Can somone please let me if there is a maximum or minimum length of cable allowable for Modbus?

 

Some people say "4000ft" but it really depends on your physical layer and/or modulation scheme. If RS-485, there are design considerations, but the design depends on you. Remember this - the longer the line, the more it looks like a transmission line. Measure the rise-time (slew-rate) of your transmitters. This frequency component will determine the length of the line (>1/4 wavelength) at which it becomes a "transmission line." If it is a transmission line you must use proper termination for your topology. For a simple termination the source and load impedances must be equal to the cable's characteristic impedance. Bym matching the impedances you get maximum power transfer and minimize reflected power.
Also keep in mind that longer lines are lossier.

 

Modbus is a comms protocol (how to frame the messages, checksums, etc.) or actually two such protocols, Modbus RTU and Modbus ASCII. The persmissible cable and drop will be determined by the electrical specification of the network on which the Modbus is transmitted, such as RS232, RS485, etc., in conjunction with the baud rate (faster rate -> shorter cable run).
--
Steve Myres, PE
Automation Solutions
(480) 813-1145

 

Typical Modbus RTU is RS-485 standard. 4000 feet (1200 meters) is generally accepted as maximum length.

 

This will be determined more by your choice of serial interface standards more than anything else. RS-485 max about 4000', RS-232 about 150', 10baseT ethernet about 300'. MOre than this for any of these, and you need to start using repeaters or change transmission modes (modems, radio etc).

 

Yes and No. Yes because there are specified limits for RS232 and RS485 interfaces, but these are influenced by the cable type, the baud rate, and the environment. Check www.bb-elec.com for great info on serial networks.

No, because if you use Ethernet or fibre optic media then distances depend upon many more factors. And then there is radio.

 

It is not a MODBUS question - but a physical interface (PHY) question - ask the same questions for RS232 or RS485 - and at which speed (Baud rate).

Meir

 

> Typical Modbus RTU is RS-485 standard. 4000 feet (1200 meters) is generally
> accepted as maximum length.

is it for one loop or is it for whole topology? Say if three transmitters are connection on one line, then total of all three lines and distance from JB to control room is to be considered.

 

Whats the max length for Modbus over Smoke-Signal then please

 

But, is there any maximum size for this cable?

For example, can i use 24AWG for a distance of 50m?

 

> Whats the max length for Modbus over Smoke-Signal then please

Depends on weather conditions, atmospheric conditions and time of day as do most proprietary protocols when they cease to operate reliably (who hasn't wanted to consult a medium when what used to work doesn't?)

My comments derive from conventional Modbus smoke signalling, given that attempts to harness environmentally-friendly smoke signals from Mt. Helens proved to be a disaster (Vulcan, Hephaistos and Pele, 1980), as we all know.

In the US midwest with its relatively high humidity and flat lands, smoke signals have an effective limit of about 4-5 miles or less due to haze in the summer, further on clear winter days, but operator frostbite from wind chill due to handling and waving the wet blanket is an operational issue in the winter, particularly this past global warming winter.

In the high plateau of the Rocky Mountain states, smoke signalers report effective line-of-sight up to 16-18 miles on flat plateau, even greater distance from mountain elevations.

Eastern Appalachian mountain distance is probably the worst, sometimes as little a mile or two.

Smoke signalling is still banned in Massachusetts in the last remaining "Blue Law" from the puritanical mid-1600s.

Anecdotal reports from England and Scotland are quite negative on smoke signaling due to near constant fog, mist and rain conditions. I'm told that the latest IEC committee research couldn't even locate a smoke signaler in tropical regions due to annual monsoon seasons.

Since the French invented semaphore but Germany controls SAP (Semaphore Alliance Protocol), smoke signalling is virtually unknown in Europe.

As in all smoke signaling, which is in lower media layer of OSI (level 1 + possibly level 2), the key is keeping the blanket wet, regardless of protocol.

As to diagnostics, the noobie use of dry leaves instead of wet leaves to create a smoke column is quite common as a quick review of the literature and control fora will show.

Poor smoke density generation is the analog to bias problems, as smoke dissipation due high winds is to ground loops. Same old, same old, no matter what the transport layer. Ain't it the truth?

It is my observation that just as the official Modbus standard has been unofficially modified over time (the real standard is the de facto standard) in areas like a defacto default serial 8-N-1instead of the 'official' 8-E-1 and the near disappearance of 2 stop bits for 'no parity'; smoke signalling has dropped start bits, stop bits and parity altogether. If you had to wave that heavy, wet wool blanket three times for every byte, you'd ignore those bits, too.