Hello ADI Team,
I am currently working with the AD9467-FMC-250EBZ board connected to a ZedBoard for data acquisition experiments involving square-wave inputs from a function generator. The test frequencies range from 1 kHz to 1 MHz (1, 10, 25, 50, 100, 200, 500, and 1000 kHz).
1. Low-Frequency Distortion Issue
The output looks clean and stable for sampling frequencies between 500 kHz and 1 MHz. However, at lower frequencies (1 kHz – 100 kHz), the acquired waveform becomes heavily distorted.
From my research and forum discussions, I understand this distortion occurs due to the input transformer and AC-coupling capacitors that attenuate low-frequency and DC components. To address this, I performed a DC-coupling modification on the board:
-
Removed and bypassed the input path transformers T100, T101, T102, and T104.
-
Verified that T105 (XFMR optimized AD balun, 300 Ω to 100 Ω) was not populated on the board by default.
-
Ensured continuity from the input SMA connector to the ADC input pins to confirm a valid signal path.
The Analog Front end actual One is below:
Following these changes, I confirmed continuity from the input SMA connector through to the ADC input pins, ensuring that a valid signal path remains present.
Despite these changes, the board still fails to accurately capture square-wave signals at lower frequencies.
Could you please suggest additional schematic or configuration changes (e.g., biasing, common-mode adjustments, termination, or sampling settings) that would allow proper DC or low-frequency signal acquisition with the AD9467-FMC-250EBZ?
2. Post-Modification SPI/Initialization Issue (One Board)
After applying the same DC-coupling modification to four AD9467-FMC-250EBZ boards, three work perfectly and are fully programmable. However, one board is exhibiting unusual behavior during firmware operation.
-
When using the no-OS 2022_R2 branch with
make TINYIIOD=y, the board successfully builds and connects, but no UART print messages appear during initialization. -
The same board fails to connect to IIO Oscilloscope, whereas the other three identical modified boards connect and operate normally.
-
Upon debugging, I observed that the AD9467 ADC initializes successfully, but the AD9517 clock generator fails to respond to SPI communication. The PART_ID register always returns 0xFF, suggesting that the MISO line is floating and not returning valid data.
Given that the same firmware, SPI configuration, and hardware setup work fine with other boards, it appears to be a hardware-level issue localized to this particular unit.
Could you please advise on potential hardware causes, such as:
-
Faulty or unpowered CSB, reset, or MISO connection for the AD9517.
-
Power integrity or clock stability issues post-modification.
-
Possible board-level damage from transformer removal or re-soldering.
Any guidance on debugging steps or key measurement points (e.g., checking supply rails, SPI lines, or reset states) to isolate the failure would be greatly appreciated.
