troubleshooting Siemens Simatic S7-300 PROFIBUS "BUSF" failure

Hi guys,

Tomorrow I'm heading down to take a look at an existing SCADA system that has failed, and I was wondering if anyone had some suggestions for what I might keep an eye out for. I've never worked with this particular site so I'm not entirely sure what to expect, though I am lucky in that the company I work for installed it (back before I worked here) so I have some pretty good system level documentation.

The system is a water treatment plant, and the master PLC is a Siemens Simatic S7-300. The specific issue is that on the master PLC, there is a "BUSF" failure. From what I've gathered from some Google searching already, this failure indicates that there is a comms failure between the master PLC and one or more of its slaves, which are communicating via PROFIBUS. The system operator has told me that the problem started after a recent storm (with lightning). They've tried power cycling everything but that does not seem to help.

I know it's a broad question, but if anyone had any suggestions for what I might look for, and/or what the BUSF failure could indicate, I'd be most appreciative. Thanks all!

BUSF can be for Profibus or Profinet.
In addition to the BUSF lamp, what has stopped working ?
Do you have STEP7 software ?
Do you have the source STEP7 project ?
You should use STEP7 to try and go online with the S7 PLC and check its diagnostics buffer, and to check the hardware diagnostics.

edit: Describe the entire setup in detail, what HMI and IO hardware is used, and how is it connected with each other.

Hi Jesper,

It's definitely PROFIBUS, so I know that much. I'm currently onsite and the whole system is down; basically the system consists of a main control panel at the water treatment plant (Filter Control Panel, or FCP) and the master PLC is the S7-300. There are two local slave PLCs onsite (filter controllers) and one remote slave PLC that operates a pump station from a nearby lake.

I do have the original software so I've been pouring over everything that I have, but one thing that alarms me the most is that I can't even connect to the slave PLCs locally at the panel with a laptop.

I do indeed have the STEP7 software and when I try to connect to the two local PLCs, I keep getting communications errors. I'm using a PPI adapter via a serial port at 19.2K and it's just not connecting, at all. I've tried several addresses and every conceivable iteration of settings just to be sure, and I can't even connect to it at the control panel. I'm seriously beginning to wonder if the COMM ports are fried? Don't know how likely that is but if I can't even connect locally, what else can I do?

Quick update: I've finally been able to talk to the local slave PLCs; I had to unplug them from PROFIBUS and when I did that, they were able to communicate with the laptop. So now I'm trying to figure out why the master PLC still doesn't see them.

Use MPI adaptor for S7-300. Set in PC/PG Interface.

Are all slaves powered and operable?

Are address dials/dips correctly set?

Are profi cables intact?

Are terminating resistors correctly set at ends of bus? Typically on where single cable enters DP connector... but system setup deterrmines this.

Are DP connectors with looped out cable switched off at the terminating resistor?

Any non-standard joints in the cables, or wet components?

Hi Silva, thanks for chiming in. To answer your questions and to give a general update:

1. The two local slaves appear to be functional; they have no fault lights and if I unplug PROFIBUS and plug in the S7 adapter, I can talk to them. When PROFIBUS is plugged in and I go to the second port, however, I cannot get it to talk to my laptop.

2. Terminating resistors appear to be functioning correctly.

3. We found out after arriving that the PROFIBUS connection at the remote pump house apparently has its connection through 18 gauge solid core electrical wire, not twisted pair, shielded cable. Don't know if that's the connection for a small part or the whole run of a couple thousand feet. Definitely a major concern. We isolated that run and plugged the two local PLCs into the PROFIBUS network alone; they were still not recognized.

4. At this point we think the CPU's PROFIBUS port is fried, and/or the PROFIBUS physical connections/cables could have some sort of problem. The DP lights never even came on at the local slaves, which if I understand the manual correctly means that there was never even an attempted connection between the master and those slaves.

Try the hardware config and go online.
Now you can see witch slave there is connectet or missing.

Also look in diagnostics buffer (ctrl + D) and see witch slave is missing or in faliure.

JERA said:

Try the hardware config and go online.
Now you can see witch slave there is connectet or missing.

Also look in diagnostics buffer (ctrl + D) and see witch slave is missing or in faliure.

Click to expand...

Hi Jera,

We did that, yeah; all three slave PLCs were shown as unavailable for the entire time we troubleshot the system even though I could connect to them locally via a laptop.

Are you using a PPI adapter ?
You cannot use that on a Profibus system or the MPI ports on older S7 CPUs.
Maybe you mean a Serial "PC Adapter". The 19.2 kbaud tells us that it is ancient, and horrible to use for troubleshooting.

The problem can be as simple as your laptop having the same Profibus node address as the CPU.
The problem can also be electrical, something to do with the cables or terminations, etc.
But without a reliable online connection it is almost impossible to troubleshoot.

I suggest that you make a detailed description of your system.
Describe the entire PLC system, all slaves, the SCADA, and how everything is connected, including which ports.

Hi Jesper,

They had both the PPI adapter and the S7 "PC Adapter"; PPI adapter didn't work as you suggest, "PC Adapter" did.

I was able to get ahold of Tech Support from the regional Siemens distributor and (quite a long) story short, we've come to the conclusion that the CPU is likely damaged so we've ordered a replacement.

