Low-Pass vs High-Pass Filter: Differences, Circuits, Cutoff Frequency, and Applications

Electronic systems often need to separate useful signals from unwanted frequency components. A sensor interface may need to remove high-frequency noise, an amplifier may need to reject DC offset, and an ADC may require attenuation above its usable signal bandwidth. Low-pass and high-pass filters address these problems in opposite ways.

Quick answer: A low-pass filter passes frequencies below its cutoff frequency and attenuates higher frequencies. A high-pass filter passes frequencies above its cutoff and attenuates lower frequencies. In a basic RC circuit, the low-pass output is measured across the capacitor, while the high-pass output is measured across the resistor.

Neither filter behaves like an ideal on/off switch. A real filter has a transition region, and its attenuation increases gradually according to its order and response type.

Low-Pass vs High-Pass Filter: Quick Comparison

The words “low” and “high” are always relative to the selected cutoff frequency. A 10 kHz signal may be in the passband of one low-pass filter and in the stopband of another.

What Is a Frequency Filter?

A frequency filter is a circuit or algorithm that changes a signal according to frequency. Important filter terms include:

1. Passband: The frequency range intended to pass with acceptable loss.
2. Stopband: The range that must be sufficiently attenuated.
3. Transition band: The region between the passband and stopband.
4. Cutoff frequency: A defined boundary or corner in the frequency response.
5. Attenuation: The amount by which a signal is reduced, commonly expressed in decibels.
6. Phase shift: The frequency-dependent timing difference between the input and output.
7. Filter order: The number of poles governing the ultimate roll-off rate.

For a first-order RC filter, and for the conventional Butterworth response, the cutoff is the −3 dB point. The output voltage magnitude there is approximately 70.7% of the passband value—not zero.

Other filter families may define their important band edge differently. For example, a Chebyshev filter’s specified cutoff can refer to the edge of its allowed ripple rather than exactly −3 dB.

What Is a Low-Pass Filter?

A low-pass filter, or LPF, allows lower-frequency components to reach the output while progressively attenuating higher-frequency components.

In a basic passive RC low-pass filter:

1. A resistor is connected in series with the signal.
2. A capacitor is connected from the output node to ground.
3. The output is measured across the capacitor.

A capacitor’s reactance is:

[

X_C=\frac{1}{2\pi fC}

]

At low frequencies, the capacitor has relatively high impedance, so little signal is diverted to ground and the output remains close to the input. As frequency increases, the capacitor impedance falls, directing more of the high-frequency signal toward ground.

A low-pass filter can therefore smooth abrupt waveform changes. If a square wave is applied to a sufficiently low-cutoff RC filter, the output edges become rounded. With a time constant much longer than the pulse period, the circuit can behave approximately like an integrator.

Typical low-pass filter applications include:

1. Filtering high-frequency noise from analog sensors
2. Limiting bandwidth before an ADC
3. Smoothing PWM into an approximate analog voltage
4. Removing DAC image components
5. Reducing high-frequency noise on references or low-current supply nodes
6. Routing low-frequency audio to a woofer
7. Suppressing harmonics in signal-generation circuits

An anti-aliasing filter must be designed from the required signal bandwidth, sampling rate and stopband attenuation. Simply placing the −3 dB corner at the Nyquist frequency does not guarantee sufficient rejection.

What Is a High-Pass Filter?

A high-pass filter, or HPF, passes higher-frequency signal components while attenuating low-frequency components.

In a basic passive RC high-pass filter:

1. A capacitor is placed in series with the signal.
2. A resistor is connected from the output node to ground.
3. The output is measured across the resistor.

At DC, the series capacitor eventually behaves like an open circuit, preventing a steady voltage from reaching the output. At higher frequencies, its reactance decreases and more of the changing signal appears across the resistor.

This makes high-pass filters useful for:

1. AC coupling between amplifier stages
2. Removing DC offsets
3. Suppressing slow sensor baseline drift
4. Reducing low-frequency motion artifacts
5. Removing rumble from audio signals
6. Routing high-frequency audio to a tweeter
7. Detecting transitions or waveform edges
8. Rejecting low-frequency interference in communication systems

When its cutoff frequency is high relative to the input waveform’s fundamental frequency, a first-order high-pass filter can approximate a differentiator. A square-wave input then produces positive and negative pulses around its rising and falling edges.

Low-Pass and High-Pass RC Circuits

The same resistor and capacitor values can produce either a low-pass or high-pass response. What matters is the circuit topology and the point from which the output is measured.

It is therefore incomplete to say that the components are simply “swapped.” Filter behavior depends on the complete impedance network and the selected output node.

