Hi everyone,
I’m working with the AD9467-FMC-250EBZ board and need clarification on whether my modified input topology is electrically correct or fundamentally incompatible with the AD9467 analog front end.
Goal
I want to feed the ADC with a DC-coupled signal from an external amplifier instead of the default AC-coupled transformer network.
What I Changed
-
Removed the transformer-based input section (T100/T101/T102/T104).
-
Replaced C109 and C110 with 0-Ω jumpers to allow DC coupling into VIN+ / VIN−.
-
Reinstalled the VCM bias network (R111, R112, C107) so VIN− can be held at the ADC’s ~2.15 V common-mode if needed.
-
Using ADA4945-1 as the external driver.
Two modes I tested
1. Differential Drive (OUT+ → VIN+, OUT− → VIN−)
This SHOULD be supported, but I am checking whether DC coupling is valid with the jumpers installed.
2. Single-Ended Drive Using VCM (OUT+ → VIN+, VIN− biased to VCM)
This is the topology I am uncertain about.
Problem Observed
In the single-ended mode:
-
The ADC output collapses to a flat line
-
Or produces unstable spikes/random vertical noise
-
As if the differential input stage cannot establish a valid common-mode reference
This happens even when R111/R112/C107 are installed correctly.
My Question (Main Point)
After removing the transformer, is it electrically supported to operate the AD9467 in a single-ended + VCM bias configuration where:
-
VIN+ receives the signal
-
VIN− is tied to VCM (via the onboard R111/R112/C107 network)
-
And only OUT+ from the ADA4945 is used?
Or…
Does the AD9467 absolutely require a true differential drive at both VIN+ and VIN− when operated in DC-coupled mode (transformer removed)?
Why I’m Asking
I am designing a custom signal-conditioning PCB using ADA4945-1.
Before I finalize the board, I need to know whether “Option C” (single-ended + VCM) is a valid input configuration, or if the ADC requires differential drive only.
Any confirmation, schematics, or official recommendation for DC-coupled operation would be extremely helpful.
Thanks in advance.
