Measuring a handlebar mass damper with ShokIMU

The setup

A mass damper inside the handlebar is meant to soak up high-frequency buzz and take the edge off repeated hits. Subjective feel is easy to argue about. ShokIMU gives a repeatable trail measurement: a compact BLE IMU on the bar, GPS on the phone, and the same analytics the app uses for roughness, vibration bands, and impact events.

Nine CSV sessions were recorded across two days. Conditions were assigned chronologically: the first session as no-damper day 1; the remaining day-1 sessions as mass damper (5 runs, 37.7 moving minutes); day-2 sessions as no damper (3 runs, 18.1 moving minutes). The day-1 no-damper recording is a short 7.2-minute baseline—useful, but not the primary comparison.

How we compared the runs

1. Parsed each session and separated IMU rows from the appended GPS sample footer.
2. Kept riding only: GPS speed ≥ 1.0 m/s, closing gaps shorter than 3 s and keeping segments of at least 5 s.
3. Resampled riding segments to 100 Hz and compared acceleration-magnitude energy, high-pass filtered from 0.5–50 Hz. Magnitude avoids axis dependence from center vs right-side bar placement; placement still confounds absolute amplitude.
4. Speed-matched the paired conditions into GPS-speed bins where both sides had at least 20 seconds of data, so a faster day does not masquerade as a harsher bar.

Frequency comparison is capped at 50 Hz because the effective BLE/export rate for these sessions is around 100 Hz—sharper buzz above that is not reliably separable here.

Condition summary

Unmatched totals (all riding time in each condition). Median RMS and impact rate already hint that day 2 without the damper was the harsher set of runs.

Mass damper vs no-damper day 2

This is the strongest unmatched comparison: multiple/longer runs on both sides. Relative change is mass damper minus baseline, as a percent of the no-damper day-2 value.

Speed-matched check

Speed bins where both conditions have at least 20 seconds of data. This is the fairer test: same velocity mix, then compare bar motion.

Mass damper day 1 vs no-damper day 2

4 matched bins · 879 s of overlapping riding time

Mass damper day 1 vs no-damper day-1 first run

4 matched bins · 324 s of overlapping riding time — short baseline, treat as directional only

What we take from this

• Against the longer no-damper day, the damper sessions show lower overall RMS, fewer hard impacts, and slightly less 30–50 Hz chatter—both unmatched and speed-matched.
• The spectral mix barely moves: terrain (15–30 Hz) and chatter shares stay within a few percent. The clearer signal is amplitude and impact rate, not a wholesale shift of energy into another band.
• The single no-damper first run is too short to stand alone. Speed-matched against that baseline, the damper day looks busier—likely line, fatigue, or sample-size noise rather than a reversed damper effect.
• Sensor placement moved from center to the right side of the bar. Directional roll/yaw and raw-axis metrics should not be over-interpreted; this write-up sticks to acceleration magnitude and relative band mix.

That is the point of the kit: clip a module on the part you care about, ride the same trail, and compare conditions with GPS-aligned motion instead of guessing from pump feel.