In practical circuits, the signal source impedance becomes part of the filter. The next stage’s input impedance also loads the output. Consequently, the actual cutoff frequency can differ from a calculation that includes only the labeled R and C.

RC Filter Cutoff Frequency Formula

For an unloaded first-order RC low-pass or high-pass filter, the cutoff frequency is:

[

f_c=\frac{1}{2\pi RC}

]

The equation can be rearranged to select either component:

[

R=\frac{1}{2\pi f_cC}

]

[

C=\frac{1}{2\pi f_cR}

]

Where:

1. (f_c) is in hertz
2. (R) is in ohms
3. (C) is in farads

At (f_c), the resistor magnitude equals the capacitor reactance:

[

R=|X_C|

]

The voltage magnitude of either filter is then:

[

|V_{out}|\approx0.707|V_{in}|

]

In decibels:

[

20\log_{10}(0.707)\approx-3.01\text{ dB}

]

The cutoff frequency is therefore not a point where the signal is completely blocked. It is a defined point within the transition between passband and stopband.

Worked RC Filter Example

Suppose a circuit uses:

1. (R=3.3,k\Omega)
2. (C=100,nF)

The calculated cutoff frequency is:

[

f_c=\frac{1}{2\pi(3300)(100\times10^{-9})}

]

[

f_c\approx482\text{ Hz}

]

For the low-pass filter, the magnitude response is:

[

|H_{LP}|=\frac{1}{\sqrt{1+(f/f_c)^2}}

]

For the high-pass filter:

[

|H_{HP}|=\frac{f/f_c}{\sqrt{1+(f/f_c)^2}}

]

The ideal unloaded results are:

The table demonstrates why “passes” and “blocks” should not be interpreted as absolute conditions. The unwanted frequency remains present but is reduced according to the frequency ratio and filter order.

Transfer Functions and Bode Plots

Using the Laplace variable (s), the transfer function of a first-order RC low-pass filter is:

[

H_{LP}(s)=\frac{1}{1+sRC}

]

For a first-order RC high-pass filter:

[

H_{HP}(s)=\frac{sRC}{1+sRC}

]

These equations describe three important operating regions.

On a logarithmic Bode magnitude plot, the low-pass response begins near 0 dB and turns downward after the corner. The high-pass response rises from strong low-frequency attenuation and approaches 0 dB above the corner.

For a first-order section, the asymptotic roll-off is 20 dB per decade, equivalent to approximately 6 dB per octave. Analog Devices notes that each pole contributes about −20 dB/decade, while a zero contributes approximately +20 dB/decade.

Filter Order and Roll-Off Rate

A higher-order filter can produce greater stopband attenuation without moving the cutoff unnecessarily far from the wanted signal.

These values describe the eventual slope, not necessarily the attenuation immediately beside the cutoff.

Higher-order filters can be constructed by cascading first- and second-order stages. However, directly connecting passive RC stages may cause loading, so the result is not always equivalent to multiplying two ideal independent responses. Buffers or a properly calculated active topology may be required.

Common response families include:

1. Butterworth: Maximally flat magnitude response in the passband, with moderate transition sharpness.
2. Bessel: Prioritizes nearly constant group delay and clean transient response, but rolls off more gradually.
3. Chebyshev: Accepts passband ripple to obtain a sharper transition and greater near-cutoff attenuation.

Texas Instruments’ active-filter guidance describes the same trade-off: Butterworth provides strong all-around performance, Bessel minimizes pulse distortion, and Chebyshev provides faster attenuation at the cost of ripple, overshoot and ringing.

Passive vs Active Filters

Both low-pass and high-pass responses can be implemented as passive or active filters.

An active filter is not automatically better. At low and moderate frequencies, it can provide gain, buffering and accurate higher-order responses without inductors. At RF frequencies, passive LC, transmission-line, ceramic, SAW or other specialized filters may be more appropriate.

Active-filter performance also depends on op-amp gain bandwidth, open-loop gain, slew rate, noise, output drive, supply voltage and input/output range.

How to Choose Between a Low-Pass and High-Pass Filter

Start by identifying the frequency components that must be retained.

After selecting the response type, define:

1. Highest and lowest wanted frequencies
2. Maximum acceptable passband loss
3. Frequency at which stopband attenuation is required
4. Required attenuation at that frequency
5. Acceptable phase shift, delay and transient distortion

This approach is more reliable than choosing a cutoff frequency before defining the passband and stopband requirements.

Practical RC Filter Design Considerations

Source and Load Impedance

For a low-pass filter, source resistance adds to the intended series resistance. A finite load also changes the voltage division and effective pole. Analyze the complete circuit using the source’s Thevenin resistance and the receiving stage’s input impedance.

