Author Topic: Talk me through this simple circuit?  (Read 6620 times)

0 Members and 1 Guest are viewing this topic.

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7855
  • Country: au
Re: Talk me through this simple circuit?
« Reply #25 on: June 01, 2019, 03:29:37 pm »
But wait!---- There's more!

Back in the day, there were two schools of thought on grounding.

One, which I might call "The British way", drew schematics with a lne at the bottom normally called the "earth rail" (Another term for what others called "ground") & one at the top called the "supply rail".

The top one was usually the positive side of the supply, & the bottom one was usually the negative side of the supply.(not always, though, as with early PNP transistor circuits, the connections were reversed)

The supply side of each circuit component ultimately was connected to the top "rail" & the ground side to the bottom "rail".

This showed clearly where all the "ground" points were connected together, but was a bit unwieldy.

The other school of thought was what I might call "The American way",
With this method, the "ground" was indicated by a special symbol, which meant that all points which had this symbol were connected

Just to show that they weren't complete "party poopers" the Americans often drew simpler schematics in the British style.
European schematics were usually "somewhere in between".

Ultimately, the "American way" won out, as, with the advent of  multiple supply rails, the other way became very messy to implement.

Unfortunately for "noobs", it doesn't "lay it on the line" as to how ground connections work, like the old "British way" did.
 

Offline Nerull

  • Frequent Contributor
  • **
  • Posts: 694
Re: Talk me through this simple circuit?
« Reply #26 on: June 01, 2019, 06:05:19 pm »
Let's take it back to the original circuit in the first post of this thread.

Let's put aside rstofer's comments about the efficacy of this circuit, i presume it at least works to some degree.

Let's place this circuit on it's own in intergalactic deep space and turn it on.

Now, this circuit contains 6 connections to ground.

My question is simply this: if we cut all those connections to ground, will the circuit still work in exactly the same way?

If not, why not? Could someone give me one or two example of why the components in the circuit won't do what they're expected to do without these ground connections?

"Ground" in a circuit without any actual earth-ground connections is just a shorthand. It represents an invisible wire connecting all the 'ground' points together, just as if you have multiple power nodes named 'Vcc' you would assume they are connected.

This is the exact same circuit, drawn two different ways:


We have chosen, by convention, to call the common return path in DC circuits "Ground", regardless of whether it has any real connection to earth ground. It really doesn't mean anything more. It seems like you are vastly overthinking this.
« Last Edit: June 01, 2019, 06:07:19 pm by Nerull »
 

Offline DimitriP

  • Super Contributor
  • ***
  • Posts: 1420
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: Talk me through this simple circuit?
« Reply #27 on: June 01, 2019, 09:41:57 pm »
Quote
This is the exact same circuit, drawn two different ways:

But ...you can only simulate one of the two. The one on the left.

   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #28 on: June 01, 2019, 09:55:34 pm »
I don't agree i'm overthinking it, i'm simply trying to get a clear picture of what is actually meant by the ground symbol in a circuit schematic!

You mentioned the "return path", this is the first time anyone has called it that in this thread and that makes things a bit clearer.

I think part of my confusion is that there are two things under the umbrella term of "ground", and they seem (to me) totally unrelated.

Firstly you have the safety issues and the connection (for whatever reason) with the real ground, chassis or whatever.....

Secondly you have this return path, whose purpose is sometimes to close the circuit (although not in my circuit example in this thread) and give you a short cut route for any component in the circuit to a common point of lowest potential, typically just before the negative terminal, or if you're using a rail splitter, to that point in between the two resistors.
So in this usage, the word "ground" could be substituted with "return path".
And if we were only talking about circuits in intergalactic space, the term "ground" would be a meaningless and confusing obfuscation.
Once these circuits are returned in a spaceship to planet earth, we could happily continue using the "return path" nomenclature, adding, when necessary, "this point needs to be connected to the chassis or real ground or whatever".

So could someone look at the circuit in the first post and choose one of those connections to ground (apart from the rail splitter connection) and explain why, at that point, the designer said to himself "right, here we need to connect to the return path"? (he probably thought of it as "ground"  :P)




« Last Edit: June 01, 2019, 10:14:38 pm by Mr D »
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Talk me through this simple circuit?
« Reply #29 on: June 01, 2019, 10:53:32 pm »
In the amplifier circuit, the ground is connected to the chassis.

