Profibus DP in SIVACON MCC installation

I am trying to troubleshoot a Profibus DP installation inside a large Siemens SIVACON motor control center and I hope somebody can help me out.

I am most familiar with using Profibus DP with conventional Profibus connectors, where you connect 1 or 2 violet Profibus cables to a big connector with a DB9 shell, and have a selector switch for termination. The more expensive ones can even run up to 12 MB/s.

This system uses terminal blocks and a trunk/drop configuration. There is a Profibus DP trunk made up of the violet cable going from terminal block to terminal block in the back of the MCC sections (up to 27 terminal block sets in the trunk) with 1 meter drop cables down to drives and Simocode Pro V motor starter/overloads in buckets.

The Simocode Pro V starters use the Profibus screw terminals, not a conventional connector with a DB9 shell.

The Profibus DP master is a DeltaV module, configured to run at 187.5 kb/s.

I realize this is on the low end of speed for Profibus DP, but every installation guide I have ever seen for Profibus DP says you should not build it with droplines. Yet this MCC was built this way by Siemens, and the Simocode Pro V overloads have screw terminals as an alternative to the DB9 shell connector, which implies that they are meant to connect to a non-connectorized Profibus DP network.

Is this the accepted way to wire up a Profibus DP network in an MCC at lower data rates ?

The DeltaV scanner is at one end and has a proper built-in terminator, but at the far end of the network instead of using a powered Profibus terminator they simply placed a 150 ohm resistor across A and B wires. Has anyone seen this kind of "informal" termination ?

The system evidently passed its factory acceptance tests but the customer is getting occasional "dropouts" of individual nodes and wants me to find the culprit.

Can anyone suggest some guidelines for 187.5 kb/s Profibus cable systems that do not use the conventional Profibus connectors ?

Ken,
Copper RS485 practices should apply. Try the CTI 2500-TAP, cut sheet attached.


Tom

Thanks for that link; I presume that the CTI device is for RS-485-like networks as well.

Every Profibus DP wiring guideline manual I have ever seen insists on a powered terminator. Yet this system lacks them.

All Profibus DP guidlines I have seen also prohibit droplines except when an active dropline tap is used. Yet this system was built by Siemens using nothing but terminal blocks.

The amount of reflection I am seeing in the signal is moderate, so I am going to recommend that an active termination block be installed, so that we can say it's "by the book".

I think there may also be installation problems because I have captured packets that are asymmetrical; they only show signal on Data A and not on Data B. I suspect that indicates an open wire on a dropline.

Has any other Forum member seen one of these MCCs from Siemens wired up for Profibus in this way ?

Ken,
The 2500-TAP is made for use on the TI505 / CTI2500 PLC RS485 remote I/O network which runs at 1.0 Mb. I have attached the pertinent section of the original TI manual. I haven't tried this same hardware on a Profibus installation myself; mostly because of all devices usually having the DB9 port available.

Tom

Does anyone else out there have experience with Profibus diagnostics using an oscilloscope ?

From what I am reading, a Profibus DP signal should be between 4 and 7 volts peak-to-peak differential voltage. I am seeing only about 2.5 volts.

Everything I am reading, from Siemens and from Texas Instruments and from the PTO, says that Profibus needs the active terminator at all data speeds and all cable lengths. But this customer says "a guy from Siemens told us we didn't need all that" at 187.5 kb/s.

I also cannot find a point where the shield has been grounded. It is jumpered with a green/yellow wire from device to device in the Simocode Pro V overload relay assemblies, but those overloads do not have a ground clip.

An example of a typical signal on this network: