2 Pressure Transmitter 500m far From PLC

I had such experience even for completely analog case. But 500 m is significant distance.

(1) you should get a high quality cable pulled in maximum far from any power one;

(2) cable should be shielded and earthed on side of your amplifier (ADC);

(3) in case of using a parallel connection of two transducers (bad practice) be sure that input impedance of every sensor is the same. I used additional resistors to balance impedances, but it is not good practice as well and should be conducted with care;

(4) assuming bridge circuit of transmitters, it is recommended to apply so-called 6-wires scheme instead 4-wires;

(5) your digital(PLC) and analog schemes should be isolated and not share the same power supply;

(6) plus to (5) it would be good to employ isolated instrumental amplifer like this http://www.analog.com/en/subCat/0,2879,759%255F783%255F0%255F%255F0%255F,00.html

you need one more power supply for this option.

(7) your transmitters should posses very high isolated level, at least 4000 Ohm for >100V testing voltage.

(8) You'll get much less headache if you place ADC and controller near transmitters. I'm sure it's achievable.

Good luck.

Since I am still a junior in this field, please kindly advise me for below case;

We have plan to use 18AWG Instrument Cable, two wire connection, connected to ABB DCS,

1. so, do we have to used an independent 24 dc power supply for the field Instrument ?

2. If we use the amplifier, is it attached at the field instrument, or can be put at PLC panel ?

thanks a lot for your advice and attention

regards,.

Hi guest!

It's nothing. My recomedations were represented for all "undesired" cases. It does not mean you should apply them all at once.

ABB is good brand and transmitting current signal (4-20mA) much more easy than mV.

You said "We have plan to use 18AWG Instrument Cable, two wire connection, connected to ABB DCS". It means: you already have separated PW supply for you two transmitters parallel connected on field side. Isn't it so? As well it means you have amplifier (believed to be insulated) and mV/mA converter embedded at your transmitters. It's very good also.

I have not any idea about your DCS/PLC but they will got prepared signal. If you'll get troubles you'll resolve its.

So I hope all will go well. If you have some doubts I would advice you before installation check your system when cable is yet coiled.

All will be well. Do not worry.

<...18AWG Instrument Cable, two wire connection, connected to ABB DCS...>

Sounds OK. 20mA will pass down wet string, almost! 4-20mA is very secure as signal transmission goes, as it is practically immune to all bar severe attempts at interference. Provided the cable is something like a shielded, twisted pair, then that is industry-standard, and rarely will problems become apparent.

<1. so, do we have to used an independent 24 dc power supply for the field Instrument ?>

Not usually. The 24VDC power supply is usually, and often, configured to supply a multitude, if not all of the analog and volt-free digital digitals. And the power the instrument needs is derived from the 2-wire loop. That's the beauty of going this way.

<2. If we use the amplifier, is it attached at the field instrument, or can be put at PLC panel ?>

It depends upon the overall loop resistance as measured: the cables in the field, any series indicators in the panel and the input resistance of the analog input all contribute their ohms, and they all add up to a figure. Check the manufacturer's information on the instrument. Most of them have a chart showing maximum loop resistance for given supply voltages. Provided one is in the non-shaded part of the chart, i.e. the loop resistance is less than the minimum for that voltage, then operation will be OK across the whole of the 4-20mA range.

If a 1:1 current amplifier is used to compensate for high loop resistance, it will be found in the panel usually, though it is possible that these days most of the functions that would have cascaded from it in twentieth-century panels can be handled as digital signals in the PLC, HMI or DCS downstream of the analog input. In the configuration outlined in the original post, it would be unlikely to be required as even over that distance, the conductor resistance is relatively small compared to other parts of the loop, though it would need to be checked either against the cable manufacturer's published information or by site measurement to gain confidence.

If there is an excess of loop resistance, one of the solutions is to increase the power supply voltage. In this case, it is desirable to have a separate DC supply for the analog and another one for the digitals. The voltage on the analog one can then be cranked up a bit to compensate for the excessive loop resistance, to get back "on-side" on the chart. Most manufacturers' nominal-24VDC transmitters will operate quite happily on 36VDC, though it pays to check their instructions just in case.

I believe replies 2, 3, and 5 are all good answers. I would like to add that I do not believe a variable power supply is needed. The resistance of a 1000 meter loop of 18 gauge copper is about 10.8 ohms. Max loop resistance for transmitters are in the hundreds of ohms range. (Of course you need to add input impedance of the AI module.) Another thing is that you will need to calibrate the analog input. With Allen-Bradley systems there is offset and gain compensation settings through the software. If you use a transducer/transmitter then the input calibration corrects for transducer, transmitter, and transmission errors. If the transducer is connected to a meter/transmitter then calibrate the meter first and calibrate the input to the meter. Then the input calibration corrects for transmitter and transmission errors.

I do not believe a variable power supply is needed

Which is why I said that 500 meters is very possible and that he try just hooking everything up before getting a variable power supply.

