19 Sampling and Anti-Alias Limits

19.1 Start With the Measurement Story

Catch One Fast Change Without Inventing Another

Picture a spinning fan that appears to slow down when its speed actually rises. The sensor is not seeing the motion often enough, so the samples create a false pattern.

Analog-to-digital converter means a circuit that turns a voltage into a number; it is often shortened to ADC. Firmware means the software stored on the device. Both the hardware filter and software rate shape the recorded signal.

Feed in a known tone below and above the planned limit. Change the sample rate and filter, then retain the source frequency, settings, samples, spectrum, error, and decision.

This runway does not prove that every real signal is captured. The deeper sections explain the sampling theorem, aliasing, pre-sample filtering, rate choice, jitter, and validation limits.

A waveform sampled too slowly can tell a confident lie. The design story begins with how fast the physical signal can change, what frequencies must be kept, and what filtering is needed before the ADC samples it.

Phoebe’s Field Notes: Why Four Samples Buys 2x, Not 4x

The mathematical gist. One ADC code step has RMS rounding noise $q/\sqrt{12}$. Averaging $M$ independent readings reduces that noise to $q/\sqrt{12M}$, so $M=4$ buys a factor of 2, 6.02 dB, or one effective bit. It cannot remove this chapter’s coherent 90 Hz tone after it has folded to 10 Hz: an alias repeats systematically rather than cancelling like random noise.

Math Bridge · guided foundationsWhy do four samples buy only one extra bit?Let Phoebe derive the square-root averaging law and its strict aliasing limit. Learn the maths

19.2 Learning Objectives

After this page, you should be able to:

• Explain why sampling happens before firmware filtering and calibration.
• Use the Nyquist rule to identify sample rates that can alias sensor signals.
• Calculate where an out-of-band frequency appears after sampling.
• Place analog anti-alias filtering before the ADC and digital filtering after it.
• Decide when oversampling helps noise performance without pretending to fix aliasing.

19.3 After Data Processing

Sensor Data Processing teaches moving averages, median filters, calibration equations, validation rules, and persistence of calibration coefficients. Those techniques assume the ADC already captured the physical signal correctly. This page checks the boundary before that assumption: the analog sensor path, sample rate, and anti-alias filter determine whether the data is real signal or a believable artifact.

A practical deployment treats acquisition and processing as one contract. If the logger samples too slowly, a vibration, motor-drive ripple, or power-line component can fold into the operating band and survive every later average, calibration, dashboard, and alert rule. Use this page before tuning firmware filters when a signal changes after a sample-rate change, appears at an implausible frequency, or arrives from an analog path without a documented bandwidth limit.

19.4 Snapshots of a Continuous World

The physical world is continuous, but a data pipeline sees only snapshots: the ADC captures the signal at a fixed sample rate fs and throws away everything in between. If the signal changes slowly compared with how often you sample, the snapshots faithfully trace it. If it changes quickly, the snapshots can lie.

When a signal contains frequencies faster than half the sample rate, those fast components reappear disguised as slow ones. This is aliasing, and it is the single most damaging mistake in sensor data acquisition, because the result looks like perfectly plausible data. The guard against it is the Nyquist rule: sample at more than twice the highest frequency present in the signal.

As a concrete acquisition check, imagine an accelerometer on a small pump. The mechanical signal of interest is shaft imbalance around 12 Hz, but the motor drive injects a 38 Hz ripple onto the analog line. If the logger samples at 50 samples/s, Nyquist is 25 Hz, so the 38 Hz ripple cannot be represented honestly. It folds to |38 - 50 x round(38/50)| = 12 Hz, exactly where the real imbalance feature lives. The trend plot now shows a believable 12 Hz vibration even if the shaft is healthy. The fix is not a smarter dashboard; it is sampling and filtering so the 38 Hz component cannot enter the ADC path.

A quick field test is to repeat the capture at a different sample rate. A real 12 Hz vibration stays at 12 Hz when you move from 50 SPS to 80 SPS. The aliased 38 Hz interference moves from 12 Hz at 50 SPS to |38 - 80 x round(38/80)| = 38 Hz, safely below the new 40 Hz Nyquist limit. If the apparent feature moves when the sample rate changes, suspect aliasing before diagnosing the machine.

19.5 Predicting Frequency Aliases

When a component at frequency f is sampled at fs, it appears at an aliased frequency:

f_alias = | f - fs × round(f / fs) |     (valid up to fs/2)
Nyquist frequency = fs / 2

Sample rate fs = 50 samples/s  ->  Nyquist = 25 Hz

A 30 Hz vibration component (above 25 Hz) aliases:
  round(30/50) = round(0.6) = 1
  f_alias = |30 - 50 × 1| = 20 Hz

A 55 Hz component aliases:
  round(55/50) = round(1.1) = 1
f_alias = |55 - 50 × 1| = 5 Hz

Turn that into a design rule before deployment. If the fastest useful vibration feature is 20 Hz, sampling at 50 SPS gives a 25 Hz Nyquist limit with almost no room for a real analog filter to roll off. A safer design might sample at 200 SPS, place an analog low-pass near 40 Hz to suppress higher-frequency electrical noise, then apply a digital low-pass and decimate to 50 SPS for storage. The stored data rate is still small, but the analog front end had enough bandwidth margin to prevent false low-frequency content.

