Compare the reading.
Return to the reference.

What number are you comparing?

A phone magnetometer measures components of the magnetic field near the device. It is not a general-purpose radiation meter. It does not provide a measurement of Wi-Fi, radio-frequency exposure or ionizing radiation.

Before recording anything, note what your chosen tool actually displays: an individual axis, total-field magnitude, peak, or change relative to a set baseline. These are different quantities. Do not compare a total in one tool with a baseline-relative change in another, even if both labels say magnetic field.

Retain the unit exactly as displayed. Apple's underlying Core Motion axis values use microteslas (µT), but an app can present another unit. Keep the same unit and readout mode throughout the log. If either changes, start a separate comparison rather than silently merging the numbers.

Baseline reset is not sensor calibration. A chosen local reference lets a tool display change from that reference. It does not make the phone a calibrated laboratory instrument or certify accuracy. Apple separately describes raw and processed sensor streams; this worksheet does not assume which stream your app uses.

Write down the setup before moving

For an ordinary, non-safety-critical observation, choose two accessible, stable positions on a tabletop or other permitted surface: reference A and comparison B. Record the same phone edge as your reference for positioning. This is a convenient repeatable edge, not a claim about where the sensor is inside the phone.

1. Keep one device and one tool. Record the phone model, app/version if known, readout mode and unit. A missing sensor or unavailable reading is not a zero reading.
2. Retain orientation. Note which edge faces the ordinary test object and whether the phone is flat or upright. Rotating the phone changes the meaning of axis readings.
3. Note nearby accessories and surroundings. A magnetic case, charger, speaker or motor can affect the observation. Remove magnetic accessories when practical and record what remains; do not move or dismantle equipment to get a desired result.
4. Describe both positions. Mark A and B and record their spacing from the same phone edge using an ordinary ruler if useful. This describes your setup; it does not infer sensor-to-object distance or detection range.

Use only an ordinary object and a location where you can work without risk. This is not a procedure for opening equipment, approaching live wiring, scanning walls before drilling, searching for hazards or checking medical-device compatibility.

A → B → A tells you more than a single peak

Use a live, current axis, total-field or baseline-relative reading for this comparison, not a retained peak. A running or recent peak can stay high after the phone returns to A; it cannot show whether the current reference reading returned. Record peaks separately if useful. A peak-only display cannot perform this return check.

Record a settled current observation at A, move to B with the same orientation, then return to A. Repeat the comparison once. Keep any app baseline unchanged during that sequence; resetting it halfway changes the reference. Record only what the display actually provides and leave unavailable information blank.

Scroll sideways on a narrow screen. These are paper-log fields, not a promise that the optional app saves a scan or exports this table.

If the current reference reading changes noticeably when you return, record the result as unsettled or setup changed. Do not discard the return readings to make a convincing detection story. If a similar A-to-B difference repeats, describe it narrowly: a repeated difference under this particular setup. That does not identify its cause or material.

This worksheet does not define a pass/fail tolerance. Phone models, orientation, surroundings and readout processing differ. An app's colour, sound, confidence label or sensitivity setting is not a health limit, safety classification or independently certified accuracy.

A copyable observation note

Paper or your own private document is enough. No app download, account, email address or purchase is required to use this guide. The website does not read your phone's sensors or upload observations.

Purpose of ordinary observation:
Device / app and version:
LIVE current quantity (axis / total / relative change):
Unit exactly as displayed:
Phone orientation / same edge used:
Case / magnetic accessories / nearby electronics:
Position A / position B / spacing method:
Baseline reference, if the tool uses one:

A1 displayed value / unit / note:
B1 displayed value / unit / note:
A2 displayed value / unit / note:
B2 displayed value / unit / note:
A3 displayed value / unit / note:

Optional axes or calibration status ONLY if displayed:
Optional retained peak, separate from the five current readings:
Mode, baseline or surroundings changed during sequence:
Did return readings reproduce the starting observation?
What was observed, without assigning a cause:
What remains unknown / next ordinary check:

Do not include unnecessary addresses, access details or personal information in a shared log. Retain uncertain or missing readings honestly.

A magnetic-field observation is not a material or safety test

A response does not establish metal type, value, buried wiring, depth or distance. A quiet reading does not establish that metal or a hazard is absent. The guide and optional phone app cannot replace a suitable professional detector, electrical checks or a qualified assessment.

Do not use a phone reading to make a medical, radiation-exposure, drilling, utility-location or security decision. No “safe” threshold, guaranteed detection range or certification is supplied here.

Primary background references: Apple's CMMagneticField documentation describes three axes in microteslas; its raw magnetic-field property includes geomagnetic field and device/surroundings bias. The processed device-motion property describes device-bias compensation. These explain why readout modes matter; they are not an endorsement, accuracy test or proof of our app's implementation.