As a practical first check, keeping the load impedance at least ten times greater than the filter’s relevant resistance reduces—but does not eliminate—loading error. Analog Devices similarly treats a high load impedance as an assumption behind the simplest passive RC analysis.

Component Tolerance and Temperature

Resistor and capacitor tolerances shift the real cutoff frequency. Temperature coefficient, aging and capacitor voltage dependence can introduce further error.

C0G/NP0 ceramic capacitors are useful where stability and linearity matter. Higher-capacitance ceramic dielectrics such as X7R can be appropriate in less critical circuits, but their effective capacitance may change with DC bias, package size and temperature.

Parasitic Effects

Very high resistance combined with a very small capacitor makes the circuit more sensitive to PCB leakage, input capacitance and stray capacitance. Very low resistance increases current demand and can load the source unnecessarily.

Texas Instruments recommends selecting values with practical parasitics, tolerance and power trade-offs in mind rather than treating every mathematically valid R/C combination as equivalent.

ADC Inputs

Many ADC inputs are not equivalent to a constant high resistance. Switched-capacitor input structures can draw brief charge pulses during acquisition. The filter must therefore be checked against acquisition time, input settling and driver stability.

A buffer can reduce the loading of a passive section and provide lower output impedance, but the amplifier and local RC network must follow the ADC manufacturer’s recommendations.

RF and High-Speed Circuits

At high frequencies, capacitor ESR and ESL, resistor parasitics, package dimensions, PCB traces, grounding and impedance matching may dominate the result. An ideal low-frequency RC equation is not sufficient for designing a precision RF filter.

Common Filter Design Mistakes

Avoid these frequent errors:

1. Treating the cutoff frequency as an on/off boundary
2. Assuming “high-pass” means removing high frequencies
3. Ignoring the location of the output node
4. Selecting (f_c) without defining stopband attenuation
5. Cascading passive stages without calculating loading
6. Ignoring resistor and capacitor tolerance
7. Using an op-amp with insufficient bandwidth or slew rate
8. Assuming every bypass capacitor forms a complete low-pass filter
9. Applying an ideal RC model directly to RF layouts

Frequently Asked Questions

What is the main difference between a low-pass and high-pass filter?

A low-pass filter retains frequencies below its cutoff and attenuates higher frequencies. A high-pass filter retains frequencies above the cutoff and attenuates lower frequencies.

Is a high-pass filter the same as a low-cut filter?

Yes. “High-pass” describes the frequencies being retained, while “low-cut” describes the frequencies being reduced.

Is a low-pass filter the same as a high-cut filter?

Yes. Both terms refer to a filter that attenuates higher-frequency components.

What happens at the cutoff frequency?

For a first-order RC filter, the output magnitude is approximately 70.7% of its passband level, equal to −3.01 dB. The signal is attenuated, not completely blocked.

Does a low-pass filter pass DC?

A basic RC low-pass filter passes DC because the capacitor behaves as an open circuit after steady state is reached. The exact DC gain still depends on source and load resistance.

Does a high-pass filter block DC?

Yes. The series capacitor in a basic high-pass filter prevents a steady-state DC component from reaching the output.

Can low-pass and high-pass filters use the same R and C values?

Yes. They will have the same ideal cutoff frequency, but the component arrangement and output measurement point determine whether the response is low-pass or high-pass.

Can low-pass and high-pass filters be combined?

Yes. A high-pass section with a lower cutoff and a low-pass section with a higher cutoff can form a band-pass filter. Stage interaction and loading must be included in the design.

What is the difference between first-order and second-order filters?

A first-order filter contains one pole and eventually rolls off at 20 dB/decade. A second-order filter contains two poles and can reach 40 dB/decade, although its behavior near cutoff also depends on damping and Q.

Which filter removes electrical noise?

It depends on the noise spectrum. Use a low-pass filter when unwanted noise lies above the wanted signal band. Use a high-pass filter when the interference is DC or lower in frequency than the wanted signal. Narrowband interference may require a notch or band-stop filter instead.

Conclusion

The choice between a low-pass and high-pass filter begins with one question: which part of the frequency spectrum contains the useful information?

A low-pass filter is appropriate when low-frequency or slowly changing content must be preserved. A high-pass filter is appropriate when DC and slow changes must be removed while faster signal components are retained.

The equation (f_c=1/(2\pi RC)) provides a useful starting point, but a reliable design must also consider required stopband attenuation, source and load impedance, component tolerance, filter order, phase response and circuit non-idealities. For demanding ADC, audio, industrial and RF systems, these practical factors often determine whether a filter works only in calculation or also performs correctly in hardware.