Point-of-Care Test (POCT) Electronics: AFE + MCU + Disposable I/F
← Back to: Medical Electronics
POCT electronics is the discipline of turning a disposable sensor into a trustworthy result within minutes, using a compact reader that tightly controls the cartridge interface, measurement timing, and low-power states. A practical POCT design succeeds by gating bad insertions early, scheduling clean sampling windows, and linking calibration + error budget to a traceable result record.
Definition: What makes a device “POCT electronics”
POCT electronics is the reader-side hardware and firmware that connects to a disposable strip or cartridge, applies controlled excitation and timing, converts sensor signals with an analog front end (AFE), runs low-power orchestration on an MCU, and delivers a traceable result through a simple display and short-range I/O.
POCT designs are constrained by small size, short test time, single-use consumables, low maintenance, low power, and result traceability. These constraints push the architecture toward a strong disposable interface + timing orchestration + low-cost error budget, rather than heavy instruments or long, user-managed calibration routines.
- Small form factor: fewer protection and shielding options; interface cleanliness and ESD become first-order design items.
- Short test time: precision comes from controlled timing windows, not from “averaging longer”.
- Disposable consumable: cartridge ID and calibration payload must be part of the signal path and data flow.
- Low maintenance: built-in self-checks and contact diagnostics reduce user actions and support calls.
- Low power: burst measurement, peripheral gating, and event-driven firmware dominate the power budget.
- Traceable results: records typically include cartridge lot/ID, calibration version, timestamp, and measurement status flags.
Boundary note: this POCT page focuses on reader ↔ disposable interface, AFE timing, and low-power orchestration. It does not expand into PCR thermocycling, high-throughput lab analyzers, or gateway/security system architecture.
System architectures you will actually see
A practical way to avoid scattered requirements is to classify POCT designs by how the disposable connects and what sensing modality dominates. Most products fall into a small set of repeatable patterns.
Common reusable core across types: AFE + ADC + ULP MCU orchestration + a simple display, with only light short-range I/O. Differences concentrate at the disposable interface and the measurement timing windows.
Disposable-sensor interface: the make-or-break layer
POCT reliability is often decided before any measurement begins. The interface must confirm correct insertion, healthy contacts, and valid cartridge identification/calibration, while protecting the AFE from ESD and preventing moisture or contamination from creating leakage paths that mimic real signals.
Practical rule: gate the measurement. Only enable AFE sampling after presence/latch, contact health, ID/cal validation, and leakage checks pass within defined windows.
- Presence detect / latch: insertion detection, latch confirmation, and tolerance-aware “ready” criteria.
- Cartridge ID & calibration payload: resistor code, 1-Wire, I²C EEPROM, or NFC; used to load gain/offset/time constants and traceability fields.
- ESD / moisture / contamination: clamp and limit surge paths; treat wet contacts and residue as leakage sources that can create false currents/offsets.
- Anti-reuse (light): one-time token, counter bit, fuse mark, or ID bind; kept simple to deter accidental reuse without building a full security system.
AFE patterns for POCT: one page, two universes
Most POCT readers reuse a common MCU/display/power core. The AFE, however, usually falls into one of two repeatable patterns: electrochemical sensing (current/voltage at electrodes) or optical sensing (LED + photodiode). The right pattern is selected by the disposable’s physics, then tuned by a cost-driven error budget.
Electrochemical POCT (module level)
- Excitation: voltage step, programmable current source, or DAC-driven bias to create repeatable chemistry conditions.
- Measurement: TIA and/or electrode potential sense into a low-noise ADC with a stable reference and input protection.
- Multi-electrode / multi-channel: MUX plus bias routing, with open/short diagnostics to catch contact faults.
- Key metrics: input bias/leakage, noise floor, polarization settling impact, and end-to-end dynamic range.
Optical POCT (module level)
- LED driver: pulsed or constant-current, sometimes multi-wavelength; pulse timing enables low average power and synchronous sampling.
- PD TIA: noise and bandwidth set minimum detectable signal; saturation recovery sets how soon a valid sample window can open.
- Timing: blanking after switching events, ambient subtraction (LED off vs on), and synchronized sampling markers.
Low-power MCU orchestration: timing is the algorithm’s chassis
In a POCT reader, reliable results come from a repeatable state machine that gates measurement, schedules short sampling bursts, and returns the system to low power immediately after each window. Timing control is the chassis that holds sensing physics, AFE behavior, and user interaction together.
Practical rule: Gate → Burst → Sleep. Gate measurement on interface checks, sample in short bursts, then power-gate AFE and I/O back to sleep.
- Idle: deep sleep; only insertion/presence path active.
- Insert: presence/latch, contact health, and ID/Cal read/validation.
- Warm-up / Prime: brief stabilization of references, bias, LED/TIA; quick baseline checks.
- Measure: burst sampling windows; gate high-power peripherals between bursts.
- Validate: quality checks; classify contact/leakage/instability as actionable faults.
- Display / Transmit: show result; optional short-range export; record traceable fields.