Also note, "return path" implies negative polarity, which is not always the common or ground node. For example, on old cars, the chassis and engine block were connected to battery positive. In that case the chassis acts as source rather than return.

Also, note that "return path" doesn't make sense in AC circuits, where current reverses direction in each cycle. In differential circuits, you have positive and negative signals which are referenced to neutral which is grounded.
« Last Edit: June 01, 2019, 10:55:16 pm by timelessbeing »
 

Offline DimitriP

  • Super Contributor
  • ***
  • Posts: 1420
  • Country: us
  • "Best practices" are best not practiced.© Dimitri
Re: Talk me through this simple circuit?
« Reply #30 on: June 01, 2019, 11:42:23 pm »
Quote
And if we were only talking about circuits in intergalactic space, the term "ground" would be a meaningless and confusing obfuscation.
Not for earth people. (pardon the pun :) )

All those terms, no matter what they are called mean what they mean in the context they are being used.

A flashlight uses a ground connection. It won't make sense until it makes sense.

Analyzing your post makes obvious two issues.
The first issue is your use of the circuit simulator. Ditch it.
The second issue is not understanding the concept of single vs dual rail power supplies.

Quote
So could someone look at the circuit in the first post and choose one of those connections to ground (apart from the rail splitter connection) and explain why, at that point, the designer said to himself "right, here we need to connect to the return path"? (he probably thought of it as "ground"
As for that first circuit.....I'm not sure what you mean by "apart from the rail splitter connection".
There are three connections available  +4.5V , ground, -4.5V
   If three 100  Ohm resistors are connected in parallel, and in series with a 200 Ohm resistor, how many resistors do you have? 
 

Offline LateLesley

  • Frequent Contributor
  • **
  • Posts: 322
  • Country: scotland
Re: Talk me through this simple circuit?
« Reply #31 on: June 02, 2019, 05:56:58 am »
Well, another name that gets used for ground is "common". It is the common point that all the other rails are "referenced" to. So all those ground symbols in the circuit are wired, or commoned, together. If you start disconnecting them all, the circuit won't work, as there is then no path for the electrons to flow. Remember circuits need to complete the loop to work. In most cases, the ground, or common is connected to the chassis (well at least in metal boxed equipment) and that got used as the reference point to measure the other voltages in the circuit. Some of them may have been positive in respect to the chassis ground, others may be negative.

Anyway, the upshot is, yes those grounds all need connecting together, and if you just start disconnecting components of the circuit from it, then it won't work as designed, as you'll have taken that component out of circuit.
 

Offline virtualparticles

  • Regular Contributor
  • *
  • Posts: 149
  • Country: us
Re: Talk me through this simple circuit?
« Reply #32 on: June 04, 2019, 05:50:47 pm »
Yet another way of thinking about a "ground" is to simply consider it to be a "net". It is common on a schematic to label a node with a "net" name, like "Vout-1". On the schematic it might be shown as a line that ends in a flag shape with the net name inside it. It might also be just be a small circle on the end of a line with the net name written next to it. Somewhere else on the page another circuit might use that output as an input and the "net" name will be repeated and a wire will be shown going from the net to the second circuit. This is done to reduce clutter on the schematic. Functional blocks are fully represented with labeled ins and outs, nets. "Ground" can be considered to be just another net.

Some circuits on here show a direct connection to the battery. One could easily have the battery symbol sitting off by itself with a ground symbol on the negative and a "+V" net name attached to the positive side perhaps with an arrow symbol. Then whenever the positive voltage is needed in the schematic one simply draws a little arrow and labels it "V+" and you have an implied connection. The same is true of the ground but the net name isn't required because the symbol itself is unique.

 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #33 on: June 04, 2019, 08:33:06 pm »
Hi folks,

I feel like i'm still banging my head against a wall, i just don't think anyone is understanding what i mean!

Let me try, at the risk of boring you all to tears, one more time, after which i'll shut up about it and move on.

What i'm not getting is whether there is anything SPECIAL about ground in a battery powered circuit where there is no physical earth to contend with, and no observer to worry about a reference.

Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

So if you agree with that (i'm sure you won't), surely a circuit could have multiple "pieces of wire"?

In the attached pix i have drawn the same (useless, do-nothing) circuit in two different ways.
In the second pic i have used two (yes, count them, 2!) grounds, on the basis that "ground" can be substituted with "piece of wire".

