HHZ Respiratory Review

How we test

Every HHZ review starts with the manufacturer's specification sheet and ends with bench measurements where we can run them. The protocol below is what we apply, in what order, with what tolerances. It is written so that a reader with access to the same equipment could repeat our measurements and expect to land inside our stated uncertainty bands.

Oxygen concentrator protocol

Oxygen purity across the working range

Purity is sampled at the outlet after a 20-minute warm-up with the device in its declared operating orientation. We record the flow selector position, the measured flow (via in-line mass-flow reference), and the oxygen fraction by volume. Sample points:

Each setting is held for 10 minutes before logging a one-minute average. We publish the averaged value and flag any setting at which purity dropped below 82% — the lower bound of the ISO 80601-2-69 acceptable range for home concentrators.

Sound level at patient position

Measured on the A-weighted scale with a Class 2 sound level meter at one metre from the device's nearest surface, at ear height for a seated user, in a quiet room (ambient <32 dB(A)). We follow the geometry specified in IEC 60601-1-8 for alarm and operating sound, and we report the slow-averaged steady-state value plus the peak observed during the one-minute window. Rated accuracy of the meter is ±1.5 dB; we publish a ±2 dB uncertainty band.

Continuous eight-hour power draw

The device is run at its rated maximum flow for eight hours on a stable 230 V ±2% bench supply, with a power meter logging true RMS current, voltage, apparent power (VA), real power (W), and power factor at ten-second intervals. We report the eight-hour mean and the 95th-percentile draw, and we flag any thermal-cycling pattern that suggests a compressor overload protector cutting in.

Sieve endurance cycling

Where a unit is available long enough, we run an accelerated cycle: eight hours at rated flow, one-hour cool-down, eight more hours at rated flow. A purity re-measurement after each 100 cycles flags sieve degradation. Most concentrators in this category are rated for 10,000–20,000 hours of sieve life; accelerated cycling reveals units whose sieves start dropping below 85% purity earlier than the spec suggests.

Altitude performance

Where we can, we re-run the purity curve in a low-pressure chamber or at a verified high-altitude location (typically 2,000–2,500 m equivalent) to quantify the drop in delivered purity. For most 5 LPM concentrators, delivered purity at 5 LPM falls by 3–7 percentage points between sea level and 2,500 m. We publish the device-specific delta when we have it.

Alarm trigger verification

We verify that each manufacturer-declared alarm actually fires and within its declared threshold: low-purity alarm (typically <82%), high-temperature alarm, power-failure alarm, and flow-obstruction alarm. We record the time from fault introduction to alarm annunciation.

CPAP and BiPAP protocol

Pressure accuracy

We measure delivered pressure at the mask connector using a calibrated digital manometer (±0.1 cmH₂O resolution) with the device driving a reference breathing simulator. We sample CPAP at commanded 5, 8, 12, and 16 cmH₂O and BiPAP at IPAP/EPAP pairs of 10/5, 14/8, 18/10, and 20/12 cmH₂O. Each setting is held for one minute after a 30-second stabilisation. We report mean delivered pressure and peak-to-peak variability.

Leak compensation

With a controlled mask-leak orifice (calibrated to 20, 40, and 60 L/min at 10 cmH₂O), we measure the device's ability to hold commanded pressure against the leak. The published value is the mean pressure deviation from setpoint across the stabilised minute.

AHI detection accuracy

Using a QuickLung-style breathing simulator programmed with scripted apnea and hypopnea events, we compare the AHI reported by the device to the reference event count over a one-hour simulated night. This catches algorithm-level optimism — a chronic issue with entry-level APAPs that undercount hypopneas compared to a polysomnography reference.

Noise at patient-ear position

Measured at 30 cm from the mask (the approximate distance of a pillow-side sleeper's ear) with the device at a typical 8 cmH₂O therapy setting and a representative mask attached. Same meter and uncertainty band as the concentrator protocol.

Destructive teardown

After performance testing completes, every loaner unit goes through a destructive teardown designed to surface manufacturing and materials choices that do not appear on any datasheet. Manufacturers are told in writing before the unit ships that it will not be returned in working condition. Teardown findings are published as a dedicated section of the review, and they feed directly into the build-quality score.

Sieve-bed zeolite analysis (concentrators)

Each molecular-sieve bed is extracted, weighed, and sampled. We run FTIR identification on the zeolite to confirm the grade (13X vs LiX and binder type), measure bed density with a calibrated tamped-volume cylinder, and record the total zeolite charge against the manufacturer's declared value. Bed weight and zeolite grade correlate directly with sieve life expectancy; a nominally-10,000-hour sieve with a 10–15% below-spec charge or the wrong binder will fail well short of its warranty window. We also run a moisture-stress sample — the sieve is exposed to 60% relative humidity for 24 hours and re-weighed to quantify moisture uptake, which is the dominant real-world degradation mode in coastal Indian installations.

Solenoid-valve teardown (concentrators)

PSA concentrators hinge on two solenoid valves cycling 10,000+ times a day. Each valve is extracted and inspected for seat material (Viton, EPDM, NBR — each has a different cycle- life curve), coil insulation class, spring-return preload, and seal durometer. Where a second identical unit is available, we run one valve to its rated cycle count on a bench jig and tear it down to inspect seat wear.

Enclosure and internal plastics — FTIR material identification

Every polymer panel — outer chassis, air-path tubing, humidifier chamber — goes through Fourier-transform infrared spectroscopy for material confirmation. Manufacturers often declare "ABS" when the actual part is a PP blend, or "polycarbonate" when it is a PC-ABS blend with reduced impact resistance. We publish the identified grade against the declared grade. Wall thickness is measured with a coating-thickness gauge at four points per panel; thin-walled PP enclosures flex under daily abuse and are a leading driver of out-of-warranty crack failures in Indian-summer installations.

