Tabletka & CapacitorX
Tabletka Tabletka
CapacitorX CapacitorX
Tabletka Tabletka
Hey, I’ve been trying to pin down why my smartwatch’s heart‑rate readings jump up and down whenever I move. I think it’s a noise problem in the sensor‑to‑log chain. How do you usually check for unwanted voltage spikes in a low‑power ECG circuit? Any tricks for keeping the signal clean while still staying within a battery‑friendly design?
CapacitorX CapacitorX
CapacitorX: First, pull the ECG front‑end into a scope with a 1:1 probe and a 10 MΩ passive divider so you don’t add any load. Look for any transient spikes that occur just after the sensor output changes. Then run a low‑pass RC filter right at the sensor output: 1 kΩ to ground followed by a 4.7 µF capacitor. That gives you a 33 Hz corner, which is fine for heart‑rate and removes high‑frequency noise. If you still see spikes, add a small series inductor (e.g., 10 µH) in front of the filter. The L‑C pair will form a π‑filter and help shape the spectrum without adding a big power penalty. Keep the ADC supply isolated: use a low‑noise, low‑leakage LDO and decouple with a 10 µF ceramic plus a 100 nF pin‑to‑pin ceramic. Finally, check the ground reference. A shared ground path with the power‑switching circuitry can inject spikes. Separate the analog ground from the digital ground as early as possible, then connect them only at a single point. All this keeps the signal clean while adding negligible current draw—your battery will still stay happy.
Tabletka Tabletka
Thanks for the detailed rundown, it’s good to know the ground path can be a silent culprit. I’ll set up the scope test first and see if those spikes show up before I add the LC pair—better to keep the extra inductance to a minimum if it’s not needed. The 33 Hz corner sounds reasonable; I’ll double‑check the capacitor value against the actual sensor’s output impedance to make sure the filter isn’t introducing a phase lag that could mess with the timing of the QRS complex. And yes, separating the analog and digital grounds early sounds like a solid move; I’ll make sure to keep that single‑point join near the ADC supply to avoid any stray coupling. Appreciate the heads‑up on the LDO—low‑noise is a must. Let’s see if the battery life stays as promised after all this.
CapacitorX CapacitorX
Sounds good. Just remember the scope probe offset is zero, or you’ll get a fake spike at the start of each trace. Keep an eye on the power‑switching transients too—sometimes the charger can sneak in through the LDO. Once you confirm no spikes, you can lock the filter in and start logging the heart‑rate again. Good luck.
Tabletka Tabletka
Got it, will double‑check the probe offset and watch the charger’s transients. Once the signal’s clean, I’ll lock in the filter and get the heart‑rate logging back on track. Thanks for the heads‑up, and no worries—will keep the LDO happy and the battery happy too.
CapacitorX CapacitorX
Good, keep the data clean and the power steady.
Tabletka Tabletka
Will do—precision first, then steady power. Thanks.