Designing a single-lead ECG patch: electrodes, spacing and signal quality
A wearable ECG is only as good as the few millimetres between the electrode and the skin. These are the decisions that matter most when you design a single-lead patch.
A hospital ECG uses wet electrodes spread across the chest and limbs. A patch has to record a usable signal from two electrodes a few centimetres apart, on a person who walks, sleeps and sweats. That changes almost every design decision.
1. Electrode spacing sets the signal size
The voltage a patch picks up grows with the distance between its two electrodes. Bring them closer and the QRS complex shrinks, while noise from the skin and muscles stays about the same. Most chest patches use a spacing of roughly 4 to 10 cm. Below that, the signal often becomes too small to read P waves reliably.
Orientation matters as much as distance. A patch that runs along the heart’s main electrical axis, from upper right to lower left of the sternum area, sees a larger signal than one placed across it. On Hridya each lobe of the heart shape carries one electrode, which gives a usable spacing in a small, friendly outline.
2. Skin contact is the biggest source of noise
- Gel electrodes (silver / silver chloride with hydrogel) give the lowest and most stable skin impedance. They are the safe default for a first device.
- Dry electrodes are reusable but have much higher impedance, especially in the first minutes and on dry skin. They need a front end with very high input impedance.
- Movement stretches the skin under the electrode and changes the half-cell potential. This shows up as baseline wander and spikes that can look like heartbeats.
A good adhesive that holds the electrode still does more for signal quality than any filter in firmware.
3. Choose an analog front end built for ECG
Dedicated ECG front ends combine a low-noise amplifier, filters, an ADC and lead-off detection in one chip, at a power draw small enough for a patch. Look for:
- input-referred noise well below the smallest feature you want to see (P waves can be under 0.1 mV);
- high common-mode rejection, so mains hum from nearby wiring cancels out;
- DC lead-off detection, so the app can say “electrode not in contact” instead of drawing noise;
- a sample rate of at least 125 to 250 samples per second for rhythm monitoring. Hridya records at 200 samples per second.
4. Filter for the use, not for looks
Strong filtering makes a trace look clean on a demo, but it can also distort the ST segment and hide real features. The common conventions are:
| Filter | Typical setting | Why |
|---|---|---|
| High-pass | 0.5 Hz for monitoring, 0.05 Hz for diagnostic | Removes breathing and movement drift. Higher cut-offs can distort the ST segment. |
| Notch | 50 Hz (India, Europe) or 60 Hz (US) | Removes mains hum. Make it selectable for the market. |
| Low-pass | 40 Hz for monitoring, 100–150 Hz for diagnostic | Removes muscle noise while keeping the shape of the QRS. |
5. Show signal quality, do not hide it
Every real-world recording contains stretches that cannot be trusted. A patch that marks them, as good, fair or poor with a reason such as leads off, movement or noise, is far more useful to a clinician than one that silently draws a smooth line through them. Our app marks every trace this way and keeps gaps visible in its PDF reports.
6. Display it the way doctors read it
Clinicians read ECGs on paper at 25 mm/s and 10 mm/mV with a 1 mV calibration pulse. If the app and reports follow the same conventions at true scale, a cardiologist can read them without a manual.
A short checklist: electrodes 4 to 10 cm apart along the heart’s axis, gel electrodes for a first build, an ECG front end with lead-off detection, monitoring-grade filters that can be changed, a visible signal-quality mark and true-scale display.
Want an ECG and SpO₂ patch under your own brand?
Our prototype, Hridya, already records ECG, SpO₂ and motion and streams them to its own app. Develop it with us into your product, or let the same team build your own device.