Next time I am down to the plant, I will make a detailed list of the system. Off the top of my head:

1. One master (PROFIBUS addr 2) (316-2 CPU) with various I/O racks in its control panel.
2. Two slaves in the same building (S7-200), addr 11 and 12; PROFIBUS goes from 12 (term resistor off) to 11 (term resistor on).
3. One remote slave at pump house (S7-200); addr 21; term resistor on.
4. At master, the remote slave plugs into the PROFIBUS port first, then the local slaves stack on top of that. This is on the X2/DP port.
5. Local PG/PC is Windows 2000 machine, connected via the X1/MPI interface via a CP5611 card. Also has the S7 PC Adapter plugged into the same port (maybe to talk to the S7-200s?)
6. Trying to contact the slaves via the PC/PG was entirely unsuccessful. Plugging into the PROFIBUS connector via "PC Adapter" at the local slaves allowed the laptop to communicate but once they were plugged back into the PROFIBUS net and the laptop was plugged into Port 0 (Port 1 is PROFIBUS) they wouldn't talk to the laptop.

That's as much as I can remember off the top of my head, if there's any other info that would be helpful to know please mention it and I'll do my best to either remember it or glean it next time I visit the site.

Thanks!

You probably mean ET200M or ET200S rather than S7-200.

if the PG/PC has a CP5611 card, then it does not need additionally a PC Adapter.
The CP5611 card is much better for troubleshooting. You can use the Diagnostics for the CP5611 card to scan for DP nodes, active and passive (slaves).

The use of 316-2 as the CPU, Windows 2000 and a PC Adapter (slow version with 19.2 kbaud) tells us that it is a quite old project (15+ years ?).

As to the original problem, for a bus system that spans a relatively large area, and with some cables going between buildings, which I believe is the case here, there should be overvoltage snubbers on the bus cables where the cables enters an area or building.
Today, the solution would have been to go Ethernet via fiber.

JesperMP said:

You probably mean ET200M or ET200S rather than S7-200.

Click to expand...

I don't think so, the slave PLCs that I was interfacing with said "SIMATIC S7-200" on the bottom left of the case and when Google searching for ET200M, what comes up isn't what I was working with.

if the PG/PC has a CP5611 card, then it does not need additionally a PC Adapter.

Click to expand...

I was wondering about this, thanks for clarifying. Seems like the PC Adapter is wholly unneeded if the CP5611 card is there but I didn't want to make any assumptions.

The CP5611 card is much better for troubleshooting. You can use the Diagnostics for the CP5611 card to scan for DP nodes, active and passive (slaves).

Click to expand...

To the best of my abilities I was doing this, or at least something similar to it. I had my contact open the Manager application to get software versions and when he tried to open the project file, it gave him an error about not being able to connect and said something about "PROTOCOL" (I was watching via a blurry Facetime connection via iPhone). There was something called "available nodes" that we opened and that allowed us to at least interface with the Master PLC, but not the slaves. I'm not sure what that means exactly, but the Siemens Tech Support Rep I've been talking to referred to that as "possibly dire", which is interesting.

The use of 316-2 as the CPU, Windows 2000 and a PC Adapter (slow version with 19.2 kbaud) tells us that it is a quite old project (15+ years ?).

Click to expand...

Yes, very old! The company I'm working for designed it originally (or at least had something to do with it, details are a bit vague), and we are actually gearing up to replace the whole system with Allen-Bradley PLCs; in the meantime, we're trying to get this system to hobble along; this system controls the filter plant for taking water out of a nearby lake, and without it, they're relying solely on wells that are running 20+ hours a day...a dire situation, especially in CA summers!

As to the original problem, for a bus system that spans a relatively large area, and with some cables going between buildings, which I believe is the case here, there should be overvoltage snubbers on the bus cables where the cables enters an area or building.

Click to expand...

I can check the documentation available to see if there are any snubbers, at least on the original plans. The system is so old though, that I'm not sure what's there now versus what was there back in 2002 or even before that.

Today, the solution would have been to go Ethernet via fiber.

Click to expand...

That's exactly what we're planning for the future system overhaul.

Thanks for the insight Jesper!

defcon.klaxon said:

I don't think so, the slave PLCs that I was interfacing with said "SIMATIC S7-200" on the bottom left of the case and when Google searching for ET200M, what comes up isn't what I was working with.

Click to expand...

It is possible to use S7-200 PLCs as Profibus DP slaves. It is a little bit unusual, but possible. Would have been an 'economic' way of making a distributed control system.

defcon.klaxon said:

To the best of my abilities I was doing this, or at least something similar to it.

Click to expand...

What can trick you is that the masters are indicated by a prominent [v], whereas the slaves are indicated by a change of color of the [ ] from pale grey to white. And that can be easy to miss depending on the monitor you are looking at.

defcon.klaxon said:

To the best of my abilities I was doing this, or at least something similar to it. I had my contact open the Manager application to get software versions and when he tried to open the project file, it gave him an error about not being able to connect and said something about "PROTOCOL" (I was watching via a blurry Facetime connection via iPhone). There was something called "available nodes" that we opened and that allowed us to at least interface with the Master PLC, but not the slaves. I'm not sure what that means exactly, but the Siemens Tech Support Rep I've been talking to referred to that as "possibly dire", which is interesting.

Click to expand...

I suggest to do this.
On the Windows 2000 PC, use the Set PG/PC Interface function to open the diagnostics for the CP5611 card. Here is the searching function for DP nodes i mentioned above, and the indication of slaves with a color change pale grey to white.

JesperMP said:

It is possible to use S7-200 PLCs as Profibus DP slaves. It is a little bit unusual, but possible. Would have been an 'economic' way of making a distributed control system.

Click to expand...

This would make sense; the system we're working on is for a very small town and as such they are going to be budget minded.

What can trick you is that the masters are indicated by a prominent [v], whereas the slaves are indicated by a change of color of the [ ] from pale grey to white. And that can be easy to miss depending on the monitor you are looking at.

Click to expand...

Thanks for the tip, I will be sure to keep this in my back pocket.