You can potentially damage the 485 receivers with too many coupled transients, particularly over a 400-ft run.

TTFN

FAQ731-376: Eng-Tips.com Forum Policies

Quote:

Electromagnetic-compatibility and application requirements determine whether you need a shield. A shield both prevents the coupling of external noise to the bus and limits emissions from the bus. Generally, a shield connects to a solid ground (normally, the metal frame around the system or subsystem) with a low impedance at one end and a series RC network at the other. This arrangement prevents the flow of dc ground-loop currents in the shield.

http://www.edn.com/archives/1996/080196/16df5.htm

David Baird

Sr Controls Designer
EET degree.
Journeyman Electrician.

400ft should be no problem at all.  I do it all the time.  The stuff that was put in was twisted pair correct?

The dedicated conduit is great.

What are the 485 devices? Just one on either end correct?

Hopefully the devices are edge limited or controlled.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

Keith.
It's really new for me. We always request shield twisted pair cable for RS485, it's also recommend at all manuals ( of corse what I know). More then this, we always design  ground connection to the shield ( for example: only one end of shiels is grounded).
But, I'm sure you have good expirience w/o too. Nice know.
Regards.
Slava

What else are you running in the conduit?

TTFN

FAQ731-376: Eng-Tips.com Forum Policies

slavag I would normally use shielded twisted pair just to avoid unpleasantries.  But I have run 485 more than a thousand feet thru a hidieous factory by just using ancient unshielded CAT3 phone cabling thru half a dozen patch boards with no issues.

Many times I have used unshielded for temporary instrumentation, "grabbing what I have", and had no issues either.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

I agree with Keith - I don't think you'll have a problem as long as you're not running any high current wiring in the same conduit.  Steel conduit is a good thing to have.

Shielded cable is better, and certainly recommended, but RS-485 is a differential signal so it is more robust than a signal referenced to ground like RS-232.  

Thanks, guys.  It's part of a data acquisition system with nine devices.  It isn't critical equipment, but I don't want to be called back if it doesn't work.  The equipment manuals spec shielded twisted pair, but the electrical contractor pulled in unshielded. We're debating weather to leave it as is and cross our fingers or have him repull the wires.  From Dave Baird, it sounds like the conduit may not be an effective shield even with only the RS485 conductors, but from Keith Cress, it sounds like it may work any way.  I'll have to sleep on it.
Thanks,
John

jfpe it also depends on the data rates.  If this is just 4,800baud or 9,600baud instrument traffic, pfffft.

Look at Ethernet, 1Gb/s and no shielding.  Look at your DSL 10,000ft+ 2~3Mb/s no shielding.  Granted these examples have a robust comm protocol.  Hopefully the instruments are designed correctly.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

It also depends on what the 'devices' are you ar connecting the RS485 to. I'm in the VFD/VSD business and an awful lot of people are a little surprised when connecting their RS485 to a VFD, switch it on and wonder why everything has stopped communicating. Follow the installation guidelines of the devices you are connecting to and you should not have problems. The guidelines provided by RS485 do not always take into consideration that you may be connecting to devices that may be emitting high levels of RFI and if the bonding is poor (usual) then the convenient route for high energy RFI can sometimes take a trip down your RS485.

The equipment that the company I work for typically connects to building management systems via Modbus, implemented on an RS-485 link.  The installations often times use CAT-5 (unshielded) and it works fine.

The performance would not suffer from unshielded cables if they have their own conduit.  Although the differential nature of RS485 ensures a better noise immunity, the common-mode noise has to stay within certain limits.  The common mode noise includes the ground noises between the two systems that are connected together.  The one thing I would be afraid of is when the units are powered on or off at each side.  Make sure that the grounds at each side are solidly connected together.

Ethernet has an electrical isolation through its transformers.  It is not the case with RS485.
This is what makes Ethernet so solid, and so popular in the telecom world where whole CPU racks are made from multiple CPU cards talking to each other through their Ethernet port, instead of through a system backplane.

Hi jfpe
I think Slava is right as the twisted pair has to be shielded [screened].
First of all the Cat 5 cable type is obsolete and Cat 5 e is instead now.
Second, if the insulation is PVC and the conduit pass through manhole, the cable is exposed for 5-8 ft and water will penetrate between conductors. The capacity will rise and it will spoil all good performances sooner or later.
Third, the cables are still limited to a maximum of 100 m (328 ft) in length (normal practice is to limit fixed ("horizontal") cables to 90 m to allow for up to 5 m of patch cable at each end).
So, if the total length is about 400 ft the pair has to be shielded at least.
Best Regards

Hi 7anoter4,

I have never heard of PVC suffering from water penetration problems before. XLPE suffers from this problem, which is why XLPE insulated cables are almost always sheathed in PVC. Some confusion?

Also the 100m limit applies to Cat 5 cable operating under 100BaseT conditions - using a relatively low frequency transmission such as RS485 won't have the same problems, although loop impedance might become problematic over a very long run.
 

----------------------------------
  
If we learn from our mistakes I'm getting a great education!

