S7 Profibus Fault

We have a S7 416-3XL00-OABO CPU . We are using Profibus to
communicate to SEW Drives , Fanuc Robots ,Encoders, HMI's , etc. Last evening we had a problem with 6 devices that came
off of a Siemens 6ES7-972-OAAO Repeater ( may not be the whole
number) It was 3 Fanuc Robots ( 2 RJ 2's and 1 RJ 3 ), 2 encoders , and a Siemens HMI. The CPU faulted with a EXTF Led
and IFM1F Led. We replaced the Repeater because we had a weak
DP2 Amber Led light on the repeater. When taking the wires off ,the tech took notice that one of the Profibus wires seemed broken ,they are solid wire. We replaced the repeater
but still were not able to clear the fault ,or the 6 devices
on that part of the network. Just to see if we could cut that portion of the network in half ,on the Fanuc Robot we pushed
the terminating resister in the up position (or Down)it still did not reconize the "Nodes" ,but when we put the terminating resister back to the previous position ,the fault went away.
Late today we had some random faults with the same section of the network ,we thought we had it solved.
We have had this machine for almost 9 years,and never had Profibus issues ,so the installation is sound.A few questions.
- Have any of you had similar problems?
- Would you have any ideas what could be happening?
- It is somewhat of a rough enviroment, but have been very
stable for years. We checked all the cables going to the
devices in question ,they do not look damaged.

Thank You so much in advance,

Rob S. said:

We have a S7 416-3XL00-OABO CPU . We are using Profibus to
communicate to SEW Drives , Fanuc Robots ,Encoders, HMI's , etc. Last evening we had a problem with 6 devices that came
off of a Siemens 6ES7-972-OAAO Repeater ( may not be the whole
number) It was 3 Fanuc Robots ( 2 RJ 2's and 1 RJ 3 ), 2 encoders , and a Siemens HMI. The CPU faulted with a EXTF Led
and IFM1F Led. We replaced the Repeater because we had a weak
DP2 Amber Led light on the repeater. When taking the wires off ,the tech took notice that one of the Profibus wires seemed broken ,they are solid wire. We replaced the repeater
but still were not able to clear the fault ,or the 6 devices
on that part of the network. Just to see if we could cut that portion of the network in half ,on the Fanuc Robot we pushed
the terminating resister in the up position (or Down)it still did not reconize the "Nodes" ,but when we put the terminating resister back to the previous position ,the fault went away.
Late today we had some random faults with the same section of the network ,we thought we had it solved.
We have had this machine for almost 9 years,and never had Profibus issues ,so the installation is sound.A few questions.
- Have any of you had similar problems?
- Would you have any ideas what could be happening?
- It is somewhat of a rough enviroment, but have been very
stable for years. We checked all the cables going to the
devices in question ,they do not look damaged.

Thank You so much in advance,

Click to expand...

I'll take a shot at this one.
Your network devices are powered up through different power sources I assume.
A momentary power loss on any of the devices (in the field) will fault the Profibus network since a node went missing.
I'd check for power supplies,lose wiring etc as you have already verified network cabling and terminations.

If I am not mistaken, most of the factory made Profibus cables we have used are all made with solid single strand wire. I have found in the past that cables can can routed in and thru some tough angles in order accomodate panel spaces. In the course of troubleshooting, upgrading panels these cables get pushed and bent and have resulted (in my past experience) in the wires pulling out of the cage clamp connector in the Profibus connector causing intermittent comm faults. I would inspect these if the problem returns.

Profibus DP is solid wire, Profibus PA is multiwire (dont know how it is said in english when one "wire" consists of multiple wires inside same insulation? please enlighten me

Do you have diagnostics programmed in your cpu that it saves dp slave addresses that have error?

Let's consider that we have the "nodes" in the following
order, HMI , Robot 1, Robot 2, Robot 3 , Encoder 1, Encoder 2.
Let's say that we have a semi broken ,or loose wire at Robot 2
would we lose Robot 2 , 3 ,Encoder 1 , and Encoder 2 ,or would
we lose everything from the repeater , meaning all 6.

For the record, of course there are Profibus DP cables with multi-stranded wires. These are used whereever the cable must be flexible, for example with robots. It is just that the most typical Profibus cable only has a single strand.

Rob. If a wire is broken so that there is an open end, then it will upset the entire Profibus network in the segment (the one side of the repeater). The nodes downstream will obviously be cut off, but the rest of network is also affected due to the reflections caused by the unterminated wire.

Hi. Rob S.
Did you found out what the propblem with det profibus?
Because I have the same problem with a network after a Repeater, but this happens just ones a week. Yesterday I`ll looked att the repeater and det DP2 LED had a weak amber and shimmered.
Sow I disconected the Repeater and mounted it correctly (it was hanging), then the DP2 LED become normal.

Could this have been the problem, and what does the weak amber and shimmered LED tell me?

Regards.

Good Morning Kas ,

What we found out was ,we had a loose Profibus wire on the repeater.
we tightened the terminal and we got a strong amber led.We thought we had it after that but still had the fault. We started looking at the nodes after the repeater ,and we moved the switch on the profibus plug
on the robot , to the "end of the line resister position" and back again and the fault went away. We still sometimes get a fault about once every 2 weeks.

Exchange the Profibus connector in the robot.
The chance that it has gone bad is relatively high.

I agree with JesperMP.
Try replacing the connectors (the switch for selection end-of-line or next output very often is the problem).Dirt in the slide- contact seems to be a factor, here.
Good luck.