So to be clear, in the second pic i'm assuming that ground 1 & 2 are not attached to each other.
(EDIT: i'm also not expecting the two versions of the circuit to perform the same way in EveryCircuit, i was just using it to create the images)

So is it correct that if we were only dealing with this limited case, you could create as many seperate (unconnected to each other) "grounds" as you might like?
Before you answer please remember i'm talking about a circuit in intergalactic space where no-one will ever observe or measure anything!
« Last Edit: June 04, 2019, 09:25:40 pm by Mr D »
 

Offline hamster_nz

  • Super Contributor
  • ***
  • Posts: 2812
  • Country: nz
Re: Talk me through this simple circuit?
« Reply #34 on: June 04, 2019, 10:00:07 pm »
Completely unscientific here... by convention, for battery powered devices, even when a design doesn't have an external Ground/Earth connection, GND is now usually the negative terminal of the main battery.

One of the first things I do when looking at a schematic is "how is the power supply connected to the GND symbol".

It wasn't always the way, if you look at old schematics from the days before EDA & CAD tools if they do have a GND it is the physical external connection, and is conceptually different from the negative power rail, even if it is connected directly to it. Most schematics for battery-powered pocket radios usually doesn't have a GND at all.

It is now not uncommon to have multiple GNDs in a design - for example, an analog GND and a digital GND, but you pretty much be sure that they are solidly connected together at some point, unlike your schematic where the two GNDs will be at different potentials.
Gaze not into the abyss, lest you become recognized as an abyss domain expert, and they expect you keep gazing into the damn thing.
 

Offline LateLesley

  • Frequent Contributor
  • **
  • Posts: 322
  • Country: scotland
Re: Talk me through this simple circuit?
« Reply #35 on: June 04, 2019, 11:16:27 pm »
If any electrical or electronic engineer looked at the second circuit, they would assume that all the ground symbols were connected together. Multiple grounds get complicated, though some circuits do have them. The thing they have in common, USUALLY (but not always), is they are the 0V point in the circuit. It is usually the point where you would measure other voltages from, so is assumed to be 0V.

So your second circuit would differ from the first, as the first circuit would have voltage on that line with 3 resistors, but in the second circuit, it would be read as being connected to the other grounding points in "Ground 1", or at least at the same potential IE 0V.

But there is nothing "special" about it, it IS just a "piece of wire" as in your analogy. The only thing special, is you don't want to be touching it if you touch something else with say 100V on it. otherwise you may get a tingle. :)
« Last Edit: June 04, 2019, 11:21:06 pm by LateLesley »
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Talk me through this simple circuit?
« Reply #36 on: June 05, 2019, 03:46:09 am »
Some pieces of wire CAN be at ground. But a piece of wire carrying an audio signal, for example, cannot be ground.

The second circuit is wrong. You cannot have two grounds at different voltages in the same circuit. There is one ground (0V) and the other is something else.

Usually, convention is that the negative would be ground.
 

Offline virtualparticles

  • Regular Contributor
  • *
  • Posts: 149
  • Country: us
Re: Talk me through this simple circuit?
« Reply #37 on: June 05, 2019, 10:03:53 am »
Once again, the ground symbol is a net. The grounds you drew in the second circuit would ALL be connected together unless you used a different symbol for the "ground" on the right side. The symbols are important. They have meaning.
 

Offline pianovt

  • Newbie
  • Posts: 2
  • Country: us
Re: Talk me through this simple circuit?
« Reply #38 on: June 05, 2019, 05:31:22 pm »
Quote
I feel like i'm still banging my head against a wall, i just don't think anyone is understanding what i mean!

We understand you better than you can imagine. Below is a detailed response to each step of your thinking:

Quote
Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

Do that, it's correct.

Quote
So if you agree with that (i'm sure you won't), surely a circuit could have multiple "pieces of wire"?

This, too, is correct.

Quote
In the second pic i have used two (yes, count them, 2!) grounds, on the basis that "ground" can be substituted with "piece of wire".

This is where you joined the dark side. Work on this last step of your reasoning. You have made the following logical error:

A. Ground is a piece of wire.
B. I can arbitrarily define any piece of wire in a circuit to be called "ground".

Therefore: I can call all the wires in the circuit "ground".

