furthest distance between two 8031

Hi,

What is the longest distance that can be achieved just using the TX, RX and GND to communicate between two 8031s at 2400 & 9600 b/s where the interface is:

  1. direct connection (TTL).
  2. through a pair of MAX232.
  3. using 26LS31/26LS32 transceivers.

Should I use twisted pair wires or straight parallel wires to improve the transmission?

Allen

------ The next war will determine NOT who is right BUT what is left.

Reply to
Allen Bong
Loading thread data ...

and

interface

If you're going to a system with different power supplies or out of the same box, do not attempt without drivers of some kind.

Your speeds are so slow that you can get reasonable distance with RS-232 drivers/receivers (the maximum in standard is 50 feet / 15 meters).

Probably the highest speed with simple drivers is reached with RS-322/RS-485 drivers/receivers. Even here the ground differences must be moderate (not much over a volt). Here with suitable cable and ground noise in control, the max distance should be at least hundreds of meters.

For longer distances, consider using a pair of inexpensive modems.

HTH

Tauno Voipio tauno voipio @ iki fi

Reply to
Tauno Voipio

The RS485/422 is specified as 1 mile or so at whatever baudrate.

Rnee

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

the

same

RS-322/RS-485

(not

the

Yep - subject to the common-mode limitation, which often is not possible in practice, even with shorter cables.

Tauno Voipio tauno voipio @ iki fi

Reply to
Tauno Voipio

: >

: > The RS485/422 is specified as 1 mile or so at whatever baudrate. : >

: : Yep - subject to the common-mode limitation, which often is not possible in : practice, even with shorter cables. :

You can get optoisolated inputs on those, which will remove that voltage difference (on ground). Done 1/4 mile no problem that way. Well realistically 1000' (the whole spool) , not a 1/4 mile, both units getting power from different panels with different grounds.

Reply to
C Wood

True. Well, the GND should be passed along with the signal.

Rene

Reply to
Rene Tschaggelar

And what do you expect to gain by doing that ?

The common mode voltage difference is still there in most cases and now you have the often dirty ground loop current flowing all around in quite thin wires and PCB strips.

The most common reason for ground potential differences at the end of the serial link are the different electric distribution systems in the ends of the link. In most systems, the signal ground is tied to the mains protective ground, which is tied to the mains neutral in the mains socket or in the local or main distribution panel.

In single phase systems and badly balanced two phase (America) or three phase (the rest of the world) distribution systems, a considerable current can flow in the neutral wire, causing some voltage drop in the neutral conductor. This potential is propagated through the protective ground to the signal ground of the equipment.

Since the neutral current and neutral wire resistance is usually different in different places, there is going to be a potential difference between the ends of the serial cable.

Adding a thin signal ground wire does not help a lot to remove the potential difference between the equipment, although some current (a few amps at most) will flow in the signal ground wire.

To be really effective in removing the ground potential difference, the signal ground wire must have a cross section area many times the neutral wire cross section and it may have to carry up to tens of amperes of ground current in the worst case.

Thus, for long distance serial connections, especially between different buildings (in which thunderstorm related ground potential differences can be quite large), the best way is to totally isolate at least one end of the link from the local ground. In this way, there can be no ground loop currents or common mode voltages at the transceivers.

However, if the ground potential difference is known to be at all times within the transceiver common mode range (e.g. in the same house and the electric distribution branch), a simple twisted pair should be enough with termination resistors at the receivers. The transceiver grounds can be connected to the respective local protective ground. If a shielded cable is used, the shield can be connected to ground at one end only.

Paul

Reply to
Paul Keinanen

This :

isolate one end. Then it is useful to have the GND belonging to the signal. Have a floating supply for the dangling end. Some even spend another wire and pass the Vcc and the GND along. Thus no floating supply is needed for the isolation.

Yes.

Rene

Reply to
Rene Tschaggelar

In this situation the only reason I can think of for a separate ground wire in a typical RS-422/485 point to point connection is to provide bias current for the receiver input transistors.

However, with a typical fail safe termination system, with the actual termination resistor between Rx+ and Rx- and pull up resistors to the local receiver Vcc and Gnd, the receiver transistors will receive the correct bias current without a separate Gnd wire even when the remote end is floating.

Yes, this is a useful configuration.

Paul

Reply to
Paul Keinanen

the

same

I need to communicate a distance of about 75 to 100 metres in the same building.

RS-322/RS-485

the

My 8031 board came with a pair of 75176 with the signals -422XMIT , GND and

+422RCV. Should I use a twisted pair for the TX and RX signal or just use audio screened wires with the ground terminated on one end?

How do I use 2 modems without going through an telephone exchange. The 2 wires coming out of the modem would be floating.

Allen

Reply to
Allen Bong

RX

the

RS-232

control,

in

I'd like to read more on this "common-mode litmitation" subject. Any recommendation for a good link?

Allen

Reply to
Allen Bong

At speeds up to 9600 (or even 38400 bit/s) optoisolated 20 mA current loop would be the simplest solution. Use the slowest practical speed and some low pass filtering matched to the line speed may also help reduce transmission errors.

Optoisolated RS-422 is also a good solution. If different variations of the transceiver chips, one rated at say up to 250 kbit/s and the other up to 2 Mbit/s, use the slower one if your application permits it, since it will better tolerate very short interference peaks. Of course external low pass filtering tailored for the transmission speed can be used at both the transmitter end (to slow down transitions and thus reduce crosstalk) and at the receiver end (to reduce sensitivity to interference peaks).

One should remember that when using metallic conductors transferring signals between a very noisy environment and to a less noisy environment, any noise picked up by the conductors (or even the metallic cable shield) will propagate through the wall to the less noisy side and radiate into it, unless some feed through filtering is used at the wall.

If the transfer speed needs to be large, there exist huge potential differences between the ends or there might be problems due to interference conducted through metallic conductors from one area to an other, using optical fibers is the simplest solution.

Paul

Reply to
Paul Keinanen

Is it possible to use voltage reference to transfer signal beside 1010? etc transfer 24 bits condition from one end to the other using voltage reference from 0 - 5V or 0 - 12V or even 0 - 24V by mean of DAC to ADC conversion.

25m
Reply to
Net Hunter

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.