I am running Fieldpoint 4.0.1 along with GE Intellution FIX32 successfully on a Windows 2000 computer. The FIX32 software communicates with Fieldpoint through OPC. The Fieldpoint I/O is done via a serial port. The system works great.
I am attempting to transfer this same software to a backup (newer) Windows XP computer. I have loaded both programs and am having problems with the OPC data exchange.
Upon a fresh boot, Fieldpoint Measurement & Automation (NIMax) runs, sees the fieldpoint hardware, and allows me to monitor I/O data in realtime. I can observe the serial-data line activity via the flashing LED's on the Fieldpoint controllers. I can also use Server Explorer to connect to the OPC Server.
However, when I start the FIX32 software - which presumably attempts to start the OPC Client - the client acknowledges a connection, but no data is transferred. In fact the OPC client appears to freeze. Also of interest, the Measurement & Automation software now reports all "0.0000" for the I/O data when attempting to view the real-time I/O. The Fieldpoint serial-line LED's are no longer flashing.
This condition will continue until I reboot the computer. I don't need to reboot the fieldpoint hardware.
Can you give me some tips to help debug?
I'm going to be looking into this situation for you, but before I do i need to know a few more things about your system.
What FieldPoint controller are you using?
It sounds like you are using the real-time OS to run your code, is that correct?
If so, is the real-time application dependent upon information from an OPC server?
I also need you to look into whether or not your FIX32 client is set up properly with Windows XP. Here is a link that discusses proper set up with WinXP.
Many thanks for your message. I was unaware of the DCOM security issues with XP. After following the instructions in the OPC foundation link you provided, the system seems to be working. I will give the system a more thorough workout tomorrow and revert if I have additional issues.
My system is a single-computer running Windows XP. The Fieldpoint hardware is a FP-1000 RS-232/RS-485 network interface (with a second FP-1001 interface attached). All software (FIX32, etc.) is running on the same local computer - theoretically no TCP/IP network access is needed.
Further testing reveals some ongoing problems.
The Fieldpoint hardware consists mostly of thermocouple modules and a few analog input/output and digital input/output modules. The digital I/O channels are used to turn various production equipment on/off and monitor the condition of various on/off relays.
When I startup the FIX32 software, it appears to connect to the NI OPC server fine and begins communicating with the NI Fieldpoint hardware via the serial COM port. The systems reads ("polls") the Fieldpoint hardware a few times a second and displays the status of all channels. In particular, the thermocouple readings display as desired. I have left the system running overnight and it seems to run in this mode without problems.
However, if I attempt to turn production equipment on/off using the digital IO channels, an intermitent problem develops. The symptoms are the FIX32 software stops communicating with the Fieldpoint hardware (as evidenced by the serial LED's on the Fieldpoint controller). I suspect perhaps the OPC server has "stopped working". In this failure mode, if I stop FIX32 and start the NI Measurement and Automation program, the system is unable to "read" the status of any of the Fieldpoints channels. If I reboot the computer the system starts working fine again.
I have reproduced the failure mode 3 times. In each case, the problem occured within 10 minutes of starting to turn on various equipment (via the digital IO channels). In all cases, I had managed to successfully turn on at least 3 DO channels before the system failed.
One more observation - after the system failure I used the task-manager to examine the services. The FieldpointOPC2 server executable was still listed. I attempted to stop it, (with the intent to restart it and see if FIX32 or NIMax would resume working), but I couldn't seem to stop the service - Windows would accept the "stop" command, but the service would remain listed in the task manager.
- Can you give me any more ideas on how to diagnose what part of this complex software system is failing?
- Is there a procedure that I can stop/restart the NI OPC Server to investigate this?
- Any other ideas?
Do you have diodes as protection across the coils of your relays? I believe I remember reading in some older threads where someone wasn't using them, and the voltage spikes that are generated when a relay coil denergizes was causing corruption in the communications of their fieldpoint modules to the controller.
Many thanks for you Raven Fan. Noise on the RS-232 line was the source of my last set of problems. I was able to solve with a RS-232 optical isolator.
My system is now working as expected. Thanks to the forum for all your help.