The fallacy is that you can call "ground" only one node in a circuit.  The word "node" is very important. A "node" consists of all the places in the circuit which are connected to each other only with wires and nothing else. The ground symbol represents a wire. You are free to pick which node you want to call ground, but you get to pick only one.

Quote
So to be clear, in the second pic i'm assuming that ground 1 & 2 are not attached to each other.

You have just violated the premise in the beginning of your post. In that premise, you said the following:

Quote
Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

You said that the ground symbol represents a piece of wire. Then you connected the two nodes in the second circuit together using the "ground" symbol, yet you say that the two grounds are not connected to each other.

If you decide that you want to call some other node in your circuit "ground", then you will have to rename your original ground and give it a different name. Specifically, in your second circuit, you have placed a short circuit across two of the (many) 1k resistors. You "shorted" two nodes.

Make a habit of using symbols like R1, R2, ... Rn instead of just values. In your second circuit, I have no way of referring to a specific resistor unless I draw a circle around it and post that picture.



 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #39 on: June 05, 2019, 07:20:38 pm »
OK!

You almost understood what i meant, but not quite, sooooooooo........... :

Here are two more pictures.

Would you agree that, in intergalactic space, with no observers within a lightyear distance, the two circuits are equivalent and the choice to place the grounds where they are is completely arbitrary?

........and in fact no ground need even be indicated in the diagram (as long as the ground symbols are replaced with a wire)?

And just to be crystal clear: someone might ask: "if there no observers, why would you even need the diagram?". So let's say for the sake of this thought experiment that a robot needs to mass produce the circuits and will need the diagram for that purpose.



« Last Edit: June 05, 2019, 07:55:48 pm by Mr D »
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Talk me through this simple circuit?
« Reply #40 on: June 05, 2019, 07:45:06 pm »
the two circuits are equivalent
Yes

the choice to place the grounds where they are is completely arbitrary?
In this particular circuit, functionally it is arbitrary where you put ground. In AC and RF circuits, not so much. You can read about the concept of the "ground plane". As was explained, we do have conventions about where we put ground. Not to mention that thoughtful choice of ground location can make calculations easier.

........and in fact no ground need even be indicated in the diagram (as long as the ground symbols are replaced with a wire)?
In this instance yes. In other situations connecting high voltage to chassis ground can be lethal.
« Last Edit: June 05, 2019, 07:49:13 pm by timelessbeing »
 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #41 on: June 05, 2019, 08:02:08 pm »
Thanks, we're now getting close to it being straight in my head.

But still just one more thing:

the choice to place the grounds where they are is completely arbitrary?
In this particular circuit, functionally it is arbitrary where you put ground. In AC and RF circuits, not so much. You can read about the concept of the "ground plane". As was explained, we do have conventions about where we put ground. Not to mention that thoughtful choice of ground location can make calculations easier.

Are you saying that in intergalactic space where a robot had to assemble a circuit from a diagram, some circuits (AA, RF?) wouldn't work in the same way if the ground symbols were replaced with a wires leading to a common wire, and some other bunch of nodes that lead to a common wire in the circuit was designated as ground? So, in the way i indicated in my two last pix?

This is the nub of what i'm trying to get at: when there's no observer to reason about it, that ground as indicated on a diagram, is completely arbitrary in any closed, not actually physically grounded circuit!

« Last Edit: June 05, 2019, 08:06:44 pm by Mr D »
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Talk me through this simple circuit?
« Reply #42 on: June 05, 2019, 08:18:13 pm »
Yes, that's what I said and I told you where you can read about it.

https://en.wikipedia.org/wiki/Ground_plane
 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #43 on: June 05, 2019, 08:24:36 pm »
OK, that's a special case where the ground is more than a wire, it has some specific physical property required for the circuit to work.

So anyway, apart from that, i guess the answer to my last question was "no"!.

Thanks all for your patience, i'm going to move on now, but i'll be back soon with some infuriating questions about op amps!  |O
 

Online RandallMcRee

  • Frequent Contributor
  • **
  • Posts: 544
  • Country: us
Re: Talk me through this simple circuit?
« Reply #44 on: June 05, 2019, 08:27:06 pm »
Thanks, we're now getting close to it being straight in my head.

