PROFIBUS 16 devices problem

Hello,

I have a system based on S400 Siemens PLC. On PROFIBUS network are 16 devices, Omron, ABB invertors and SIWAREX modules.
We tried to add one more inverter and when we add him whole profibus cracks. When he is removed everything works OK. We tried to remove one of existing inverters and add new one as 16. and again net stops responding after 17. has been added.

Is 16 somekind limit for profibus?
Does someone had similar problem?

In the PLC .. Hardware Diagnostics, are there any error messages ?
Is the station address set correctly on the new inverter ?
Did you add the new inverter in the HW Configuration ?
Did you save and compile the HW configuration ?
Did you download the modified HW configuration to the PLC ?

JesperMP said:

In the PLC .. Hardware Diagnostics, are there any error messages ?
Is the station address set correctly on the new inverter ?
Did you add the new inverter in the HW Configuration ?
Did you save and compile the HW configuration ?
Did you download the modified HW configuration to the PLC ?

Click to expand...

-Yes
-Yes
-Yes
-Yes, when I do that profibus stop responding. I lose all comunication with all devices, SCADA...

There is no error messages when compiling and downloading.

Hello Lesa;

when we add him whole profibus cracks

Click to expand...

profibus stop responding. I lose all comunication with all devices, SCADA...

Click to expand...

How many SCADA stations do you have setup on Profibus? With a S7-400 you should have up to 32 connections for communications, so it should not be a problem, but still...

When you lose the bus, how does your CPU/CP react? BUSF LED, SF LED? Do you run OB86 and other diagnostic OBs in your program? Please check the info in the diagnostic buffer when you lose comms, it could be very helpful.

What is the length of your segment and bus speed used for comms? Is the segment properly terminated? You could try reducing the network to 500 kB (for example) and see if you still see the same problem.

It could be helpful if you look at the HardwareConfig editor ONLINE while the fault is present, it will point out errors in the bus symbolically (you can then look at diagnostioc info per module).

Hope this helps,
Daniel Chartier

Please state everything relevant about that Profibus network. You didnt say anything in the initial post about SCADA on the same Profibus network.

If you have SCADA (or any other masters) connected to the same Profibus network, then it is critically important that they are included in the calculation of the Profibus bus profile. What SCADA is that ? Is it integrated into the same STEP7 project ?
Did you setup this project, or did someone else ?

And please, when you have the error, then check the PLC .. Hardware Diagnostics in STEP7.

You might also want to check the Highest Station Address (HSA) in the properties of the network. Some programmers reduce it to the maximum required to (slightly) decrease the bus cycle time.

Lesa said:

We tried to add one more inverter and when we add him whole profibus cracks.

Click to expand...

Where do you add the extra inverter? Is it at the end of the profibus line or somewhere in between? And how do you connect it, between 2 devices or with a T-branch (how long)?
What is the profibus speed? How much cable is used in your system?

Lesa said:

Hello,
We tried to remove one of existing inverters and add new one as 16. and again net stops responding after 17. has been added.

Click to expand...

I don't understand this. Do you mean that the new drive works if you remove an existing one? The maximum amount of devices per network section is 32 if you respect the cabling rules, after device 32 you need to add a repeater (32 = 31 devices + the repeater).

Sorry guys,
now I'll explains a little bit better my configuration
I have two PROFIBUS networks. One is being used for communication with inverters and weighing modules and other for communication with SCADA. Both are 1.5Mbps.

dchartier said:

How many SCADA stations do you have setup on Profibus?

When you lose the bus, how does your CPU/CP react? BUSF LED, SF LED? Do you run OB86 and other diagnostic OBs in your program? Please check the info in the diagnostic buffer when you lose comms, it could be very helpful.

What is the length of your segment and bus speed used for comms? Is the segment properly terminated? You could try reducing the network to 500 kB (for example) and see if you still see the same problem.

It could be helpful if you look at the HardwareConfig editor ONLINE while the fault is present, it will point out errors in the bus symbolically (you can then look at diagnostioc info per module).

Hope this helps,
Daniel Chartier

Click to expand...

-I have one SCADA station.
-BUSF led blinks.
-In Online mod in HW config all communication has being lost. All devices has be unavailable. (little red line over address number)

JesperMP said:

What SCADA is that ? Is it integrated into the same STEP7 project ? Did you setup this project, or did someone else ?

Click to expand...

SCADA is WinCC, not integrated. I made PLC program, SCADA is made by colleague.

