I have three new Siemens Oxymat 6 analyzers connected to DeltaV using a new "M" series Profibus DP Interface card. At various baud rates, I experience frequent "drops" where communication changes from "BAD" to "GOOD" as seen by DeltaV Diagnostics. The analyzers appear to be responding as messages like "Configuration Accepted" appear in some of the diagnostic dialog boxes, which later changes to "online".
Hitting "update" in the "Profibus Device Information" dialog indicates the status is changing from "online" to "Configuration accepted" even though no download is going on. The status changes might be in synch with the communication drops, about every 5 seconds . . .
The instruments - Siemens O2 Analyzers - only have three settings for Profibus: Address, Tag, and "Ident Number" which the manual says is "used to set the configuration response of the device". Goes on to say "must have value of 1 to work with the provided GSD and DD". So it's set to "1". I downloaded what appears to be the latest GSD file from the Siemens website.
I only have slots configured for one analyzer (AI, DI's, Calibration) and when monitoring those they change from a status of "Good" to "Pic Device error". Any idea what that is?
The behavior has been the same regardless of baud rate - we started at 9600 and have tried 19.2 k, 93.75k, 187.5, and 500 kbaud. Statistics appear good - "valid responses" is 12000+ and increasing, No responses = 51 and steady. CRC errors are 0 for both transmit & receive, Asynch and Synch transactions both steadily increasing. Transaction timeouts is zero.
Double-checked for terminators on both ends - they are jumpered in at the extremes.
Running 13.3.1, MQ controllers, Profibus card is SW rev 23.8 and HW rev 1.50.
Any advice to eliminate the constant communication drops?
John, I hope someone with more recent profibus experience jumps on this call.
I did a little research on "Pic Device Error", but nothing conclusive. Google shows this could be due to a registry issue and offers a "tool" to fix your computer, or suggests you recover from a back up. Not convinced and certainly not ready to hand over my DeltaV computer to an Internet "tool".
Found some references but not your situation exactly. Mostly pointing to a misconfiguration (parameterization) of the device. One indicated the device had to be set to emulate a previous device version.
Have you checked the Profibus diagnostic screen:
Looking for things like Master Address, Station Not Ready which indicate binding of device to master. Since you get intermittent data, I expect it has an address but maybe some other flags will pop up and help diagnose.
I can't find a direct definition of what Pic device error means. One individual explains it is due to physical layer issue, but in one case was resolved by a setting in the device. In another, the wire length was problematic and they reduced baud rate.
No magic bullet for you. If you have a scope to observe the wave form, that might also reveal a physical layer issue, but I'm fishing. Obviously a badly deformed wave would implicate a physical issue.
Have you tried connecting a known good device to prove out the segment? Bring it down to one device?
Sorry I can't be of more help. good luck.
Andre Dicaire
In reply to Andre Dicaire:
Thanks Andre - I have been staring at the "Device Information" dialog box. No boxes are checked, unless I enable the "watchdog timer" in the placeholder. The device state shows "Online" or "Configuration accepted". I have discinnected two of the three analyzers and curiously Diagnostics shows "GOOD" for communications to the disconnected devices. I noticed diagnostics on the ProPlus might show "GOOD" while DIagnostics opened on a workstation would have persistent yellow "?". Port stats show some curious values which - unlike the "realtime card statistics" - rarely change or increment . . .
Here is an example of the "device information" for one of the analyzers that's no longer connected . . .
Best Regards,
John
Any thoughts on shield grounding? Sounds important in the literature. Should the shield terminal on the Profibus DP card be jumpered to the DCS high-quality ground? Should the shields be landed / grounded at the panel which is over 100 meters away and may not be an "equipotential" grounding point?
The instruments are powered out of our UPS in the house - the same one that supplied the DCS power subsystem (12 & 24 VDC) - so maybe there's a chance the grounds are "equipotential"?
In reply to John Rezabek:
The Siemens Oxymat is not terribly complex or esoteric, but does have two ambiguously labled DB9 connectors.
This problem was the result of connecting to the wrong one . . . communications are good since switching to the correct DB9 connector.
Curious though that DeltaV Diagnostics detected connected devices from time to time, even though we were not connected to the DP jack.