Simatic problem

M

Thread Starter

mareki

I have very big problem with Simatic controllers (CPU 316-2DP), exactly with communication between controllers and operating stations. Configuration of my system is: two controllers, and two operating station (WinCC 5.0) communicate each other by Profibus S7 (1.5 Mbit/s). Computers is equipped with communication cards CP5613. Distance between first and last component of the network is about 15 meters. Everything would be great, but the communication between PLC and OS is disconnected. It is happen irregularly, sometimes every 15 minutes, sometimes 5 times per minute. Sign of this effect is visible on the screen of the OS: all of the analog variables is dissapeared for 1-2 seconds. During this effect I has observed error number 8304 in Channel Diagnotic (software which is include with WinCC). Unfortunately I haven’t found description of this error in documentation of WinCC. Additionally I have observed that “Cycle Overflow” is incremential after disconnection. Amount of the variables which are read from controllers is about 900.
I have created very similar network in laboratory, but I have used CPU 315-2DP. In this case, everything is great – communication is not disconnected even once per week.
I have spoken to someone from Siemens, but they don’t know why communication is disconnected so often. They told me, that I should try to use acyclic read variables from PLC. It really helped it, but now refresh the variables from PLC is very, very slow (about 10-12 seconds). It is unacceptable. I have interested why the communication with CPU 315-2DP is great, and with CPU 316-2DP is not.

Is anybody who can help me???
Regards
MI
 
J

Josh Dalzell

OK, I can't promise anything definite, but these are the things that popped immediately into my head. I talked to a colleague who's pretty good with technical issues, and this is what we've come up with to at least get on the road to a solution. This may not apply as much as we hope, but we'll see.

What do the PLC Diagnostic Buffers say when the communication is lost? How much of the PLC memory is devoted to the running program? Does your program have a really big nested routine it is calling irregularly? Along those lines, is the process in the same state whenever you see the problem (maybe running a lot of data acquisition in addition to answering the WinCC poll)?

On the WinCC side, do you have channel logging and trace files activated? If so, this could cause some degradation. On the other hand, if not, turn them on and see if you can capture the data logged during the problem. This can be done in the Configuration tab under Channel Diagnostic. It saves the file under \siemens\wincc\diagnose.

I couldn't find any explanation for Error Code 8304, either. Found 8305 and 8404. Mighty close. Wish that counted for something.

What percentage of your data is coming into WinCC as analog values? Also, how much is Discrete?

I realize that this isn't a whole lot in the way of an answer, but it's a good basis for gathering everything to get on the way. I will try to get a hold of some people I know who are pretty good with this stuff, and see if they might be able to help.

I hope this has been of some assistance.

-Josh Dalzell
jd-hotline (at) unsuave.com
(at) .eq. @
 
P

Peter Hasler

I have experienced a similar problem with S7 315-2DP and wincc v5.0. The only difference was that my interface between PLC and SCADA was via ISO Industrial ethernet. The way i solved my problem was to remove a very nasty piece of code (that ate into my plc cycle time). With the reduction in cycle time my comms was vastly improved! Please check on the cycle time and let me know where it is (also give me the longest cycle time). Maybe we can sort it out this way.

Good Luck

Regards

Peter Hasler
Senior Control Engineer
Quad Automation
 
Few weeks ago, I thought cycle time is too long, and this is cause of my problems. Then, cycle time was about 500 ms. Later I shortened it to 300 ms (max 320 ms), and problem was still existing. So I decided to cut program, except DBs, OB1 and one small function. Then cycle time was about 4 ms. Unfotunately, connection was disconnected from time to time. I made big program in CPU 315-2DP (cycle time = 900 ms), and everything was great!!!

I think, it is not problem with cycle time.
I have observed in the Channel Diagnostic following parametres:

CPU 315-2DP:
AS Cycle: 3-4
Max AS Cycle: 65536

CPU 316-2DP:
AS Cycle: 3-4
Max AS Cycle: 4

When the AS Cycle is the same like Max Cycle, then connections is disconnected. Do you know what is AS Cycle, and Max AS Cycle?

Thank you very much for help!

Regards
MI
 
> What do the PLC Diagnostic Buffers say when the communication is lost?

Bus collisions.

> How much of the PLC memory is devoted to the running program?

About 89%.

> Does your program have a really big nested routine it is calling irregularly?

Only OB35 is called irregularly, every 250 ms. Cycle time of the OB35 is about 20 ms.

> Along those lines, is the process in the same state whenever you see the problem (maybe running a lot of data acquisition in addition to answering the WinCC poll)?

I haven't observed any processes during, and before disconnection.

> On the WinCC side, do you have channel logging and trace files activated? If so, this could cause some degradation. On the other hand, if not, turn them on and see if you can capture the data logged during the problem. This can be done in the Configuration tab under Channel Diagnostic. It saves the file under \siemens\wincc\diagnose.

I don't know, if the channel logging and trace file is activated. Unfortunately, this application is about 300 km from my place of residance, so I am going to check it next week.

> What percentage of your data is coming into WinCC as analog values? Also, how much is Discrete?

60% - analog values, 40% - discrete.


I was talking to Siemens last week. One of them told me (unofficially) that it is problem with CPU 316-2DP. In two weeks, I am going to test this application with CPU 318-2DP.

I am surprised, that following configurions is great:

CPU 315-2DP + CPU 315-2DP + OS + OS = no problem!

and there is problem after replace one CPU 315-2DP to CPU 316-2DP:

CPU 315-2DP + CPU 316-2DP + OS + OS = disconnection with CPU 316-2DP from time to time.

In the both cases, I downloaded the same program to both of the PLCs, and the same project in the operation stations.

Thank you very much for help. Sorry, it takes so much time to answer you letter.

Regards
MI
 
The only error discription I can find is 'FDL connection not established' This could be because there are not enough resources on the PLC ( or PC) ,or that the transmission is done to often ( PLC hasn't time to complete request)
Have a look in the Siemens manual NCM S7 profibus

That the only help I can offer
 
Top