Sensors
9-axis IMU capture window
A 2.5 s slice from the middle of a real 9-axis log: accelerometer, gyroscope, magnetometer at 100 Hz. The page states the hardware path and the firmware capture flow, then plots that window.
Sensors
IMU 9-axis data capture
ESP32
01 Overview
This is not an ESP-IDF example recap. It is a landed 9-axis capture from the lab, opened on a page so the numbers can be seen.
The raw file starts with a long header and unused GNSS columns. This page drops those fields and keeps accel, gyro and mag. 250 samples (~2.5 s at 100 Hz) are taken from the middle of the log.
Device AA-DB816B4C, firmware v1.0.3. Full scale: accel ±16 g, gyro ±2000 dps, mag 800 µT. Body axes, right-handed; xyz = forward, left, up.
GNSS is invalid in this slice, so there is no map. The plot is what the nine channels do while the board is moving.
Captured window (mid-log)
Device AA-DB816B4C, firmware v1.0.3, 100 Hz. 250 samples (~2490 ms) from the middle of the log. Accel / gyro / mag only — header and GNSS dropped.
t_ms accX_g accY_g accZ_g gyroX_dps gyroY_dps gyroZ_dps magX_uT magY_uT magZ_uT 0 -0.1113 -0.2515 0.9697 -0.366 -26.402 0.122 -35.333 11.433 2.567 10 -0.1172 -0.2388 0.9878 0.061 -24.939 -0.061 -34.967 11.567 2.4 20 -0.1313 -0.2295 1.0005 0.122 -24.756 0.122 -34.9 11.433 2.9 30 -0.1577 -0.2388 1.0088 -0.61 -25.427 -0.061 -34.567 11.433 3.467 40 -0.1899 -0.2637 1.0132 -1.341 -25.915 0.915 -34.467 11.333 3.3 50 -0.2178 -0.2764 1.0244 -1.402 -26.585 2.317 -35.067 11.333 3.067 60 -0.2329 -0.2568 1.0454 -1.89 -28.354 2.927 -34.267 11.567 3.3 70 -0.2314 -0.2261 1.0649 -4.146 -29.756 2.805 -34.967 11.267 3.933 80 -0.2153 -0.2031 1.0884 -5.549 -30.915 2.805 -34.267 11.333 4.033 90 -0.1919 -0.187 1.1089 -6.951 -32.988 2.744 -33.267 11.733 3.4 100 -0.1753 -0.1812 1.1152 -8.598 -35.854 2.561 -33.267 11.633 3.933 110 -0.1758 -0.1846 1.1211 -8.476 -38.415 3.476 -33.067 11.333 4.4 120 -0.1943 -0.1929 1.1294 -6.89 -41.829 5.976 -33.433 11.167 4.8 130 -0.2202 -0.2021 1.1289 -4.695 -47.195 8.841 -33.333 11.267 5.367 140 -0.2363 -0.21 1.1196 -2.317 -53.232 11.524 -33.5 11.567 5.467 150 -0.2363 -0.2168 1.1045 0.427 -58.72 14.268 -32.967 11.1 5.7
Table shows 16 rows; the charts plot the full window.
02 Hardware
MCU: ESP32. It paces the sample, packs a line, and writes TSV.Sensor: 9-axis IMU — 3-axis accel + 3-axis gyro + 3-axis mag. This window is aligned at 100 Hz. Mag native rate is 200 Hz; firmware time-aligns it.Scale: accel 2048 LSB/g (±16 g); gyro 16.4 LSB/dps (±2000 dps); mag 30 LSB/µT (800 µT). The curves are already in physical units.Bus: IMU on SPI or I2C (see the board schematic). The log is UART text, one sample time per line.Power and mount: keep the IMU off inductors and antenna keep-out. Soft-magnetic housings bend the mag; leave a quiet volume.Axes: right-handed, x forward, y left, z up. Firmware writes acc/gyro/mag in this body frame. Do not rotate them again on the page.
03 Software flow
1. After boot, init the IMU: full-scale, ODR 100 Hz, data-ready interrupt or a timer.
2. Every 10 ms read nine raw registers, scale to g / dps / µT. Attitude is optional; this page plots the nine raw channels.
3. One text line: timeMs plus nine numbers. GNSS writes nan when there is no fix; this page drops those columns.
4. A full log can be huge. The page cuts 250 points from the middle so the header does not get in the way.
5. The browser draws three charts — accel, gyro, mag — and a 16-row table so the magnitude can be checked by eye.
04 Core points
A shared 100 Hz time base matters more than running each sensor at its maximum rate.
A wrong full-scale setting doubles or halves the plot. The page uses the scales from the same log header.
The magnetometer hates steel and speaker magnets; the accelerometer hates mount resonance. Check this mid-log window for saturation first.
This window has no GNSS. A track is a different link. Do not mix it onto the nine-axis plot.
Need this type, or a close variant?
Pick the module that matches your idea. If you do not see it yet, WhatsApp Leo the type — the catalog is still being filled in.