10-18-2017 11:53 AM
Hey Everyone,
I am working with a cRIO-9024 chassis with a 9870 card also attached for additional serial ports, bringing my total available up to 5. Currently, 1 device is plugged into the 9870 card and communicates properly via serial/VISA with the controlling computer connected to the cRIO via ethernet.
We have a second instrument, a Keithley 6517B electrometer, that I would like to communicate via preferably the built in RS-232 port on the cRIO, however it refuses to accept commands when talked to through the cRIO, and this is a new problem that has occurred. About 4 months ago, we were able to talk to this instrument through the cRIO just fine, but now are unable to (There was a break in using this unit between then and now, where undetermined amounts of software changes on the cRIO happened). Serial communication does work because the first device mentioned earlier is communicating via VISA the entire time contact is attempted with the Keithley unit. Additionally, the unit displays "REM" when it receives serial commands, and does so accordingly when communication is attempted through the cRIO, so the keithley unit is at least recognizing it is receiving commands from the cRIO.
The Problem:
In NI MAX (COM Port):
Opening the VISA Test Panel, changing none of the port settings, while connecting the keithley instrument directly to a RS-232 port on a computer, the keithley responds immediately to the *IDN? command. This has worked with both VISA 15.0 and 17.0 on the computer (Upgraded to see if 17.0 may have been the problem)
In NI MAX (cRIO Ethernet):
On the same exact computer as used above, with VISA 17.0, the VISA Test Panel is opened on the corresponding serial port the keithley is connected to on the cRIO. Once again, no port settings are changed from the default settings in the test panel, and the cRIO port settings are the exact same from the hardwired COM port setup above. When the *IDN? command is sent in this manner, write is shown as okay, the "REM" icon is displayed on the keithley, but either the test panel times out during the read or receives responses from the keithley such as "\FF" "\FE\FC", etc.
For the cRIO, I have formatted the device multiple times, and installed CompactRIO 15, 16, and 17 on the unit with no change in. I have tried talking to the cRIO with other computers than the one used above, with no change in results.
Has anyone experienced something like this, or seem to agree that this may be a bug?
10-19-2017 04:06 PM
If you extend the timeout period of the read, do you always get a response back that is consistent? I guess in this case always "\FF" or "\FE\FC"?
Does the cRIO have the proper VISA drivers installed? If I remember right this is a separate driver from the CompactRIO drivers that needs to be installed to understand VISA commands properly.
10-19-2017 04:48 PM
Thanks for the response!
I varied the timeout between 500ms and 5000ms, and there does not seem to be a correlation between what is returned and the length of time. Anywhere from 0 to 8 bytes of "\FF", etc. come back.
Looking into the software installation on the cRIO, VISA 17.0 is installed, but I have also tried with VISA 15 as well.