The 50 SPS logger cannot represent the machine vibration assumed in this example. Inspect Figure 19.1 to compare an invalid sample rate with a defensible acquisition chain before any digital filter is applied.

Figure 19.1 first identifies the Real 50 Hz Signal and its 20 ms period. The Undersampled (Aliased) panel then uses an 80 Hz sample rate, whose 40 Hz Nyquist limit excludes the input; the samples reconstruct a false 30 Hz component from |80 − 50|. In Properly Sampled, 500 Hz provides ten samples per cycle and an Anti-alias filter at 180 Hz blocks higher-frequency content before conversion, allowing the FFT to match the mechanics. The final inequality, fmax < fsample / 2, is therefore an analogue-front-end contract, not a post-processing option. For the 50 SPS logger below, even the fundamental 50 Hz component is already outside that contract, so the remedy is to change acquisition hardware or narrow the physical bandwidth before the ADC.

19.6 Aliasing and Oversampling Limits

Once a fast component has folded into the low band, the sampled sequence is identical to what a genuine low-frequency signal would have produced. There is no information left that distinguishes them, so no algorithm can separate them after the fact. That is why the anti-alias filter has to be analog and has to sit before the sampler.

The optional mathematics panel below works only with numbers already used in this chapter: the 12 Hz pump feature, the 38 Hz ripple, the 50 SPS and 80 SPS field test, the 200 SPS safer capture, the 40 Hz analog cutoff, and the four-sample oversampling example. It shows the arithmetic so the sampling decision can be checked rather than trusted.

Oversampling also improves noise handling, but only under the right assumptions. Averaging four independent samples can improve signal-to-noise ratio by about 2:1, roughly one extra effective bit, because random noise partially cancels. That benefit does not apply to a coherent 90 Hz interferer that has already aliased to 10 Hz; every decimated record will faithfully preserve the false tone. In practice, use oversampling to relax analog-filter steepness and reduce broadband noise, while still ensuring the analog input is band-limited before the first sample is taken.

The order of operations is the whole lesson: band-limit in analog, then sample. Reverse that order and the data is corrupted before your firmware ever touches it.

19.7 Rescue a Motor Trace Before the ADC

In the sampling trace, a healthy motor turns with a real 12 Hz vibration while its drive adds a 38 Hz electrical ripple. At this anti-alias boundary, if a logger takes only 50 samples each second, both motions can land on the same plotted frequency. Once the ADC has confused them, a dashboard cannot tell which 12 Hz trace came from the shaft.

Read Figure 19.1 from the continuous input toward the stored samples. The wanted signal and unwanted high-frequency energy meet the analogue low-pass filter first. The ADC then takes spaced snapshots, and the digital analysis receives only those samples. The anti-alias filter sits before the ADC because filtering after sampling cannot unfold a frequency that already has the wrong identity.

At 50 samples/s, Nyquist is (50/2=25\ \text{Hz}). The 38 Hz ripple folds to |(38-50)| (=12\ \text{Hz}), exactly on the shaft feature. Raising the rate to 100 samples/s moves Nyquist to 50 Hz, so 38 Hz remains distinct. A filter that strongly reduces input above 20 Hz offers another route when the 12 Hz motion is the only required result.

Averaging four independent samples cuts random noise by √4, or a factor of 2. It does not erase the coherent 12 Hz alias because that false wave repeats in step with the samples. This is the practical split between random noise reduction and alias prevention.

19.8 Release Checklist

Before relying on a processed sensor stream, confirm the acquisition chain has evidence for each item:

Start by the highest expected signal frequency and likely interference frequencies are documented. Then the chosen sample rate leaves margin above twice the highest frequency that can reach the ADC. Next the analog anti-alias filter cutoff is below Nyquist and before the ADC input. After that a sample-rate change test distinguishes real features from movable aliases. Finally digital filtering is used only after acquisition has prevented irreversible aliasing.

19.9 See Also

Start by Sensor Data Processing for filtering, calibration, validation, and persistence patterns. Then Signal Conditioning for analog front-end design before ADC conversion. Next Signal Processing for IoT Sensors for additional filtering and transform tools. Finally Multi-Sensor Data Fusion for combining already trustworthy sensor readings.

19.10 Next

Return to Sensor Data Processing when the acquisition boundary is verified, then choose the digital filter and calibration method that matches the captured signal.