute.
So at 60 hertz, one full wavelength would be 5,000 kilometers.
So you're not likely to run into transmission line effects at 60 hertz.
At 30 kilohertz, it's 10 kilometers.
So that means that transmission line effects, all that fancy transmission line stuff, doesn't
really apply to audio frequencies.
Audio frequencies are mainly dealing with resistance.
And that's about it.
Because they're not getting a lot of transmission line effects.
Because 10 kilometers wavelength means that our wires are all real short compared to the
wavelength.
At 1 megahertz, we start to get to where we might have a wire long enough for transmission
line effects to come into play.
That's 300 meters.
And at 100 megahertz, we're down to just 3 meters where we have to start worrying about
transmission line effects.
So what is a transmission line effect?
Well, a transmission line, you have a circuit voltage is only something in relation to
another voltage.
So we have to look at both sides.
We can't just look at our signal and not look at our ground or vice versa.
We have to look at both wires.
And those wires, in relation to each other, their physical arrangement is going to determine
a capacitance.
If you all are familiar, a capacitor is just two metal plates hold a charge in relation
to each other.
How far apart they are, how close they are, what's in between them, determines the
capacitance.
It determines how much capacitance there is.
So when you're building a wire, you're also building a capacitor between, or sorry, when
you're building a cable that has two wires, you're building a capacitor as well.
It's just a long stretched out tubular capacitor.
So you have to control the distance between the wires, control the material between the
wires to control the capacitance between the wires.
And then every wire that has current flowing through it is also an electromagnet.
You all may be familiar with that.
Anything with current flowing has a magnetic force.
So wires also have inductance.
So between the inductance and the capacitance, that looks like a resistance.
It's technically an impedance.
But it looks like a resistance.
And if the wire or the cable is consistent along its whole length, it can be said to
have a characteristic impedance, which is what the cable looks like for it until infinity,
assuming it's terminated.
So we're going to maybe, maybe get this other demo on.
It's not working.
Josh, where'd the browser go?
There it is.
Does anyone have any questions while I'm playing with this?
Trying to get the browser back up.
The neutral ground.
Yeah.
The five slides ago.
Yeah.
The outside of the house ground that shunned off electrical for .
Yeah.
Thor lightning.
What is that?
What is it in your house?
All right.
So like your plug has three, three, three.
All right.
So you got live, neutral, and ground.
Yeah.
So live, neutral, and ground.
Those three wires go back to your panel.
At your main panel, if you have a subpanel, this isn't necessarily true, but at your
main panel, the neutral and ground just connect together.
Okay.
So it's just like a second path for neutral.
But it's a path that doesn't carry current.
So the neutral ground is the working ground.
The safety ground, the big longer pin.
The safety ground doesn't carry current.
So you got a working ground and the safety ground on your outlets.
And they're basically a parallel path, an alternate path that the current can take if
something shorts out to the safety ground.
And that can be useful as well because like ground fault interrupters will detect when
the neutral current doesn't match the live current, which means that something is going
out the ground where it's not supposed to be going.
It's used in the back door.
When it should be used in the front door.
Common problem.
All right.
So let me get this going again.
Sorry if I'm breathing on the mic.
I've kind of got to stand a certain way here.
All right.
I've almost got it.
And all right.
We got it back up.
So I can actually, maybe if I don't break this circuit.
This is an LC tank circuit.
So you can kind of see how an inductor takes a little while to charge up.
And then it will ring with this capacitor.
This is actually a basic oscillator that is used quite a bit in radio and everything.
So an inductor and capacitor like this.
It's not too different from what we were talking about.
But let me get.
Let's take a look at this example first off.
This is.
Let me open it and reset it.
So I said the transmission line effects only really matter when you're in a
regular pulse is getting long in relation to your wire length.
Or your wire length is getting long in relation to your pulse.
This is actually just a very.
If you look at the time step down there, five picoseconds.
So we're looking at this on a very small time step.
This is just turning a light bulb on basically.
And you notice the power just doesn't start flowing.
It bounces back and forth.
It homes in on the actual amperage.
And this is just a regular.
Regular old circuit.
You know it's going to vary a lot because our wires like in our wall aren't impedance controlled.
So we couldn't hit this exactly.
But you see it actually hunts down to the actual voltage.
We got 100 volts and 100 ohms.
I think it's going to land on one amp eventually.
But this is all happening in such a short time scale.
We aren't even at 700 nanoseconds yet.
So it doesn't matter because it looks like it's instantaneous.
Because we're looking at it slower than this is happening.
So that's why you can say if your frequency is high enough then it starts to matter.
Because you're looking at it on a short enough time scale that you can actually see this happening.
So I got another demo here.
So the characteristic impedance.
As defined by the capacitance and inductance.