Oxygen purity sensor validation (concentrators)

The purity sensor is identified (galvanic electrochemical vs ultrasonic vs paramagnetic), its response time measured, and its calibration checked against a reference gas mix (certified 93% O₂ balance N₂). Drift from the factory calibration is recorded; units whose sensor is already reading 3+ percentage points off after a few hundred hours of operation are flagged. We also inspect the sensor's electrical interface — solder joint quality, strain relief on the cable, connector-pin plating grade — because most purity- alarm nuisance trips trace to the interface, not the sensor element itself.

Voltage stress test

The unit is driven by a programmable AC source through the edges of realistic Indian mains tolerance: 160 V (deep brownout, typical of Tier-3 city evening loads), 270 V (surge, typical of unregulated rural three-phase unbalance), and 30 minutes of square-wave inverter output (the single most common cause of compressor motor burnout in Indian home deployments). Post-test inspection flags compressor winding discoloration, capacitor bulging, PCB trace carbonisation, and any borderline voltage-regulation components that survived the stress but show thermal stress markers.

Compressor and blower teardown

For concentrators, the compressor is removed, opened, and inspected: piston or diaphragm condition, crankshaft bearing grade (needle-roller vs sleeve vs ball), cylinder-wall scoring, seal material, and the presence or absence of a thermal-fuse safety on the motor winding. The compressor is the single longest-lifespan-determining component on a concentrator, and its teardown tells us whether the published duty cycle is plausible. For CPAP/BiPAP, the blower turbine is extracted: bearing grade, magnet material (ferrite vs rare-earth), impeller balance, and counterweight quality. Cheap turbines develop audible bearing wear by 2,000–3,000 hours.

Intake and HEPA filter verification

Filters are weighed before and after a controlled dust-exposure run using ISO 12103-1 A2 Fine test dust at a fixed flow rate, plotting pressure drop against accumulated loading. A filter labelled "HEPA" that fails H13 minimum 99.95% filtration at 0.3 μm is flagged as mislabelled. Intake-filter gsm weight and media-area are recorded against the declared service interval.

Heat-stress test to thermal cutoff

The unit runs at maximum flow inside a 40 °C ambient enclosure (representative of an un-air-conditioned Indian bedroom in May–June) until thermal protection activates. We record the run-time to cutoff, the internal temperature profile at three points (compressor head, power-supply board, zeolite bed outlet), and — post-cutoff — whether the unit resumes normally after cooldown or surfaces any permanent damage. Any unit that cannot complete a 4-hour run at 40 °C fails the Indian-summer use case and is flagged.

Cooling-fan bearing life

The enclosure fans run continuously at rated operating temperature until audible bearing wear or mechanical failure. Cheap sleeve-bearing fans typically fail at 3,000–5,000 hours; dual ball-bearing fans rated 30,000+ hours are the premium-tier signal. A concentrator rated for 10,000-hour duty with a 5,000-hour fan will fail its cooling before it fails its compressor.

PCB and power-supply teardown

The main control board and power supply are removed and inspected: electrolytic capacitor temperature rating (85 °C vs 105 °C grade), voltage-regulation circuit quality, heatsink coverage on power components, input-side surge protection presence, reverse-polarity and thermal-cutoff protection, and trace-width adequacy for the rated current. A concentrator's PCB is its second-most-common failure surface after the compressor, and the teardown usually predicts failure modes before they surface in field reports.

CPAP/BiPAP-specific teardowns

In addition to the shared enclosure, PCB, and voltage-stress protocol, CPAP and BiPAP units undergo: heater-plate teardown (thermal-fuse rating, aluminium grade, thermistor placement), pressure-transducer calibration-drift test against a reference manometer over a 30-day continuous run, flow-sensor type identification (hot-wire anemometer vs differential-pressure orifice vs ultrasonic time-of-flight), humidifier-chamber material and silicone-gasket durometer, and mask-port seal fatigue over 500 connect/disconnect cycles. Where the machine has wireless connectivity, we also audit the cellular or Wi-Fi module for data-handling behaviour — specifically whether patient-identifiable data leaves the device without a documented consent flow.

How the teardown changes the review

Findings feed the build-quality score directly. Spec-sheet claims that collapse under teardown — undersized zeolite charge, sleeve-bearing cooling fans, below-rated capacitors, mislabelled HEPA filters, compressor motor without thermal-fuse protection — are flagged in the review's cons and explicitly called out in the final verdict. A unit that tears down clean earns the build-quality points it claims on paper; a unit whose teardown contradicts its datasheet loses them, and the gap is named.

Test equipment

Performance bench:

Teardown bench:

Measurement uncertainty

Every numeric claim on the site is published with its measurement uncertainty:

When the difference between two devices on a measured metric is smaller than the combined uncertainty, we call it a tie. We do not award fractional scoring that the measurement cannot support.

When bench data is not available

We do not always have access to every device. For new releases not yet in distribution, or for models where a loaner is not forthcoming, we say so explicitly at the top of the review and score only against the paper-spec rubric (features, warranty, service network, price-to-spec ratio, manufacturer reputation). We mark such reviews with a "paper-spec review — no bench data" notice, and we revisit the review when hands-on access becomes available.

Rubric and scoring weights

Each category uses a weighted scoring model that we publish alongside this methodology. For oxygen concentrators the weights are: purity at rated flow (25%), sound (15%), build and service network (20%), warranty terms (15%), price-to-performance (25%). For CPAP/BiPAP: pressure accuracy (25%), leak compensation (15%), noise (15%), mask and humidifier ecosystem (15%), warranty and service (15%), price-to-performance (15%). The underlying score-to-stars mapping is identical across brands.