Post Go back to editing

JESD204B Rx link up, ILAS correct, but framer sends constant zeros when noise is expected; DecPower shows signal

Category: Software
Product Number: ADRV9026
Software Version: 7.0.0.14

Custom board, Zynq UltraScale+ (Xilinx JESD204 v7.2 + PHY v4.0), LMK04828 clocking, device clock 245.76 MHz. Use case: 4 Rx / 2 Tx, Rx framer 0 with L=4, M=8, F=4, K=32, N′=16, Subclass 1, lane rate 9.8304 Gbps. All files (profile, stream, ARM firmware, gain tables, init.c) generated by TES from the same profile. API version: [7.0.0.14], firmware: [7.0.0.11], TES: [7.0.0.15].

Status:

  • MCS status 0x17, PllStatusGet 0xF, init cals complete with no errors
  • JESD link up, received ILAS configuration matches (DID=1, BID=0, L=4, M=8, F=4, K=32)
  • Scrambling disabled for debug: lanes show the /F/ and /A/ replacement pattern of constant zero data
  • adi_adrv9025_RxTxEnableGet (SPI mode): readback Rx 0xFF, Tx 0x00 (didnt enable tx on purpose to solve rx first); Rx1 decimated power ≈ −59 dBFS when enabled, 0 when disabled
  • Framer 0 crossbar: Rx1 I/Q … Rx4 I/Q; framer 1 and 2 unused
  • FramerTestDataSet with PATTERN_REPEAT at SERIALIZER works; however, RAMP at FRAMERINPUT (all framers) has no effect (gives constant zero data)
  • RF ports were not terminated during init cals. 

My Init sequence: adrv9025_HwReset → PreMcsInit_v2 → PreMcsInit_NonBroadcast → MCS → PostMcsInit → framer link / SYSREF steps → RxTxEnableSet

Question: what could prevent Rx datapath samples (and the framer-input test ramp) from reaching framer 0, when RxDecPowerGet shows signal power?

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.