In this guide
How Oscillometric Blood Pressure Measurement Works: Cuff, Envelope, and the Two Numbers
Every blood pressure reading you have ever seen, two numbers separated by a slash, comes from one physical event: an artery being squeezed until blood can barely pass, then released slowly while a sensor watches the pressure fight back. The oscillometric method is how almost every automatic monitor, from a pharmacy arm cuff to a wrist wearable, reads those numbers without listening for heartbeat sounds. This guide explains the method from first principles, shows where the two numbers actually come from, and names the places where a student build honestly differs from a commercial device.
Context: why the cuff still matters in the wearable age
In October 2026, Huawei debuted its WATCH D3 wearable blood pressure monitor at the ISH-ECS hypertension conference in Dubai, a reminder that even the newest wrist devices still chase the same measurement the arm cuff has owned for a century. Wrist wearables miniaturise the sensor and hide the algorithm, which is convenient for daily tracking but useless for learning. The arm cuff remains the reference because its physics is open: a bladder squeezes the brachial artery, a pressure sensor reads the cuff, and the algorithm is a curve any student can draw and defend. That is why a student build starts with a cuff, not a watch.
How it works: four stages of one measurement
Stage 1 — Inflate past systolic. A pump fills the cuff to a target above the expected systolic pressure, typically around 180 mmHg in a student build. At this pressure the brachial artery is fully closed, no blood passes, and the cuff pressure trace is smooth: the sensor reads only the air pressure in the bladder.
Stage 2 — Deflate slowly and watch for pulses. A valve releases air at a controlled rate, about 3 mmHg per second. As cuff pressure falls below systolic pressure, blood starts forcing its way through the partially open artery in spurts, once per heartbeat. Each spurt pushes a small pressure pulse into the cuff, typically a fraction of a mmHg at first, growing as the artery opens further.
Stage 3 — Find the envelope peak. The firmware separates these small pulses from the slow falling cuff pressure, usually with a high-pass filter, and records each pulse amplitude against the cuff pressure at that moment. Plotted together, the amplitudes form a bell-shaped curve called the oscillometric envelope. The envelope peaks where the artery is most free to pulse, and that peak pressure is the mean arterial pressure (MAP), the average pressure driving blood through the body.
Stage 4 — Read the two numbers off the envelope. Systolic and diastolic pressure are not measured directly; they are read from the envelope by ratios. A teaching build uses 55 percent of peak amplitude on the rising side of the envelope for systolic pressure, and 85 percent of peak amplitude on the falling side for diastolic pressure. The pulse rate comes free: it is the spacing between the pulses themselves. A typical healthy reading of 120 over 80 mmHg therefore means a MAP near 93 mmHg, with the two ratio points sitting on either side of the peak.
Why ratios, and why they are approximate
The ratios work because of artery mechanics. Above systolic pressure the artery is shut, so no pulse appears. Just below it, a small spurt gets through, so pulses are small. At MAP the artery wall is unloaded and swings most freely, so pulses are largest. Below diastolic pressure the artery stays open all cycle, so the pulsing nearly vanishes. The exact percentages vary between manufacturers, and commercial monitors use proprietary, population-tuned algorithms rather than two fixed ratios. A student build that states its ratios openly will differ from a commercial monitor by several mmHg, and that difference is a result to report, not a failure to hide.
What the numbers mean
Systolic pressure is the peak force when the heart contracts and pushes blood out. Diastolic pressure is the resting force between beats when the heart refills. Mean arterial pressure is the time-weighted average that actually drives blood flow to organs. Pulse pressure, the difference between the two numbers, widens when arteries stiffen. None of these is measured by a pressure sensor directly; all of them are inferred from the envelope, which is why the quality of the envelope decides the quality of the reading.
Practical factors that move a reading
Cuff size. A cuff too small for the arm reads high, because extra pressure is needed to close the artery through more tissue; a cuff too large reads low. Record the cuff size with every reading.
Arm position. The arm must be supported at heart height. An arm hanging below the heart adds hydrostatic pressure, roughly 0.77 mmHg per centimetre of height difference, and shifts the reading.
Motion and talking. Muscle movement and speech shake the cuff pressure trace and distort the envelope. A good build flags such readings as bad instead of reporting a number.
Repeat readings. Blood pressure varies naturally between sittings. Taking three readings and comparing them is not indecision; it is the correct way to report a variable quantity.
The sensor arithmetic a student should check
A common student pressure sensor, the MPX5050, measures 0 to 50 kPa with an output of 0.2 to 4.7 V. Since 1 kPa equals 7.50062 mmHg, the sensor range covers 0 to about 375 mmHg, comfortably above the 180 mmHg inflation target. At 90 mV per kPa, one mmHg corresponds to about 12 mV, so even a basic 10-bit analogue input resolves well under 1 mmHg. The sensor is never the weak link; the envelope and the ratios are. That is a useful lesson: in instrumentation, the algorithm usually limits accuracy before the hardware does.
Frequently asked questions
Why not just use a stethoscope and listen? That is the auscultatory method: a clinician listens for Korotkoff sounds as the cuff deflates. It needs a trained ear and a quiet room. The oscillometric method replaces listening with signal processing, which is why it can run automatically on a microcontroller.
Can a wrist wearable match an arm cuff? Approximately, not exactly. The wrist artery is smaller and the sensor geometry differs, so wrist devices need their own validated algorithms and careful positioning. For learning and for reference readings, the upper-arm cuff remains the standard.
Why do two monitors disagree on the same person? Different envelope algorithms, different deflation rates, different cuff sizes, and natural variation between sittings. A disagreement of a few mmHg between devices is normal and must be reported as a range, not hidden.
Is a student-built monitor a medical device? No. It is an educational instrument for demonstrating a measurement principle. Its readings must never be used to diagnose, treat, or monitor any health condition, and any unusual reading must be checked with a proper medical device.
Limitations of this explanation
This guide describes the standard oscillometric model taught in instrumentation courses. Real monitors add artefact rejection, irregular-pulse handling, and proprietary envelope processing that this guide deliberately does not claim to reproduce. Arrhythmias distort the envelope in ways a simple ratio method cannot correct, and no fixed-ratio student build should be presented as clinically validated.
The project behind this guide
The companion student build puts this method on a bench: WATCH D3-Inspired Cuff-Based Blood Pressure Monitor inflates a real arm cuff with a pump and valve, samples pressure with an MPX5050 sensor on an ESP32, draws the envelope on a dashboard, and reports systolic, diastolic, and pulse rate from the same ratios explained above. Reading this guide first makes the project page read like a lab manual instead of a mystery.