SignalForge

Introduction Most DSP filter designs fail not in theory — but in embedded deployment. Fixed-point systems introduce: quantization overflow feedback amplification limit cycles Without explicit engineering safeguards, instability is inevitable. Why Floating-Point Designs Break in Fixed-Point Hardware Key failure modes: coefficient truncation reduced dynamic range nonlinear saturation feedback noise growth IIR structures are especially vulnerable. Scaling as a First-Class Design Variable Proper deployment requires: per-section scaling headroom budgeting bounded internal states Blind normalization is insufficient. ...

Introduction Many DSP pipelines behave differently each time they run. Detection thresholds shift. Filters change. Results drift. This non-determinism is often blamed on “noise sensitivity.” In reality, it is almost always caused by fragile pipeline architecture. In low-SNR environments — where estimator variance, drift, and noise bursts dominate — naive DSP workflows amplify uncertainty instead of controlling it. This article presents a complete deterministic detection architecture for real-world low-SNR DSP systems. ...

Introduction Industrial measurement systems operate under constantly changing physical conditions. Rotational speed, load, temperature, and power quality fluctuate over time. These variations cause harmonic interference patterns to drift continuously. Simple static filtering approaches rapidly lose effectiveness. This article explains the nature of drifting harmonic interference in industrial systems and outlines robust characterization strategies. Physical Origins of Drift Common sources include: variable-speed motors load-dependent vibration modes power electronics switching variation thermal expansion effects These mechanisms shift fundamental frequencies and all associated harmonics together. ...

Introduction FIR and IIR filters both appear stable in textbook theory. In embedded DSP systems, their real-world behavior can be dramatically different. Engineers frequently discover that designs which simulate perfectly become unstable, noisy, or fragile once deployed. This article explains the practical stability differences between FIR and IIR filters under real numerical constraints. Theoretical Stability vs Practical Stability Mathematically: FIR filters are always stable IIR filters are stable if poles remain inside the unit circle Numerically: ...

Introduction Frequency magnitude plots rarely reveal phase behavior. Yet in real-time DSP systems, phase distortion and group delay often determine overall system performance. Applications sensitive to timing include: control loops audio monitoring feedback systems instrumentation pipelines Ignoring phase characteristics can destabilize otherwise correct magnitude designs. What Is Group Delay? Group delay measures: how long different frequency components are delayed In FIR linear-phase filters: delay is constant predictable In IIR filters: ...

Introduction Notch filters are highly effective at suppressing narrowband interference — when the interference stays exactly where it is expected. In real systems, it rarely does. Engineers frequently encounter interference that: drifts with temperature shifts with load or aging wanders slowly over time appears intermittently across a frequency band Designing a narrow notch at a single center frequency often works in the lab and fails in the field. This article explains why frequency drift breaks traditional notch designs and how STFT-based drift tracking enables robust suppression in real-world DSP systems. ...

Introduction False tonal detection is one of the most common structural failure modes in automated DSP pipelines. In noisy environments, PSD estimators frequently produce spurious peaks caused by: estimator variance leakage ripple random noise bursts If filters are synthesized directly from these peaks, systems end up suppressing noise instead of interference. As shown in: Why PSD Peak Detection Fails in Low SNR Signals How STFT Cross-Validation Improves Low-SNR Tone Detection frequency magnitude alone is insufficient for deterministic detection. ...

Introduction In low-SNR environments, PSD-based peak detection often becomes unstable. Spectral variance, leakage, and noise ripple cause dominant frequency bins to shift randomly between measurements. As discussed in Why PSD Peak Detection Fails in Low SNR Signals, the core issue is not mathematical correctness — it is the loss of determinism. Short-Time Fourier Transform (STFT) introduces temporal structure into spectral analysis, enabling engineers to separate true tonal interference from stochastic noise behavior. ...

Introduction Accurate noise floor estimation is fundamental to spectral analysis, detection thresholds, and filter verification. Yet many DSP pipelines still rely on simple averaging: mean PSD levels RMS magnitude global spectral averages In real signals, these methods frequently produce unstable and misleading results. This article explains why average-based noise estimates fail in practice and how percentile statistics provide robust noise floor measurement for engineering-grade DSP systems. The Reality of Real-World Noise Ideal Gaussian noise assumptions rarely hold in production systems. ...

Introduction Most DSP algorithms assume stationary frequency content. Real systems violate this constantly. Drift arises from: temperature variation mechanical speed changes power supply instability oscillator tolerance Ignoring drift produces fragile filters. Physical Sources of Drift Common mechanisms include: crystal oscillator offset motor RPM variation thermal expansion nonlinear load behavior Drift is structural — not noise. Statistical Drift Envelope Modeling Engineering pipelines estimate: minimum frequency maximum frequency bandwidth expansion percentile motion limits Rather than single-point frequency. ...