Hi ScottyUK
With all due respect you are wrong, sir!
First of all the jacket is made of PVC –power cables mainly- due to other properties of PVC as mechanical resistance and flexibility-but there are other materials as hypalon, chloroprene[neoprene] , fire-resistant non-halogeneted materials and other. The XLPE insulated core has no problem with humidity for low and medium voltage cable.
For high-voltage cable, longitudinal and transversal water barrier [moisture barrier] is provided. PVC moisture resistance is low-not so low as polyamide [nylon] but still low.
Also the permittivity factor is high[ 7-10] and the tang.delta too. This material is indicated nor for telecommunication cable as dielectric attenuation is high neither for medium voltage cable.
In the old good time the telecommunication cable core was paper insulated and lead sheathed. All the electrical properties as capacity and capacity unbalance was very good and for long time.This cable was very expensive .So the polymer used to be employed instead.No cable –for long and medium distance –was insulated with PVC except switchboard cable-short length and in dry location.For outside location the cable was insulated with LDPE[Low Density Polyethene] namely foam[cellular]and as the PVC jacket was not moisture barrier a sort of Petroleum Jelly used to fill the air gap between core and jacket.After a while, a moisture barrier of longitudinal applied aluminium foil, both side poliethene covered was employed, so PJ was not required yet and the core was insulated with solid HDPE[high density poliethene-as LDPE seems to be carcinogene].Up today telecommunication cable insulated PVC is limited for dry location and limited length.
By the way, the signal attenuation for 100 Khz or 1000 Khz does not seems to differ.
Best Regards

Hi 7anoter4,

Seems we both may have some errors then....

Expand a little on this statement: "Second, if the insulation is PVC and the conduit pass through manhole, the cable is exposed for 5-8 ft and water will penetrate between conductors." How will this occur? Water penetration of PVC sheathed cables doesn't take place to any practical level. If PVC allowed water to penetrate the cable to any significant level, why would cables designed for service in wet environments be sheathed in PVC?

Signal attenuation? I don't disagree. I do wonder why you think there is a 100m limit when using the cable with RS485. The spec for 100BaseT permits 100m over a Cat5 cable. I don't think there is a published standard for distance using RS485 over Cat5 cable, but if there is feel free to post a link to it. RS485 can run up to about 1200m with appropriate cable. I doubt that this would be achievable with Cat5, but to imply that there is a 100m length restriction using RS485 over Cat5 based on the limitations of 100BaseT over the same cable is more than a little dubious.

  

----------------------------------
  
If we learn from our mistakes I'm getting a great education!

Hi ScottyUK
I agree with you is no standard where the length of cable for RS485 is limited to 100 m. But Cat 5/5e cable performance characteristics and test methods are defined in TIA/EIA-568-B.2-2001.I have not this standard now, but - if I well remember-there is somewhere stated the maximum length. I am sure it is possible to use it for a longer distance even the signal amplitude is under the value required by this standard but still tolerable, as you said.
I don't agree with you about using PVC jacket in wet location. Only PVC jacket does not avoid water penetration into the cable. We need a moisture barrier as aluminum foil and/or polyester wrapping over the core assembly, under the jacket.
For example, see
http://www.jembo.com/public/product.php?type=26
I agree with you the water penetration depends on water level in the ground.But if a manhole is involved, no matter how dip the  ground water level is, the manhole will be full of water for a while or for ever.
Best Regards

7,

As Scotty mentioned earlier, those length limits you mention are based upon a high baudrate 100BaseT signal... at the baudrates 485 is running, you won't have nearly the same issues with signal integrity.  Capacitance (the biggest killer here) and inductance will be less of an issue at the slower rates.  Therefore, you're going to get significantly longer distances out of the Cat5 cable than 100m.

Dan - Owner
http://www.Hi-TecDesigns.com

Thank you macgyvers2000.I agree with you.

Water ingress into wet location listed cable doesn't happen without a failure of the mechanical integrity of the insulation.  For signal cable the results of water ingress can be minimal, but for power cables the results of water ingress can be spectacular; doesn't happen all that often considering the amount of cable in place.

Hi.
Of course it's depend on the application, in lot of cases, today, for avoid all such " problems" we used RS485 to FO convertrs. FO and convertors are so cheapper.
Regards.
Slava

I am very sorry and I have to apologize for my bad behavior against ScottyUK.
My hate in front of PVC and mainly against Cat 5 has a long history. A long time ago I designed a web of 400-500 ductbanks c/w manholes for a fossil fired power station.
Through this web I routed about 30 pair cables –I am not sure if they were really cat5 but very similar-for a distance of 2000 ft[or more].I don't remember what frequency was the carrier [may be 100kHz].At the beginning –let say 2-5 years – all the system worked excellently. And suddenly all it was changed. The signal was jammed and the system lost its stability.
So I checked the routing. First of all the manholes were full of water. The dedicated steel duct for instrument cables was partial used for power cables –on a fragment was even a medium voltage shielded cable. The end patch was deployed from the original place and so on. We decided to substitute all these cable for a screened [shielded] pairs cables. After then no trouble of this kind was noted.
But as ScottyUK said may be was a change in carrier frequency and the 100Khz was not from the beginning. Never the less the shielded pair mended the outage.

No offence taken, so no need for any apologies.
 

----------------------------------
  
If we learn from our mistakes I'm getting a great education!

So 7anoter4, were you sending just 485 down those 2000 footers cables?

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

600+ meters of unshielded data running parallel to power?  Yikes!  No wonder you have a bad taste in your mouth about it.

Dan - Owner
http://www.Hi-TecDesigns.com

I worked once with RS422 in a chronometer system for those 1/4-mile car races.  We used a cheap Ethernet UTP cable. RS422 was chosen to meet the full-duplex mode requirement.  The system worked successfully at 9.6kbps for months, until a bulldozer cutted the cable, and it was replaced by a wireless system  <:(

We left the 4 unused wires tied to GND to get a partial shielding.  As stated before, twisted pair has a better common-mode noise balance on both wires, so we just left 100ohm shunt resistors at both ends for cable impedance matching, along with zeners and ceramic caps in parallel, to sink some line noise.  All data was Manchester encoded to prevent loading lines  (like Ethernet), and the bitrate was chosen low to stay away from any resonance.

Good Luck!

The shielding wouldn't have helped against a dozer...  LOL

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

haha, agreed :D