Restart roadmap

[trabucayre]

• [x] Fix gateware
• [x] Fix driver
• [x] Fix software part
• [x] SoapySDR (fix, GNU Radio detection test) (6h)
• [x] DMA loopback with GNU Radio (2h)
• [x] SDR loopback with GNU Radio with low samplerate (6h)
• [x] RX by writing samples to file first, and with antenna/SMA (FM Radio) (8h)
  • [x] alignment IA/QA/IB/QB using tx_pattern (needs updating lms core for a CSR used to fill chan with 0 (only I or Q for channel A or B)
  • [x] stream content: no 0 and no samples lost or wrong aligned
  • [x] Two channels soapy -> gnuradio
  • [x] RFIC test with a real signal applied to each antenna (external RF generator)
• [ ] TX with antenna/SMA (HackSDR or PlutoSDR for RX, Carrier for TX) (1 and 2 channels) (3h)
  • [x] alignment check using litescope
  • [x] coherency between buffer received from gnuradio and RFIC
  • [ ] lms loopback samples sent must match samples received
  • [x] two channels gnuradio -> soapy -> gnuradio
  • [x] RFIC tests with an external RF receiver (HackRF and/or plutoSDR)
• [x] Complex/useful signal decoding (FM, GPS, ...)
• [ ] Complex signal/protocol generation (FM, ADSB, ...)

[enjoy-digital]

For the RX test with DMA/Soapy/Gnu Radio, I would suggest:

• Add JTAGBone to the design if not yet present.
• Add LiteScope on DMA's Sink.
• Use TX-Pattern -> MUX -> TX-RX-Loopback -> DEMUX -> CDC -> 32 to 64 -> DMA RX path (as can been seen in https://github.com/enjoy-digital/litex_xtrx/blob/master/gateware/lms7002m.py#L17), which involves:
  • Enabling TX Pattern: https://github.com/enjoy-digital/litex_xtrx/blob/master/gateware/lms7002m.py#L88
  • Enabling TX-RX-Loopback: https://github.com/enjoy-digital/litex_xtrx/blob/master/gateware/lms7002m.py#L196
  • Checking that TX Pattern is seen on LiteScope/DMA's Sink when software is running.
• When TX pattern is correctly seen with LiteScope, if not received correctly with Soapy/GnuRadio, the issue is related to software and can be investigated further.

[enjoy-digital]

For the TX test with DMA/Soapy/Gnu Radio, I would suggest:

• Add JTAGBone to the design if not yet present.
• Add LiteScope on DMA's Source.
• Play a know pattern from software and verify we are able to see it with LiteScope.
• When pattern is correctly seen with LiteScope, if issue on RF TX, the issue is related to LMS7002 init, datapath or RF frontend configuration, but the Software -> DMA part can be considered correct.

[enjoy-digital]

DMA Loopback with GNU Radio could be validated when RX and TX have been validated separately (at least on the digital part), otherwise potential issues will be too complicated to investigate and TX/RX issues will be mixed.

[enjoy-digital]

With DMA Loopback validated, RFIC Loopback could be tested.