This video documents the Open Research Institute's (ORI) FPGA Meetup from September 2, 2025, featuring project updates from team members. Key discussions revolve around three main projects. First, progress on porting a DVBS2 encoder to OpenCPI, which is currently blocked by a VHDL 2008 compatibility issue, with a proposed workaround involving a block design. Second, the team addresses persistent transmission timing gaps in the 'Opulent Voice' end-to-end demo, investigating whether the problem lies in the MSK modulator's HDL, data frame size, or system-level integration by developing a series of VHDL test benches. Third, updates from the Deep Space Exploration Society (DSES) Haswell facility detail hardware revisions for the Drift Scan project, specifically troubleshooting non-functional gain controls in GNU Radio blocks by manually adjusting the hardware signal chain. The meeting also includes a recommendation to use CocoTB for more advanced test bench development and concludes with a brief virtual tour of the DSES facility.
ORI FPGA Team Project Updates and Challenges
0:32 OpenCPI and VHDL 2008 Challenges: Aaron discusses his work on porting a DVBS2 encoder to OpenCPI. He identifies a major roadblock: the toolchain's incompatibility with VHDL 2008. His proposed solution is to implement the encoder within a block design as a temporary fix, with a long-term goal of updating OpenCPI itself.
4:13 Investigating Transmission Gaps in Opulent Voice: Paul details the ongoing effort to diagnose timing gaps and significant delays when streaming data to the MSK modulator. After ruling out the USB-to-Ethernet interface, the current hypothesis is that the software is sending incorrectly sized data frames (under half the expected data), leading to unexpected behavior. He is currently coding a fix to send the proper, fully encoded frames.
9:06 Developing Test Benches for the MSK Modem: To support the debugging effort, a series of VHDL test benches are being created to isolate the source of the timing issues. The goal is to either confirm the HDL is working correctly or pinpoint a specific module causing delays. The benches will incrementally test basic data streaming, properly framed data, and inter-module communication timing.
15:17 Hardware Revisions at DSES Haswell: Reporting live from the Haswell facility, Rick explains issues with the Drift Scan project. The GNU Radio software blocks do not allow for remote gain adjustment, forcing a manual hardware solution. He is physically reconfiguring the signal path with new splitters, cables, and attenuators to set a usable gain level.
20:53 Virtual Tour of the Haswell Facility: Rick provides a brief tour of the DSES control room, showing the science equipment racks, the future ham radio station, and the view of the large 60-foot radio dish outside.
31:22 Recommendation for Advanced Test Benches: Matthew suggests moving from pure VHDL test benches to CocoTB, a Python-based framework. This would provide a more powerful and flexible environment for simulating, driving data, and analyzing the modem's behavior directly from Python.