Common ground thread - continuation

[Southpark]

It may be beneficial to add a few comments to the 'common ground' thread from : link here

There is a diagram (pasted below) that has labels of '0V' on both the device1 and device2 side. And also labels 'GND' on both device1 and device2 sides.

Those labels should not be the same. Instead, the 'GND' label at device2, should be called something different, like GND2.

It should be pointed out to newcomers to electronics that '0V' is merely chosen as a voltage that we consider as being a zero volt reference. We only choose one particular node to be the main voltage reference ----- the zero volt reference. So if the GND node on the device1 side is going to be chosen as '0V', then the device2 side should not have a '0V' label appearing there (at device2) ------ at least not in the existing diagram.

The original diagram should ONLY be labelled in the way that it is for the case where the negative power terminals of device1 and device2 are electrically connected together. And the diagram currently has the negative power terminals of device 1 and device 2 not connected together.

This isn't to take anything away from Perry's excellent thread post. It's only to make a comment.

The newcomers to electronics will understand things more if they understand voltage referencing. So for the existing circuit, if the device1 GND is set to be zero volt reference, then output1 voltage will be referenced to that device1 GND.

But the satisfactory operation of device2 will require an appropriate design voltage to appear at the input (ie. input2) ...... and that design voltage needs to be at the correct (appropriate) level RELATIVE to the 'local ground' on the device2 side ----- and that local ground is not 0V for the case where the local GND of device1 is already chosen (reserved) as 0V. So device2 is not in a position to operate properly because the 'local ground' on the device2 side is not at 0V ........ also meaning that device1 can't apply a voltage of say 5 volt at input2 (relative to GND2) ------ and this is because that 5V is relative to the device1 GND only.

[PerryBebbington]

Hi Southpark,
Thanks for the feedback, much appreciated.

Those labels should not be the same. Instead, the 'GND' label at device2, should be called something different, like GND2.

I agree, I will change them.

As for the other stuff...
The tutorial took me a long time to write, not because common grounds is a difficult concept to illustrate but because I initially wanted to include all sorts of stuff such at the meaning of 0V, ground, earth, protective earth, floating inputs, pull up resistors and probably other related things I can't think of right now. The tutorial got created when I realised that the thing I really wanted to address was the repeated questions from newbies about a signal from one circuit to another that didn't work because they didn't understand that circuits are called circuits because they provide a path for the charge carriers to, err... circulate. Once it was clear to me that was what needed to be explained all the other stuff got left out deliberately.

0V, ground, earth and voltages being relative are worth covering and are on my list of things to add to that tutorial as extras, not in the first post. I want to keep that as it is, not withstanding your comments about different ground labels. The problem that has held me back, apart from concentrating on other tutorials, is that first I don't like calling any part of the circuit 'ground' unless it is actually connected to something that could reasonably described as a ground, so at least a ground plane on the PCB or the chassis of the equipment or protective earth. 0V and ground are generally the same thing, but at the very least the different names have be explained. Then you have protective earth, which might or might not be connected to 0V, meaning you might or might not have 3 different names for parts at the same voltage. Then you can add in the incorrect assumption that the negative pole of the power supply is always 0V or ground or whatever you want to call it and by the time you explain all that you have something horribly complicated, or at least something I can't think of a way to explain simply. This is made worse by the fact that I don't see consistent use of the words 'earth' and 'ground' on this forum, particularly on different sides of the Atlantic.

For all these reasons I've avoided addressing it.

[Edit] Don't forget that protective earth is provided in different ways in different countries, or even within the same country.

Do you want to write it?

Perry

++Karma; // For raising some important points.

[PerryBebbington]

What do you think of this?

I don't want to use it until it has been commented on; I have to ask the mods to unlock the topic before I can post, so I don't want to do that until I'm sure.

[JCA34F]

Circuit common should be called that and only called "ground" if it is connected to EARTH. We need a new, agreed upon symbol for circuit common / 0V, maybe we could have a contest and...
My klutzy entry:

[Paul_B]

The traditional term for it is "chassis", and for a very practical reason.

[Southpark]

Perry --- your thread article is excellent already. And your definition '0V for device 1' and '0V for device', and also 'gnd for device1' and 'gnd for device2' looks good too.

I can attach some diagrams where I use the pics to explain how connecting GND1 and GND2 together allows a particular voltage at any selected node of both devices to be the same voltage (with respect to both local grounds).

For the case where GND1 and GND2 are disconnected from each other ------- device1 has no means to apply 5V to the input of device2 (relative to GND2). Device1 is able to generate 5V with respect to GND1. But is unable to make the voltage at input2 equal to 5V with respect to GND2. So device1 has no control of device2 when GND1 and GND2 are disconnected.

There may be holes and gaps in my diagrams and notes. But just contributing too. Once again ----- excellent thread you made Perry. An excellent sticky.

Also - the information in your thread - about circuital loop (closing the circuit - for currrent/power flow etc) is very good. Just mentioning this, as that is definitely very helpful information.

[Ballscrewbob]

Southpark:
Perry --- your thread article is excellent already. And your definition '0V for device 1' and '0V for device', and also 'gnd for device1' and 'gnd for device2' looks good too.

I can attach some diagrams where I use the pics to explain how connecting GND1 and GND2 together allows a particular voltage at any selected node of both devices to be the same voltage (with respect to both local grounds).