If in case, the current is not getting through (for whatever reason), he would at least knowing that a variable power supply may be needed.

Every instrumentation book and instrument manual I've ever read says that 1000 meters is the upper limit for transmitting a 4 to 20mA signal. So, off hand, I'd say that 500 meters is very possible.

Most current loops use 24 volts for loop power. I'd advise you to use a variable voltage power supply that can go up to 32 or even 48 volts.

If you have a transmitter simulator, inject 20mA into the loop and measure the current at the analog input. If it's lower than 20mA, slowly increase the power supply's output until the current reads 20mA. Then add two or three volts more for good measure.

If you don't have a simulator, use the pressure transmitter as the current source.

Before you go and buy a variable voltage power supply, why not just try connecting everything to check if it works as it is. It'll save time and money.

<...use a variable voltage power supply that can go up to 32 or even 48 volts...>

Seconded!

It's handy to have the facility just in case, even if it isn't used.

Hey guy's

I do a lot of this kind of thing -- not hard (I have done up to 7.3 miles).

But it is necessary to be aware of the environment at both ends if your transducer is

on a chunk of pipe near a stand pipe on hill it can get dicey.

Worst case is use a lightning arrester of the sealed variety on the transducer, I use a Transtector unit built into a 1/2" pipe nipple it is specific for voltage and for 4-20 ma.

Use twisted shielded pair wire ground the shield at the transducer only.

At the PLC end use a powered low impedance input full isolator isolated at all three points the best place to put the separate power supply for the loop is at the transducer , but that is rare to have the power ckt there so at the plc is ok.

The power supply should be a fully isolated out put up to 2500vac rating.

Use a second lightning arrester at the PLC end with fuses fast blow type 1/4 a max. locate the fuses and arrester at the first available point in the PLC enclosure, ground the arrester well.

As far as loop power supply voltage goes make certain that the transducer can take the voltage at less than 1 ma and considering the voltage drops in wire and arresters, as well as the placement of the p/s and possible or likely shorting points.

All of this is for worst case , of course, but that gives the bigger and worst case lookouts.

FIRST WE SHOULD TRY TO CONNECT THE SIGNAL DIRECTY TO PLC WITH THE 5OOMETER CABLE, I THINK IT WILL WORK WITH OUT ANY SIGNAL LOSS.

EVEN THOUGH IF ANY SIGNAL LOSS IS THERE WE CAN USE SIGNAL ISOLATOR

HAVING IN PUT 4-20 mA & OUT PUT ALSO 4-20 mA WHICH IS TO BE INSTALLED NEAR TO TRANSMITTER.THE ISOLATOR WILL WILL TAKE CARE OF CABLE RESISTANCE & GIVE CORRECT OUT PUT AT THE INPUT TO THE PLC.

I just recycled my response from a question yesterday to control a relay 600M away.

If I were going to pay for the expense to pull a 500M cable then I would be dang sure to plan for future expansion. What would you do if next month you wanted to add more control points?

Many of the common PLC's feature open standard network communications. For example, a simple Allen-Bradley DH+ (Data Highway Plus) trunkline can span up to 3000M. It will be very easy to add more analog or digital I/O to the existing network.

Click for info:
http://www.ab.com/en/epub/catalogs/12762/2181376/214372/1491278/

Regards, Delmar Schmidt
Melfi Technologies Houston
www.melfitechnologies.com

If there is funding I would invest in a "profi bus" system or even an Ethernet system. you will then be making your sensors more intelligent, you will be able to move the data faster and farther.

Ethernet is only good for about 100M length cable segment. Don't know about Profibus

No offense here either. I comply with the KISS procedure for any of my projects. However pulling another 500M twisted pair for every I/O expansion will not be a simple task.

I agree that the older Ethernet networks are not real-time and can be easily bogged down by excessive network traffic. We typically specify newer Ethernet/IP, and always dedicate our network to only reading the loadcells and controlling our motor drives. We dedicate a second Ethernet/IP network card for communicating to a customer supervisory system.

It is possible to communicate long distances over Ethernet/IP using fiber switches. However the added complexity and cost would be overkill for this simple application monitoring only a few pressure points.

(why are the responses to this thread being hidden?)

Back to fundamentals-and KISS !

Problem: Reach a 4-20mA signal from point X to point Y (500m away)fairly accurately at minimum cost,max reliability.

Solution:

At X the originating 4-20 mA be dropped through series-- 2 small silicon diodes+500 Ohms. You get ~1V+(2>10)V signal . Call this C

(At X)-Take a small NPN silicon Transistor .Feed DCV ~16>20 to collector. Feed Cto base. Load 1000 Ohm Emitter to ground. You get ~0.5+(2>10)V at Emitter to Ground. Call D

Send D to 1Km >2Km> ( which is point Y )via a twisted pair and feed +D to base of another single transistor signal repeater like at C. Here you load Emitter to Ground=500 Ohm and take out 4-20 mA between Power supply +and Collector.