RGB said:

You might also want to check the Highest Station Address (HSA) in the properties of the network. Some programmers reduce it to the maximum required to (slightly) decrease the bus cycle time.

Click to expand...

HSA is 126. Should I change it?

Jobbe said:

Where do you add the extra inverter? Is it at the end of the profibus line or somewhere in between? And how do you connect it, between 2 devices or with a T-branch (how long)?
What is the profibus speed? How much cable is used in your system?
I don't understand this. Do you mean that the new drive works if you remove an existing one? The maximum amount of devices per network section is 32 if you respect the cabling rules, after device 32 you need to add a repeater (32 = 31 devices + the repeater).

Click to expand...

-Extra inverter is added at the end of bus. But we tried to add him in middle of profibus - same problem. We tried to remove some of working inverters and add new one on his spot. It works fine. And when I add that "old" inverter as 17th device, profibus dies.
- I don't know how is it connected, firm that sold us inverter installs it.
-Not a lot cable. Branch for inverters maybe 20m and branch for SCADA maybe 40m
-Yes. Only 16 devices works, it doesn't matter which 16, new ones, old ones, this direction of cabling, that, addresses changed...

Your problem would make sense if Profibus is somehow seeing each of your devices as TWO devices. 2 X 16 = 32, and when you add the next one, Bingo! you have exceeded the limit....(Just a thought to brainstorm your thinking).

Turn the speed right down...187K or less (97% of most devices autobaud so only the master needs to be done). If the problem goes away it's cabling or termination...if it's still there then you have other problems. If a few nodes don't come up try power them down and up. Some devices only check speed once.

I got a suggestion. Make a sketch of your network and scan it for someone to look at. Indicate all nodes and distances also show master and where you terminated etc.

Nick

If you just want it to work without working out problem? Slip in repeater in middle terminate via built in switches on both sides and your done.

Have you tried attaching the device to the SCADA network to see what happens? In this way you could just hook it up on the bench and add 24V.

Hello Lesa;

This describes the meaning of the BUSF LED for the integrated DP port on a S7-300 CPU:




Seen from here, it looks like it is either a cable/termination problem (less likely, as you have exchanged drives and le bus worked) or one of the stations camnnot be addressed (e.g. 2 slaves have the same address).

If the problem still persists, I would start over from the beginning: check all slave addresses on their own hardware/software, rebuild the Profibus configuration in HWConfig and download to the CPU.
Do you have the diagnostic OBs loaded in your project, as suggested before? If so, what diagnostics are in the diagnostic buffer?

I have never seen a Profibus setup show the reaction yours does unless there is a fundamental fault in the design.

Hope this helps,
Daniel Chartier

I am not sure that this is what causes the fault, but I am very sure that it is a problem with not configuring the other master (the SCADA PC) in the configuration of the Profibus Network.
You may have exceeded the Target Token Rotation time by adding the other station, thereby causing the fault.

The best remedy is to integrate both PLC and HMI into the same project. Even if you are working on it separately, I recommend to do like this anyway. Then in NetPro you compile everything to get the correct Bus Profile calculated.

If you cannot integrate the HMI in this way, then for the Profibus Network, go to Network Settings tab .. Change Profile from DP to Standard, select the Options button, Network Stations tab, select Include Network Configuration Below, then increase Active Stations with FDL/FMS/S7 communication load with 1.

------------------------------------------------------------------

Here is a nice lecture for anyone using Profibus.

1. Configure a simple netwok with 1 master and maybe 2 slaves.
Then check the Bus Parameters, in particular Ttr (Target Token Rotation time).
2. Now, like described above, change the Bus Profile from DP to Standard, and add one more "Active Station with FDL/FMS/S7".
Again check Ttr.

Questions:

What happens with Ttr ?
What is the reason for the change of Ttr ?
What can happen if there is a master that is not accounted for ?

And for the 3rd time:

And please, when you have the error, then check the PLC .. Hardware Diagnostics in STEP7.

Click to expand...

Hello guys;

First, thanks to Jesper i had not thought of the possibilitythat the SCADA could be taking to much of the TTR; worth investigating in any case. The simplest would of course, as you suggest, that the SCADA machine be included in the Step 7 project.

Next, for Lesa: at no point have you come back to us with the contents of the diagnostic buffer. Possibly you do not know how to access it; in that case here is a help file (from Step 7 Lite, should be similar in Step 7):




May I suggest that when the problem appears again yopu open the diagnostics buffer, save the entries as a text file, and then copy those here so we may examine them?

Hope this helps,
Daniel Chartier