Profibus Communication on S-7

D

Thread Starter

darshan

we have s7-400 (pcs-7,ver.6.1) plc with 417cpu and 4 cp for four different profibus routes with min 60 nodes on each route. we have got problem of link failure during running plant and got whole plant stoppage.

we used 1.5mbps speed on data transfer.
can anybody tell me that which type of care/maintenance should be done with profibus to prevent this failure?

how may repeaters we can use in one line and how may instrument earthing we can done in one route?
 
C

Curt Wuollet

If there is a way to prevent this, I haven't found it. Some plants work flawlessly and some do not with profibus. Connections are a big issue. Poor quality connectors will make life miserable. That is the first thing I address and that usually gets the plant running. But for some reason, with some networks, it doesn't seem to matter, they work even with crappy connectors or even mashed cables. And some, no matter what you do, and I've seen a whole network redone, are prone to issues that no one can figure out. But, I would start with the connectors.

Regards
cww
 
S
> we have s7-400 (pcs-7,ver.6.1) plc with 417cpu and 4 cp for four different
> profibus routes with min 60 nodes on each route. we have got problem of link
> failure during running plant and got whole plant stoppage.

Does the network come back by itself when it fails?

Does the plant stop because you need to be in continuous contact with the instruments or because the network failure causes the S7 to go into fault mode and that shuts everything down?

There is an OB you can install to handle (or ignore) network glitches and keep the PLC in run mode, but that won't help you if the disruption of the sensor data is equally damaging.
 
> we have s7-400 (pcs-7,ver.6.1) plc with 417cpu and 4 cp for four different profibus routes with
> min 60 nodes on each route. we have got problem of link failure during running plant and got whole plant stoppage.

> we used 1.5mbps speed on data transfer.
> can anybody tell me that which type of care/maintenance should be done with profibus to prevent this failure?

Do you have checked shieldings issues ? Please do measure the voltage of the shields related to ground at different locations.

> how may repeaters we can use in one line and how may instrument earthing we can done in one route?

You can use only 4 repeaters in one line ... that means every
communication path in your network should not include more than 4 repeaters.

Best Regards
Armin Steinhoff
 
Does the CPU go into Stop? If so you need to use OB85 or OB86 (can't remember which one) to handle the rack failure.

Have you looked in the diagnostic buffer to see if there's any clues causing the error?
Also ensure that the end slaves are fed from a permanent supply so that the termination resistor works correctly.

--
Tom Chubb
 
> how may repeaters we can use in one line and how may instrument earthing we
> can done in one route?

-----------------------------------------------
In our exp. with profibus-dp in SIEMENS PLCs/SIMOCODES/DCS (min 40-60 devices) the following were observation while facing Link failures.

I. General bus fault in a particular line and all the equipments loose communication with higher level system like DCS. In effect tripping of the down level process equipment.

Diagnostics with 3rd party Vendor using profitrace S/w tool:
1. Disturbed telegrams.

2. Close proximity of DP cables with High voltage power cables (Physical).

3. Cable shielding not done properly at Each node. (Physical)

4. Poor connectors used (Physical).

5. End resistors not utilized in some Long Lines having optical repeaters. (Physical)

6. Rated speed of 1.5Mbps found very high as many of the devices like S7-300 PLCs/SIMOVERT/SIMOCODEs unable to respond within cycle time.

7. Poor cable connections followed.

Corrective Action Taken:
1. Replaced the profibus cable where ever found not good in terms of quality in connections, and routing in view of the huge pile of cable routes.

2. Checked the signal quality in each nodes/ lines.

3. Installed procentec Hubs in some lines where chances of electrical noise suspected. This arrangement give us some confidence to isolate individual devices from the hub without disturbing the particular line/bus.

4. cable shielding done properly at each device end/ line.

5. Active end terminator resistors introduced in long lines.

6.Reduced the baudrate to 187.5Kbps after analyzing each nodes/ segments signal waveforms.

7. Checked the signal quality at each nodes/ segments as well as at Master(DCS) end and found to be satisfactory.

Conclusion:
Our corrective actions proved effective to resolve the frequent tripping as a result of Profibus general bus fault. Based on general comments from profibus network specialists client should have a diagnostics tool set up in order to check the bus healthiness once in 6 months or so in order to alert the team either instrument or electrical for appropriate action from their respective devices during plant stoppage.

You also can avail certification courses in profibus-DP trouble shooting.

Finally the number of repeaters/hubs max. allowed in a line:
The answer lies in the electronic components of the repeater. These determine the actual "maximum number”. The digitalization from and to the opto-couplers in the repeater will make bits smaller or wider. If you put too many repeaters in series, the bits differentiate too much and messages are not recognized anymore. There are also repeaters available with no limit in serial placement.

Check the manual of the repeater and should state the maximum amount. For example: the Siemens repeater is 9, the ProfiHub is unlimited.
 
thanks very much.

hear we have no on-line analysis of profibus signal available. we just check it through hardware route via changing connector, cable, and earthing of profibus so it may helpful.

can you pl. tell me the best on-line profibus tester which will reduce the break-down time and we can predict the failure.
hoping for a replay.

darshan
 
Top