Can Optocoupler works when connected to different ground in floating way?

Thread Starter

altisclimax

Joined May 12, 2013
9
Hi guys, this is the first time i encounter the application which the signal side only have +VE input but did not provide its gnd/-VE (the system are not common grounded). as in attached image.
Will the optocoupler works if ground connected to ground in this way?
Need yours opinion, thanks.
 

Attachments

Tealc

Joined Jun 30, 2011
140
In your design the optocoupler Pin 2 needs to go to Ground A, otherwise it's not properly isolated.

I'd imagine it would work but it's not an ideal use of the device.

Is it not possible to tap into Ground A?
 
Last edited:

Ramussons

Joined May 3, 2013
1,409
The "Ground A" and "Ground B" need not be at the same "absolute potential". The difference would depend on various factors including distance.
If both the "grounds" are in close proximity, the circuit may perform as expected. If the earth potentials are different, the circuit may malfunction. There is also a possibility of damaging the Opto coupler.

Ramesh
 

Thread Starter

altisclimax

Joined May 12, 2013
9
In your design the optocoupler Pin 2 needs to go to Ground A, otherwise it's not properly isolated.

I'd imagine it would work but it's not an ideal use of the device.

Is it not possible to tap into Ground A?
yupp, there is no way to tap the signal (customer not allow)
 

crutschow

Joined Mar 14, 2008
34,452
In that circuit the opto coupler serves no purpose and the circuit will not work if GND A and GND B are not connected in some manner. You must connect pin 2 to GND A for the circuit to work as an isolator.

If you customer won't allow any connection to GND A then you will need to tell him the circuit won't work. Options are to use a fiber-optic cable or RF to send a signal between the two points.
 

Thread Starter

altisclimax

Joined May 12, 2013
9
In that circuit the opto coupler serves no purpose and the circuit will not work if GND A and GND B are not connected in some manner. You must connect pin 2 to GND A for the circuit to work as an isolator.

If you customer won't allow any connection to GND A then you will need to tell him the circuit won't work. Options are to use a fiber-optic cable or RF to send a signal between the two points.
Actually the main concern for the application is not to isolate the two circuit. The major purpose is to detect the high/low state of the signal side. If the optocoupler is able to turn on whenever ther is a 24V, then the LED will lit on to indicate High state. What i worry is the the optocoupler will not work.

As long as the optocoupler is able to function, then will be fine for me.
 

hexreader

Joined Apr 16, 2011
581
Just noticed the word "floating" in the thread title.

If either ground is floating (which means it is not a ground at all) the circuit will not work, and no amount of wishful thinking will make it work.
 

Thread Starter

altisclimax

Joined May 12, 2013
9
Just noticed the word "floating" in the thread title.

If either ground is floating (which means it is not a ground at all) the circuit will not work, and no amount of wishful thinking will make it work.
OMG, but i will try the hardware test later, hoping a small current flow due to potential different will drive the optocoupler, or maybe i will choose a more sensitive one like 6N139 which require minimum 0.5mA to function.
 

t06afre

Joined May 11, 2009
5,934
Just noticed the word "floating" in the thread title.

If either ground is floating (which means it is not a ground at all) the circuit will not work, and no amount of wishful thinking will make it work.
That is correct. Unless A and B is connected together. No current will flow in the circuit. But connecting A and B together may not be something that is preferable. I think we need a lot more details in order to give any good answers here
 

Thread Starter

altisclimax

Joined May 12, 2013
9
That is correct. Unless A and B is connected together. No current will flow in the circuit. But connecting A and B together may not be something that is preferable. I think we need a lot more details in order to give any good answers here
The signal side (24V and Gnd A) is floating (not earth), both GND A and B are not connected. The +5V and Gnd B (Not sure whether real-earthed or not). The design is to read the state of the signal side (24V & GND A) . That is the information for both source.
 

Potato Pudding

Joined Jun 11, 2010
688
The signal connection would be by optical transmission.

Convince the customer that pin2 is allowed and must be connected to GNDA.

This will NOT repeat NOT connect GNDA and GNDB with the disclaimer that the opto isolater will have only a few kiloVolts of rating (2 to 25kV are likely) that might break down in truly horrible situations, even with only pin1 connected.
 

t06afre

Joined May 11, 2009
5,934
To me it also make no sense allowing access to a +24 volt signal and not any signal ground. What kind of equipment is this.
 

Potato Pudding

Joined Jun 11, 2010
688
It is basically a telltale indicator that shows that the 24Volt power and 5V power are both there as I understand it.

That is paraphrasing where a quote might have worked better.
 

Potato Pudding

Joined Jun 11, 2010
688
I think there is a disconnect here between relative grounds and floating grounds. GNDA is relative to the 24V signal. GNDB is relative to the 5V power.

Circuit A and circuit B are floating relative to each other.

That is actually the reason why you would use an Optoisolator.

If I have misunderstood, the intent of request, and how others are reading it then I can't say I am surprised because there seems to be a lot of confusing spinning around this.
 

Potato Pudding

Joined Jun 11, 2010
688
Almost done. There is a methodology for bringing grounds towards each other that is very similar to ESD discharge practices.

You connect the grounds with a lot of ferrite chokes, high megaohm resistances, and metal oxide varistors.

Nothing I have spent much time with and probably not relevant beyond being aware of the fact that not all floating circuits are meant to float completely free.

Just asking if this might be something that needs to be added to the discussion.
 

Thread Starter

altisclimax

Joined May 12, 2013
9
I think there is a disconnect here between relative grounds and floating grounds. GNDA is relative to the 24V signal. GNDB is relative to the 5V power.

Circuit A and circuit B are floating relative to each other.

That is actually the reason why you would use an Optoisolator.

If I have misunderstood, the intent of request, and how others are reading it then I can't say I am surprised because there seems to be a lot of confusing spinning around this.
Yupp, that is what i mean.
 

BillB3857

Joined Feb 28, 2009
2,570
If you were to assume the 24 volt source was a battery, would you expect anything at all to work if only connected to the positive terminal of the battery? I didn't think so.
 

kubeek

Joined Sep 20, 2005
5,795
It is kind of IC tester, and the 24V signal is the telling the state / condition, for example: pass/ fail/ error.
24V singal in respect to WHAT exactly? If you don´t have access to the common point it is supposedly referenced to, it could just as well be a piece of wire connected nowhere and no one would ever be able to tell the difference.
 

Thread Starter

altisclimax

Joined May 12, 2013
9
If you were to assume the 24 volt source was a battery, would you expect anything at all to work if only connected to the positive terminal of the battery? I didn't think so.
Yupp, U are right.
I already do the hardware test, it wont work at all.
The only solution is to common the ground.

Thank you guys for ur information.
 
Top