Hi all,
I am part of a research team working with radar waveform simulation using the USRP-2944 / X310 software defined radio. We connect the SDR to a host PC via PCIe, and we use LabView 2018 (x64) to write to the FPGA and read the data. Until recently, this has worked well with no issues; we built the code around the Simple Tx/Rx Streaming example project, so it was straightforward to get everything running.
However, at a time during which we made NO changes to the VIs or the LabView project, the connection stopped working for no discernible reason. Specifically, the LabView program hangs when trying to run the simulation. It does not fully "freeze" in the sense that Windows marks it as "Not Responding", but the pause and abort buttons cannot be interacted with, and LabView must be closed via task manager. I did some investigating using breakpoints and probes, and identified the culprit as a VI very early in the execution order that attempts to establish connection and load the bitfile onto the device (specifically getDeviceInfoFromScs.vi). This VI is not customized in any way, and is one of the VIs included in the original Simple Tx/Rx Streaming example.
Fast forward a week, and after digging through this roadblock with what feels knowledge-wise like a plastic spork, I guess my main goal in writing this post is to get some advice on where else to even look. I am relatively new to LabView and could use the guidance, as I am more of a software person still learning the ropes of hardware integration. Thank you for reading!
Troubleshooting attempts thus far:
- Ran the earliest possible version of the code - i.e. the original streaming example. The same issue occurs, with the execution hanging at around the same point in the diagram.
- Tried to run a few of the more basic VIs included in the "niusrpio_tools" library to confirm that, yes, any VI that tries to establish a link to the device has the same freezing issue.
- Ensured that the drivers for the USRP and RIO were up to date in the device manager - they were. I also updated the firmware and FPGA image using the USRP_utils.exe program. I could not update the firmware, but I wrote a new FPGA image from the same bitfile in case it got corrupted at some point. No change in behavior, unfortunately
- At this point, I wanted to test if it was a basic communication issue, so I used both the uhd_find_devices and uhd_usrp_probe utilities to see if they could find the device via PCIe, and both of them work! I also briefly ran one of the calibration programs included in the same utilities folder, and confirmed that not only does it work, but the Link LED turns on (unlike before).
- This indicates that the issue could be with LabView's connection protocol rather than the physical PCIe connection.
- We have another setup with a second USRP-2944R connected to an identical PC (albeit with older revisions of the LabView projects), and both devices work on this second PC's Tx/Rx Streaming example. Similarly, neither device works on the PC of interest. I considered simply moving the current project over to the second PC, but I am afraid of also copying whatever this error is.
Other potentially helpful clues I've stumbled upon:
- I tried to use the NI-MAX panel to help with the troubleshooting, but it would not come up. The window with "NI Measurement & Automation Explorer" "Initializing..." appears and then stays on the screen for over an hour, but never opens the program.
- The IT department here has been making some changes to the network over the past few months. Around the same time that this issue appeared, I suddenly stopped being able to remotely connect to this PC with my laptop. So, these could be related somehow.
- I tried changing the firewall settings and turned them off temporarily, but no dice. I also still cannot connect remotely, but that is a topic for a different forum.