But still just one more thing:

the choice to place the grounds where they are is completely arbitrary?
In this particular circuit, functionally it is arbitrary where you put ground. In AC and RF circuits, not so much. You can read about the concept of the "ground plane". As was explained, we do have conventions about where we put ground. Not to mention that thoughtful choice of ground location can make calculations easier.

Are you saying that in intergalactic space where a robot had to assemble a circuit from a diagram, some circuits (AA, RF?) wouldn't work in the same way if the ground symbols were replaced with a wires leading to a common wire, and some other bunch of nodes that lead to a common wire in the circuit was designated as ground? So, in the way i indicated in my two last pix?

This is the nub of what i'm trying to get at: that ground as indicated on a diagram, is completely arbitrary in any closed, not actually physically grounded circuit!

Ok, Ground point is not completely arbitrary in any real circuit. It is intended to be the point at zero volt potential. This is important when you are actually going to connect one circuit to another (think: one piece of equipment to another). Even in intergalactic space this must be done non-arbitrarily. Also, in any analog circuit using a dual power supply (e.g. +-15volts) the ground is almost always chosen as the zero volt point at equal potential to the positive and negative supply because this maximizes various specifications, e.g. for an operational amplifier CMRR, voltage swing, etc. etc.

It will also become important in intergalactic space when one arbitrary circuit is grasped with one hand and the other grasps another arbitrary circuit at a potential difference of more than 40 volts. Perhaps.

 

Offline Mr DTopic starter

  • Regular Contributor
  • *
  • Posts: 63
  • Country: nl
Re: Talk me through this simple circuit?
« Reply #45 on: June 05, 2019, 09:08:33 pm »
You've set the cat amongst the pigeons there, mate!

But i don't think it really has a bearing on the basic fundamental that i'm trying to grasp, as any of the things you described could be considered as actually just being part of the same circuit!

Sure, once you insert a reasoning agent into the thought experiment that needs to decide: "how do i connect this circuit to that circuit to get XYZ result", i guess it becomes an issue?!
« Last Edit: June 05, 2019, 09:13:00 pm by Mr D »
 

Offline larrybl

  • Regular Contributor
  • *
  • Posts: 150
  • Country: us
Re: Talk me through this simple circuit?
« Reply #46 on: June 05, 2019, 09:40:32 pm »
What i'm not getting is whether there is anything SPECIAL about ground in a battery powered circuit where there is no physical earth to contend with, and no observer to worry about a reference.

Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

Ground can also be a metal chassis as depicted in this lawn mower diagram.
 

Offline larrybl

  • Regular Contributor
  • *
  • Posts: 150
  • Country: us
Re: Talk me through this simple circuit?
« Reply #47 on: June 05, 2019, 09:56:06 pm »
I meant to add in the above diagram that there is "not" a wire connecting all the connections on the right side. Most all ground connections are connected to the chassis.
 

Offline timelessbeing

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: 00
Re: Talk me through this simple circuit?
« Reply #48 on: June 06, 2019, 02:05:26 am »
battery powered circuit ... Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

That doesn't sound right to me.  I think that is has nothing to do with the power source. Take for instance a portable radio. The antenna is a piece of wire, but it cannot be "ground". If you connect it to the chassis then surely the radio won't work right. The same goes for various other parts of the input circuit.
 

Offline hamster_nz

  • Super Contributor
  • ***
  • Posts: 2812
  • Country: nz
Re: Talk me through this simple circuit?
« Reply #49 on: June 06, 2019, 02:50:38 am »
battery powered circuit ... Under these limited conditions, i can see no reason not to substitute the word "ground" with "piece of wire".

That doesn't sound right to me.  I think that is has nothing to do with the power source. Take for instance a portable radio. The antenna is a piece of wire, but it cannot be "ground". If you connect it to the chassis then surely the radio won't work right. The same goes for various other parts of the input circuit.

Are you sure? my cell phone (and maybe even Walkmans) use the GND of the earphones as the FM antenna...

If you want to see how it is most likely done... https://electronics.stackexchange.com/questions/273469/separating-an-fm-antenna-from-a-topology-which-uses-headphone-ground-as-antenna

(spoiler: there is an inductor between the headphone ground and the design's local ground).
« Last Edit: June 06, 2019, 02:59:11 am by hamster_nz »
Gaze not into the abyss, lest you become recognized as an abyss domain expert, and they expect you keep gazing into the damn thing.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf