Maker Pro
Maker Pro

RS422/485 and ground wire.

I really dont want to cause you a headache. I really try to understand.
Your circuit shows a resistor as a differential component between A and B lines. In that case any current flowing into the resistor via the A line comes out via the B line
How is it possible ? there is no potential difference between the + of battery A and the + of battery B. Note that only the - sides of each batterys are tied to GND. and the resistor in that picture is not connected to it.

ground connection is required as the receiver only evaluates the voltage drop across R. Current returns to gnd of the transmitter via the "batteries" shown. You have a closed circuit and that is fine.
I pictured the resistor to be the "receiver" or at least some oversimplified equivalent.

Also how can we measure potential if we have no reference point ?
Are the rs422 circuits using lines A and B in turn as reference points to measure voltage between them ? Or maybe use the A line as ref point and measure pot difference to B line ? Thats the only case i can think of that will not need an extra common ground used from the "listener".
 
Last edited:

bertus

Moderator
Hello,

Have a look at page 20 (page 23 of the PDF) for choosing the cable for RS422/RS485.

Bertus
 

Attachments

  • 485appnote.pdf
    410.9 KB · Views: 2

Harald Kapp

Moderator
Moderator
My bad, I didn't see that you have one of the batteries backwards. This is how it should look:
upload_2020-4-23_11-55-56.png
Differential signalling means that when A is pos. B is neg. and vice versa.
 
Differential signalling means that when A is pos. B is neg. and vice versa.
True. So there may be some tiny ammount of current flowing due to non ideal conditions ? Is that why they say that the A and B lines do not serve as return lines ?

Hello,
Have a look at page 20 (page 23 of the PDF) for choosing the cable for RS422/RS485.
Bertus
It says:
A two wire system then requires two twisted pairs and a four wire system requires three twisted pairs.

Meaning that is 2 wire system we need one pair for A and B signals and one other pair for A gnd and B gnd ? or just use one wire of the 3rd pair for both A and B grounds ?
In 4 wire system why 3 pairs ? one A and B for TX and one A and B for rx and one gnd for all ?
 
Last edited:

Harald Kapp

Moderator
Moderator
As if i was to put my dmm with its negative lead to A line (always used as ref point) and its positive to B line and get reading ranging from -5V to + 5V ?
No:
- you get - 10 V when A = +5 V and B = -5 V
- you get + 10 V when A = -5 V and B = + 5 V
You get the reverse voltages when the minus lead from the multimeter is connected to B.

In differential signalling there is no reference point. The information is in the difference between the signals A and B.
This is completely different from single ended signalling as e.g. in RS232 or common TTL signals. Here gnd is the reference and the signal levels are high or low with respect to this reference.
For example:
- RS-232: High < -3 V, low > + 3 V with respect to gnd
- TTL: High is > 2.4 V with respect to gnd, low = < 0.8 V with respect to gnd (levels for inputs. Output levels are a bit different: 2.8 V and 0.4 V respectively to ensure some safety margin).

Meaning that is 2 wire system we need one pair for A and B signals and one other pair for A gnd and B gnd ? or just use one wire of the 3rd pair for both A and B grounds ?
Neither of these. Use one pair for A+return the second pair for B+return. That's exactly what you stated yourself in post #5.
 
Neither of these. Use one pair for A+return the second pair for B+return. That's exactly what you stated yourself in post #5.
Yes, forgive me im confused over this. So indeed some people say it has to exist while other say that there is no need for it to be there since the devices have their own ground.
Not using a common ground does make sense to me when we only measure the potential difference between 2 lines and we do not have current flowing (ideal condition) or we have tiny ammounts.
But what happens when we do provide a common ground ?
I have drawn two circuits one with out a common ground and the other with common ground.
The way i have come to understand it now is that both will work correctly. Yes ?
 

Attachments

  • rs422 example.png
    rs422 example.png
    67.8 KB · Views: 5
If my thinking is correct then in the below circuit we should be fine since in theory no ground loop would exist if say both transformers where fed from the same mains socket.
Only when the two circuits where to be powered from other power sources that may give the same potential difference but with respect to different grounds ( eg T1 is fed with ac voltage ranging from 110V to 220V which would still present pot difference of 110V while T2 fed with ac voltage ranging from 0 to 120 V which would still present pot difference of 110V) then a ground loop would be formed causing trouble.
 

Harald Kapp

Moderator
Moderator
The way i have come to understand it now is that both will work correctly. Yes ?
Mostly yes.
As long as the maximum allowed common mode voltage at the receiver's input is not exceeded (the datasheet of the receiver will tell you the max.). A common mode voltage may occur, as explained before, due to stray impedances, external influences etc.
A common return will greatly reduce common mode disturbances
The ground connection will also improve noise immunity and reduce emission of electromagnetic radiation.
I will always include the common return in my circuits.
 
nly when the two circuits where to be powered from other power sources that may give the same potential difference but with respect to different grounds ( eg T1 is fed with ac voltage ranging from 110V to 220V which would still present pot difference of 110V while T2 fed with ac voltage ranging from 0 to 120 V which would still present pot difference of 110V) then a ground loop would be formed causing trouble.
Is that correct ?
 
Did you see this in the given document?
i did.

Bottom line is that the system may work with only one twisted pair of wires but it is strongly recommended (given that all grounds are okay) that a common ground is used. And from what i understand if this common ground is connected to the shielding of the wire, even better.
 
I keep on reading on the net about this issue and it seems like i keep stumbling on different opinions.
Others say that a return line has to be used or ground can act as return line.
Interesting is what the following youtube video says at 03.00 about ground loops that can corrupt data. How can we fight that ?
 
I keep on reading on the net about this issue and it seems like i keep stumbling on different opinions.
Others say that a return line HAS to be used or ground can act as return line.

The operative word in question is whether it HAS to be used,
Systems that are operating with just two (A&B) conductors are working fine.
I will add that where possible I either take the power common to each side where possible and/or use a grounded shield.
But I have used the two conductor devices with no problems. Especially where the slave device in question has no provision for any other conductor but the signal pair, and the USB/RS485 interface devices quoted earlier do not provide for it either..
M.
 
Not the ones I have picked up on ebay, they have various types and configurations, as I said, the VFD's I use, there is only provision for the A&B lines on the unit.
M.
 
The date is sent between two conductors, that is all you need, in many cases the shield is connected to earth GND and/or to the common of the transmitting device,
They may, or may not use the same supply, generally each has its own.
M.
Data is sent between the two conductors, but if there is any significant difference between 0V at each end, you will have problems. Low at one end must be below the low threshold at the other, it is not isolated differential as Ethernet is. I have run RS422 over 50m, using the ground wire to hold 0V of each end close together.
 
I have run RS422 over 50m, using the ground wire to hold 0V of each end close together.
The application I mentioned was in DIY CNC machine(s) so the run is usually less than 50m.
As I mentioned in a previous post, I do use the ground where possible in longer runs.
The point being made here however seemed to be the confusion as to whether the ground was mandatory in order for the system to work at all.
M.
 
Top