We mounted two 100 mV/g accelerometers on a TDS-11SA last month and caught a carrier bearing trending up at 2× running speed before it forced downtime. If you’ve tried similar setups offshore, what sampling rate and mounting locations gave you reliable alarms without a flood of false positives?
@xavierf21’s 5 s persistence helps, but what really cut our false alarms was gating: only evaluate the 1.8–2.2× band RMS when speed is steady (±2% over about 8 s) and torque >30% rated, and ignore run‑up/auto‑driller ramps. Have you tried that on the 11SA?
We killed most of the “2×” falses on a TDS-11SA by ditching mag bases and using an M8 stud mount on the carrier cover, torqued about 2 N·m with a dab of 243, plus proper strain relief so the cable doesn’t tug. Kept the same two 100 mV/g sensors and 2 kHz sampling was plenty once the mount was rigid. @xavierf21 did you notice mag bases ringing when the VFD was humming?
Quick win: 12.8 kHz bursts with a 2–8 kHz envelope and alarm on 2× in the envelope spectrum cut our mis-alarms about 60% on a similar top drive. If you can’t drill, an adhesive stud at the planet-carrier endplate load zone held up fine; @jordan_hall92, can you grab a tach off the VFD analog out to keep it speed-synchronous?
On a TDS‑11SA, 25.6 kHz continuous with a 1–8 kHz envelope worked well, but we only raise a ‘2×’ alarm when the two sensors show magnitude coherence >0.6 in that band. Put one radial near the loaded side of the carrier bearing and a second axial on the thrust face; a narrow pump‑stroke notch acts like noise‑canceling to kill drillstring bleed‑through. If you can stream the VFD RPM tag, set speed‑state baselines — what’s the high‑end bandwidth on your 100 mV/g pair?