What we want to do is we want to put a resistor on the end of that line.
That makes it look like that impedance continues out to infinity.
And when we do that there isn't much reflection.
This thing gives you a little shadow of a reflection even when it's impedance matched.
But there isn't much reflection.
If there is an impedance mismatch on the end of the line.
The signals actually start bouncing off the end of the line.
And then you get a very noisy signal.
This is why you always need to terminate a lot of these buses.
Or have some self termination.
Because the impedance needs to appear to continue on to infinity.
Makes the wire look like it's infinitely long.
And I've got one more demo.
How am I on time?
Is anyone going to have time?
20 minutes.
Oh we've got plenty of time.
Alright so.
Yes.
Does the resistance of the matter.
Are those 75 ohm resistors?
The reason that this has to match the characteristic impedance of the line.
So we've got a 75 ohm transmission line with a 75 ohm resistor.
75 ohm is actually a common line for like television coax.
And that characteristic impedance is based on the construction and the materials that the cable is made of.
Like the plastic that's actually in the coax helps define that.
And they handle all that at the factory.
But they can say that the characteristic impedance is 75 ohm.
So if you stick a 75 ohm resistor on the end of this.
Things shouldn't bounce off the end of it.
At high enough frequencies for it to matter.
So I have a 75 ohm line here.
And a 500 ohm line here.
So that's why they're terminated like that.
If we take off the termination in the middle of the line here.
This is actually like an impedance matcher.
It's got.
You know.
It's got 75 on this side and 500 on this side.
So if we take this off.
The signals still go through.
But we're getting reflections on the first side.
So not quite as messy.
But it's still not good.
And which one haven't we done?
Lightning.
All right.
So this is an interesting little demo.
You remember I said.
Even at lower frequencies.
Lightning has generally got a pretty fast rise time.
It's surging pretty quick.
So it can be viewed as a higher frequency.
But in the short scale.
You can actually get reflections off of a surge protector.
With lightning.
So you can almost get a secondary surge.
From the lightning reflecting back off your surge protectors.
So if you have anything important.
Try to put a surge protector on each one.
So that they don't reflect back and surge your other devices.
So let's do this.
Let's start the lightning going.
Maybe.
Oh crap.
All right.
There it goes.
All right.
So this is the lightning surge.
We're looking at a very fast time scale.
Or very slow.
Depending on how you look at it.
You know.
This is all happening in pico seconds.
Nano seconds.
This is happening quickly.
All right.
So let's say that the MOV.
This is the way a surge protector works.
It basically shorts it out.
So when it detects a voltage over a certain level.
It basically shorts out your live and neutral.
Which is why a lot of them burn out.
And if they fail.
They generally have a fuse that cuts them off.
To keep them from shorting out your whole house.
So but if you look.
That MOV is now reflecting.
It's reflecting the surge back onto the other device.
And you see the unprotected device went from 10 kilovolts to 19 kilovolts.
It actually amplified the surge.
So this can happen.
It happens on very short time scales.
It may not matter.
But if you have MOV.
If you have surge protectors or what not.
You should have them on every device.
So that you don't get these potential reflections hitting your other devices.
So and also it helps to have a surge protector at your service panel.
And on your devices.
Because if we were shorting this out at the service panel.
It wouldn't be bouncing around as much.
So got any questions about this?
This one.
So you actually put your devices at more risk by having one or two surge protectors.
Yeah.
In your environment.
It could.
I mean this is a very idealized thing.
In that there's a lot of resistance in your wires.
So it's going to decay down a lot faster.
But in theory yeah you can.
If you have a good surge protector over here.
And you have no surge protector over here.
You can actually send a reflection back to this first device.
And it can actually.
I think it was.
Square D has a paper on this.
A white paper on it.
About reflections from surge protectors actually causing more damage.
Obviously they want to sell you surge protectors for every device.
But you know the simulation bears out that it can reflect.
And it can cause a secondary surge.
That could actually be stronger than the first surge.
Is this only printed for per circuit?
Or for your entire breaker?
Well it's going to reflect.
I guess you know based off of where that.
I mean we're talking about lightning.
So the ground rod is relevant.
So generally I mean it's hard to say.
Because like if you have two wires running right next to each other.
They can induce currents on each other.
Especially when you're talking about lightning levels.
So it's really kind of hard to say.
But the bottom line is you just want to have protection on every important device.
You know things like motors.
Generally you know you might blow a hole in their insulation.
But they can tolerate a lot more than electronics generally.
So just have.
Ideally have protection at your panel.
Protection at your device.
At any important device.
And that's the best you can do.
So.
Any other questions?
All right.
Well I'm Giggs.
I'm over in the hardware hacking village.
We're helping people learn to solder and various other things.
So come see us over there.
And thanks for your time.
Thank you.