- Time-sliced power: enable AFE/LED/reference only inside defined windows.
- Peripheral gating: ADC, bias, LED driver, pull-ups, and UI backlight are state-scoped.
- Burst sampling: short bursts reduce average power and isolate transients from steady samples.
- Event-driven firmware: insertion, button, timer, and threshold events replace polling loops.
- Self-test: baseline/offset sanity, ADC saturation checks, and reference-in-window tests.
- Reference drift check: compare prime vs measure windows; retry prime if drift exceeds limits.
- Contact anomaly checks: open/short/leakage detection and stability scoring across bursts.
Display & comms (keep it minimal, POCT-specific)
POCT output is optimized for fast readability, low average power, and minimal interference with sensitive measurements. Local UI and short-range interfaces should stay simple: show a clear result, export only what is needed, and always attach traceable fields.
- Segment / simple LCD: lowest power, excellent sunlight readability, best for numeric + status UI.
- Small mono LCD: adds icons and step prompts with modest power increase and low EMI risk.
- Small TFT: richer guidance and error details, but backlight power and EMI management become design constraints.
Accuracy & calibration: error budget you can actually manage
POCT accuracy becomes manageable when error sources are grouped into a small set of buckets, then matched to a calibration placement strategy and a lightweight field verification plan. The goal is a repeatable error budget that can be verified during insertion and measurement, without turning the system into a maintenance-heavy instrument.
- Disposable variance: lot-to-lot and cartridge-to-cartridge gain/offset differences and chemistry/optics variability.
- Temperature: reaction rate, LED/PD temperature behavior, leakage changes, and reference drift sensitivity.
- Optical / electrochemical drift: LED aging/contamination, electrode polarization and fouling, response settling shifts.
- ADC / reference: gain/offset/noise floor, reference stability and drift within the measurement window.
- Mechanical positioning: insertion alignment, optical path repeatability, contact resistance variability.
- Control line: a must-pass indicator that the disposable chain is functioning.
- Internal standard: a quick reference reading to spot optical drift or path changes.
- Sanity checks: baseline, saturation, noise, and repeatability gates that stop invalid runs early.
IC role mapping (example part numbers for POCT)
This role map keeps POCT-specific priorities in view: low average power, repeatable timing orchestration, and an error budget that can be linked to cartridge ID/cal data and traceable result records. Part numbers below are examples to represent each role; selection depends on sensor physics, cost targets, and interface constraints.
- ADI ADuCM355 — integrated electrochemical measurement + MCU class integration; supports “read ID/cal → configure → measure → record” loops.
- ADI AD5940 / AD5941 — low-power electrochemical/impedance-oriented AFE family; useful for repeatable excitation + measurement blocks.
- TI LMP91000 — configurable analog front end for electrochemical sensing; helps adapt a reader to multiple disposable behaviors with tunable settings.
- TI AFE4404 — integrated timing-friendly optical measurement front end; useful as a reference pattern for synchronized sampling windows.
- TI LMP7721 — ultra-low input bias amplifier; helps control bias/leakage error terms in very small current measurements.
- ADI ADA4530-1 — ultra-low bias with guard buffer concept; useful where leakage paths dominate the error budget.
- TI OPA381 — photodiode TIA-friendly op amp option; supports fast settling for pulsed LED sampling windows.
- TI ADS1220 — low-power 24-bit ΔΣ ADC with PGA/IDAC options; useful for small-signal measurement and repeatable gain paths.
- ADI AD7124-4 / AD7124-8 — multi-channel precision ΔΣ ADC family; supports stable, low-drift measurement paths where channel switching is needed.
- ADI ADR4525 — precision voltage reference example; stabilizes the ADC/reference budget term across temperature and time.
- ADI AD5683R — low-power DAC with internal reference option; useful for programmable excitation/bias in a compact BOM.
- Nordic nRF52840 — BLE SoC example for short, user-triggered exports; supports deep sleep and event-driven bursts.
- Nordic nRF52832 — smaller BLE SoC option when feature set is minimal and power/size are prioritized.
- ST STM32L4 — ULP MCU family example for state-machine orchestration and aggressive peripheral gating.
- TI MSP430FR series — ULP MCU example suited to event-driven control and long standby with fast wake.
- NXP PCF8576C — segment/LCD driver example for ultra-low-power numeric + status UI.
- Holtek HT1621 — common low-power segment LCD driver option for simple POCT displays.
- TI TLC5947 — multi-channel constant-current LED driver for stable illumination and indicator LEDs.
- TI TLC59711 — higher precision LED driver option when current matching and repeatability matter.
- TI BQ25120A — highly integrated charger/power-path style solution for small batteries and low-power systems.
- ADI/Maxim MAX17048 — low-power fuel gauge option to support battery-aware policies and traceable logs.
- TI TPS62840 — low-IQ buck regulator example for efficient “burst then sleep” power behavior.
- TI TPS7A02 — ultra-low-IQ LDO example for quiet always-on rails with minimal standby drain.
