Ultrasound Receiver Overload: Verify Recovery Beyond the Last Clipped Sample

By ALEVA | Orange County, California, USA

Engineering question: After a strong receive-path disturbance, when can the next weak echo be treated as valid data?

An ADC output that no longer reaches full scale is an incomplete recovery criterion. Analog stages can overload before conversion. Maxim’s ultrasound design note, now hosted by Analog Devices, explains how saturation and recovery in the receiver affect subsequent signals. The absence of clipped digital codes therefore does not, by itself, demonstrate that the complete analog path has returned to its intended response. [1]

Recovery specifications also have conditions. The historical TI AFE5804 Rev. C datasheet specifies a two-clock-cycle recovery entry with a stated overload level and error criterion. Its application discussion separately addresses transmit feedthrough, strong echoes and external protection circuitry. That component specification should not be copied into a complete probe-and-board acceptance requirement without reproducing the relevant conditions. [2]

A proposed paired-capture experiment

The following is a proposed engineering protocol, not measured ALEVA performance. First capture a repeatable weak electrical test burst without a preceding disturbance. Preserve the input waveform, termination, gain schedule, bandwidth and sampling settings. This is the reference for the same receiver configuration.

Next precede that burst with a controlled strong disturbance, within the approved limits of the test fixture and device. Sweep the interval between disturbance and weak burst. Compare the recovered burst with the reference using amplitude error, residual offset and waveform error; include phase when the intended processing depends on coherence. Repeat captures to distinguish a stable bias from random variation.

Define acceptance tolerances and the required valid observation interval before examining results. Record both the last clipped sample and the first interval that meets the complete acceptance criteria. They answer different questions. A quiet interval with no weak test burst cannot establish that small-signal sensitivity has recovered.

Trade-offs and boundaries

Repeat the experiment across relevant gain schedules, disturbance polarities, durations and channels. A single large-input test at one gain setting is a narrow qualification. TI’s discussion of protection and overload behavior also makes the external receive path relevant: changing the protection network can change recovery and noise behavior. Document the probe interface and protection configuration alongside the result. [2]

A longer invalid-data interval may simplify downstream handling but excludes more early receive data. A shorter interval requires stronger evidence that the selected error limits are met. For coherent processing, specify how an invalid channel or sample interval is represented through IQ processing and beamforming; silently treating it as ordinary valid data obscures the test outcome. This is a system-design recommendation, not a claim about a particular implementation.

The OEM deliverable

Request a recovery envelope tied to configurations, stimulus files, raw captures and acceptance rules. Separate electrical recovery evidence from acoustic phantom evaluation and clinical claims. The useful handover is a reproducible validity decision, rather than one recovery-time number detached from its test conditions.

Sources and scope

[1] Maxim Integrated, Don’t Be Blinded by Your Amplifier’s Slow Overload Recovery, Design Solutions note hosted by Analog Devices; publication date not printed in the reviewed text. Historical manufacturer reference, checked November 7, 2026.

[2] Texas Instruments, AFE5804 datasheet SBOS442C, Rev. C, June 2008, revised October 2011; electrical characteristics and overload-recovery discussion. These references are not new-product announcements, and no commercial use of either manufacturer’s device by ALEVA is asserted.