求助：为何RaspberryPi采集的双极性模拟交流电流波形非正弦？

Hey there, let's walk through the most likely reasons you're not getting that clean sine wave you expect from your current measurement setup. Based on your setup details, here's what to check first:

1. Bipolar Signal vs. Unipolar ADC Mismatch #

Most 12-bit ADCs (especially common ones paired with Raspberry Pi, like MCP3208) are unipolar—they only accept input voltages from 0V up to their reference voltage (usually 3.3V). But your AC current sensor outputs a bipolar signal (-3.3mV to +3.3mV for 0.3A peak current).

Without a proper biasing circuit to shift this signal into the ADC's valid input range, the negative half of the sine wave will get clipped at 0V, resulting in a distorted half-wave or flat-lined signal. You’ll need to add a DC offset (e.g., 1.65V, half of 3.3V) to center the AC signal in the ADC’s input window.

2. Signal Amplitude is Way Too Small for 12-bit Resolution #

Let’s do the quick math:

• 0.3A peak current → sensor output = 0.3A × 22mV/A = 6.6mV peak-to-peak
• 12-bit ADC with 3.3V reference has a step size of ~0.805mV (3.3V / 4096)

That means your entire peak-to-peak signal only spans about 4 quantization steps. That’s nowhere near enough resolution to resolve a smooth sine wave—you’ll just get a blocky, distorted mess instead. You need to add a precision instrumentation amplifier (like the INA128) to boost the sensor’s output to fill most of the ADC’s input range (e.g., 0V to 3.3V, paired with the biasing circuit from point 1).

3. Insufficient Sampling Rate (Nyquist Violation) #

For a 50Hz sine wave, the Nyquist theorem says you need a sampling rate of at least 100Hz to avoid aliasing. But to get a smooth, accurate sine wave, you should aim for 10x the signal frequency (500Hz or higher).

If your Raspberry Pi sampling code is running too slow, or if system load is causing uneven sampling intervals, you’ll get aliased waveforms that look nothing like a sine wave. Double-check your code’s loop timing—use hardware timers if possible to ensure consistent, fast sampling.

4. Ground Loops & Noise Interference #

Your setup involves AC power (charger) and low-voltage electronics (Raspberry Pi, ADC), which is a recipe for ground loops and electromagnetic interference (EMI). The tiny sensor signal is super vulnerable to this noise, which can swamp the sine wave and create distorted, erratic readings.

Fixes here include:

• Using a differential input ADC or instrumentation amplifier to reject common-mode noise
• Shielding the sensor’s wiring with grounded copper tape
• Ensuring all devices share a common, clean ground (avoid connecting the charger’s ground to the Pi’s ground directly if there’s a potential difference)

5. Wiring, Calibration, or Sensor Issues #

Don’t overlook the basics:

• Double-check sensor wiring: Is power connected correctly? Is the output pin securely attached to the ADC?
• Calibrate the sensor’s zero: If there’s an offset in the sensor’s output, your sine wave will be shifted, leading to clipping or incorrect peaks.
• Verify sensor sensitivity: The 22mV/A rating might be nominal—test with a known current to confirm the actual output matches expectations. Also, make sure the ADC’s input impedance is high enough (most sensors require high-impedance loads to maintain accuracy).

6. Unstable ADC Reference Voltage #

If you’re using the Raspberry Pi’s 3.3V pin as the ADC’s reference voltage, that’s a problem—this voltage can fluctuate with the Pi’s load (e.g., when USB devices are connected). An unstable reference will cause your measurements to drift or distort. Swap in a dedicated, high-precision voltage reference (like the REF3033) to keep your ADC’s reference rock-solid.

内容的提问来源于stack exchange，提问作者Ranjan Pal