The transfer function for a high pass filter is a mathematical expression, typically written as H(s) or H(jω), that defines the ratio of output voltage to input voltage across different frequencies, effectively blocking DC and low frequencies while passing high frequencies. In a real circuit, applying this function changes the signal by stripping away unwanted DC offsets, attenuating low-frequency rumble or 50/60Hz mains hum, and introducing a frequency-dependent phase shift that approaches +90° as frequencies drop toward zero.
To visualize the physics, think of the series capacitor as a flexible rubber membrane stretched tightly across a water pipe: it completely blocks a steady, continuous flow of water (DC), but rapidly vibrating pressure waves (AC) easily flex the membrane and transmit through to the other side. Below, we break down the exact math, run a bench-ready numeric example, and provide a concrete decision path for selecting your physical components.
The Core Math: Deriving the Transfer Function
For a standard first-order passive RC high pass filter, the circuit consists of a capacitor (C) in series with the signal path and a resistor (R) shunting to ground. The output voltage is measured across the resistor. Using the voltage divider rule in the Laplace domain, the transfer function H(s) is expressed as:
H(s) = (sRC) / (1 + sRC)
Where s is the complex frequency variable (s = jω), R is resistance in ohms, and C is capacitance in farads.
When we move to the frequency domain by substituting s = jω (where ω = 2πf), we can extract two critical pieces of data that dictate how the filter behaves on your workbench:
- Magnitude (Gain): |H(jω)| = (ωRC) / √(1 + (ωRC)²). At the cutoff frequency (fc), the magnitude drops to 1/√2, or roughly -3.01 dB.
- Phase Shift: φ = arctan(1 / ωRC). At very high frequencies, the phase is 0°. At the cutoff frequency, the phase is +45°. At very low frequencies, it approaches +90°.
The cutoff frequency (the -3dB point) is defined by the equation fc = 1 / (2πRC). This single equation is the bridge between your theoretical signal requirements and the physical components you pull from your parts bin.
Worked Numeric Example: Designing a 1 kHz Audio Coupling Filter
Let’s say you are building an audio preamplifier and need to block the DC bias from a previous transistor stage while passing the audio band (20 Hz to 20 kHz). You decide on a target cutoff frequency (fc) of 1 kHz to safely roll off sub-sonic mechanical noise without touching the fundamental frequencies of most instruments.
Step 1: Pick a standard capacitor value.
In practical design, it is almost always easier to select a standard capacitor value first, as resistor ranges are much denser (E24 or E96 series). We select a 100 nF (0.1 µF) capacitor.
Step 2: Calculate the required resistance.
Rearranging the cutoff formula to solve for R:
R = 1 / (2π × fc × C)
R = 1 / (2π × 1000 Hz × 100 × 10⁻⁹ F)
R = 1 / 0.0006283
R ≈ 1591.5 Ω
Step 3: Select a physical resistor and verify.
1591 Ω is not a standard value. If we use the common E24 (5%) series, the closest value is 1.5 kΩ. Let's calculate the actual cutoff frequency with a 1.5 kΩ resistor:
fc = 1 / (2π × 1500 × 100 × 10⁻⁹) = 1061 Hz.
Where You Meet This in Practice
You will encounter the high pass transfer function across almost every discipline of electronics. Here is where it physically manifests on the jobsite or lab bench:
- Oscilloscope AC Coupling: When you press the 'AC Coupling' button on your scope, the instrument switches in a hardware high pass filter (typically with a cutoff around 10 Hz). This applies the transfer function to block the DC offset of a power rail so you can zoom in on the millivolt-level switching ripple.
- Audio Tweeter Crossovers: In a passive speaker crossover, a series capacitor acts as a first-order high pass filter, blocking low-frequency bass energy from reaching and destroying a fragile tweeter voice coil.
- Biosignal Amplifiers (ECG/EEG): Skin-electrode interfaces generate massive DC offset potentials (up to 300 mV). Medical front-ends use high pass filters with an extremely low cutoff (e.g., 0.05 Hz) to block this DC while passing the 1 Hz to 40 Hz ECG waveform.
Common Confusions: Brick Walls and Loading Effects
When engineers and hobbyists first work with the transfer function for a high pass filter, two major misconceptions lead to broken prototypes.