For the case where GND1 and GND2 are disconnected from each other ------- device1 has no means to apply 5V to the input of device2 (relative to GND2). Device1 is able to generate 5V with respect to GND1. But is unable to make the voltage at input2 equal to 5V with respect to GND2. So device1 has no control of device2 when GND1 and GND2 are disconnected.

There may be holes and gaps in my diagrams and notes. But just contributing too. Once again ----- excellent thread you made Perry. An excellent sticky.

How old are your kiddies ? Tell them well done for the pictures

[Southpark]

ballscrewbob:
How old are your kiddies ? Tell them well done for the pictures

I drew them myself BSB ...... I'm a big kiddy hahahaha. Ok ..... a kiddy at heart. And thanks for the kind comment! The colours turned out ok hehehehe. I used 'paint.net' (that's what the paint package is called .... paint dot net ..... it's free).

[Ballscrewbob]

Southpark:
I drew them myself BSB ...... I'm a big kiddy hahahaha. Ok ..... a kiddy at heart. And thanks for the kind comment! The colours turned out ok hehehehe. I used 'paint.net' (that's what the paint package is called .... paint dot net ..... it's free).

Lets stick with the kids and save your reputation LOL.

[Southpark]

ballscrewbob:
Lets stick with the kids and save your reputation LOL.

My reputation on arduino forums is very good, and will remain very good heheh. You can clearly see that I tried to be neat with the 'sketch' and also used colour coding. There is value in my painting (ok..... work of art). But I'm only mentioning it for fun. It does look like colour art. But certainly the purpose of the set of diagrams is to add extra pointers regarding connecting together the local grounds.

[Southpark]

Just answering this one too ---- link

Why can the current from device 1 not simply flow off into ground of device 2? I understand that there must be a current flow. But why does the current have to flow back to the SAME ground? For batteries, I imagine that there could be some kind of capacity/limit of how much current can be dumped into their ground. But if the power supply uses a power socket ground, the need for common ground is completely mysterious to me.

For circuit 1, why does it matter if some current is lost along the way?
For circuit 2, why does it matter if its ground has to sink a little more than the + side of the power supply provides

For circuit 1, the output current is equal to zero. That's because there is only actually a single line (the green coloured one) that connects device 1 and device 2 together. There is no current returning line that goes from device 2 back to device 1, even though I have included a hypothetical (theoretical) return line, but it has infinite resistance.

So device 2 is actually the circuit that is hanging off the output of device 1. And due to no return line, the resistance of device 2 that appears at the output of device 1 is infinite. That is, the resistance seen at the output of device 1 is infinite. And infinite resistance is going to mean zero current. No current flows through the green coloured line.

Also ------ the definition of a 'ground'. A ground is generally nothing more than a point of voltage reference in a circuit. And usually that point of voltage reference is just 0 volt (relative to itself .... ie relative to its own point ---- is zero volt). We normally choose a single 'ground' point where we simply assign the voltage at that point (relative to itself) as being ZERO volt. It is for voltage referencing purposes - used in circuit analysis.

The symbol (such as a triangle) that is often included in a circuit diagram for the 'ground' reference ........ well, it is a triangle symbol ...... and that symbol is just a symbol. It is not a source and it is not a sink. That symbol is just meant to be a landmark point on a circuit, for referencing ...... that's all it is.

And the other important thing about batteries .... or a battery is ....... if the battery is supplying electricity to a circuit, then a current flowing out of the positive terminal of a battery at any one moment in time, will be exactly the same amount of current that flows into the negative terminal of the same moment in time. This is the 'ideal' assumption about batteries in basic circuit theory, and we need to remember this theory.

[SteveMann]

JCA34F:
Circuit common should be called that and only called "ground" if it is connected to EARTH. We need a new, agreed upon symbol for circuit common / 0V, maybe we could have a contest and...

Earth ground and Chassis ground are defined in the International Electrical Code:

Earth ground, IEC60417-5017:

(The earth ground symbol is a notation similar to Nicola Tesla's patent drawings that were a triangular section of earth with a plate inside.

Chassis ground, IEC60417-5020:

The origin of the chassis ground was the terminal lug,

Digital ground:

[PerryBebbington]

Thank you everyone for your suggestions.

I will change the illustrations to more explicitly show that the separate circuits do not share a common ground so there is no return path for current between them.

As for the other stuff, all interesting but I did not set out to write a basic electronics course, I set out to address the common problem asked here a lot of why a single wire between 2 otherwise separate circuits doesn't produce the required result. Since posting the tutorial I think there has been a significant drop in such questions, and when they do arise there is now some where to point people to. I have therefore largely achieved my aim in writing the tutorial.

The other common problem that needs to be addressed is floating inputs. This is partly and indirectly covered by common grounds as a wire between the output of one Arduino to the input of another with no common ground produces a floating input on the second Arduino. That is not, of course, a complete answer to the problem of floating inputs.

Maybe one day I'll do the floating inputs, for now I am not happy with 'how to get the best from this forum (full version)', but I am taking a rest from that the moment.

++Karma; // To everyone who commented, much appreciated.

[Southpark]

Hats off (and on!) to you Perry for that sticky about the ground. Having the sticky there and mentioning the importance of connecting those local grounds together ---- ie. the title of the sticky thread itself -- is a massive help for everybody.