Confusion 1: The -3dB point is a 'brick wall'.
A first-order passive RC filter does not instantly eliminate frequencies below fc. The transfer function dictates a roll-off slope of -6 dB per octave (or -20 dB per decade). If your cutoff is 1 kHz, a 500 Hz signal is only attenuated by about -7 dB, not eliminated. If you need a steeper drop-off, you must cascade multiple stages or use an active topology like a Sallen-Key filter, which alters the transfer function denominator to increase the pole count.
Confusion 2: Ignoring the load impedance.
The math above assumes the output is connected to an infinite impedance (an open circuit). If you connect your 1.5 kΩ RC filter directly to a 1 kΩ load (like a low-impedance headphone driver or a poorly buffered ADC input), the load resistor acts in parallel with your filter resistor. Your effective R drops to 600 Ω, and your cutoff frequency violently shifts upward to 2.6 kHz. Always buffer passive filters with a high-impedance op-amp voltage follower.
Component Selection Decision Tree
Calculating the math is only half the battle; picking the right physical part ensures the transfer function holds up in reality. Capacitor dielectrics introduce parasitic effects that warp your filter response. Use this decision path to select your components.
| Application Scenario | IF your priority is... | THEN select this Capacitor | AND select this Resistor |
|---|---|---|---|
| High-Fidelity Audio AC Coupling | Low microphonics, zero piezoelectric ringing | Wima MKS / MKP Film (Polyester/Polypropylene) | 1% Metal Film (e.g., Yageo MFR-25) |
| General Purpose Signal Conditioning | Compact footprint, stable capacitance over voltage | Murata C0G / NP0 Ceramic (Avoid X7R/Y5V entirely) | 1% Thick Film (e.g., Panasonic ERJ) |
| Ultra-Low Frequency (ECG / Sensors) | Massive capacitance in a small space without leakage | Panasonic OS-CON Polymer or Tantalum (observe polarity!) | 0.1% Precision Metal Foil |
| RF / High Speed Digital (>10 MHz) | Low Equivalent Series Inductance (ESL) | ATC 600S Series (RF Ceramic, 0402 package) | Thin Film 0.1% (e.g., Susumu RG) |
The Concrete Pick: If you are building a general-purpose, sub-100 kHz signal conditioning circuit and don't have a specific constraint, default to a 100nF C0G/NP0 ceramic capacitor (like the KEMET C315C104J1G5TA) paired with a 1.58kΩ 1% metal film resistor. C0G dielectrics have near-zero voltage coefficients and temperature drift, meaning your physical filter will actually match your mathematical transfer function across temperature changes and varying signal amplitudes.
FAQ: Troubleshooting and Edge Cases
Why does my high pass filter output look like a distorted triangle wave instead of a sine wave?
You are likely experiencing slew-rate limiting or dielectric absorption. If you used a cheap Y5V or X7R ceramic capacitor, the dielectric absorbs charge and releases it slowly, smearing the waveform. Switch to a C0G/NP0 ceramic or a film capacitor.
Can I just use an electrolytic capacitor for large values?
Yes, but you must account for polarity and Equivalent Series Resistance (ESR). Electrolytics have high leakage currents and wide tolerances (often ±20%). If your transfer function relies on a precise cutoff, electrolytics will introduce massive errors. Furthermore, if the signal swings negative, a polarized electrolytic will be reverse-biased and fail. Use a back-to-back series pair or a non-polarized bi-polar electrolytic if you must.
How do I calculate the transfer function for an active high pass filter?
For a first-order active filter (an RC network feeding a non-inverting op-amp buffer), the transfer function magnitude remains identical to the passive version, but the gain is multiplied by the op-amp's closed-loop gain (1 + Rf/Rg). For a second-order active Sallen-Key topology, the denominator of the transfer function gains an s² term, introducing a Q-factor and a steeper -12dB/octave roll-off. Refer to the Texas Instruments Filter Design Guide for exact Sallen-Key component scaling.
Understanding the transfer function for a high pass filter moves you from guessing component values to engineering predictable signal chains. Respect the -3dB slope, buffer your outputs, and always verify your dielectric material.
