DC-Blocking Capacitors: Selection and Design Guide

DC-blocking capacitors are connected in series with a signal path to prevent steady DC current from passing while coupling the required AC signal between circuit stages. This article explains their operation, application limits and selection for design engineers and component engineers working with audio, sensor and RF circuits, and distinguishes signal coupling from capacitor functions in power converters.

Knowles A-Series RF Attenuators Target Compact 20 GHz Designs

Electrification Raises Demands on DC-Link Capacitors

Knowles Cornell Dubilier 105C Flatpack Aluminum Electrolytic Capacitors Target Low-Profile High-Density Power Designs

The article develops the original Knowles Precision Devices explanation into a practical selection guide covering circuit bias, low-frequency attenuation, effective capacitance, parasitic impedance and validation. [1]

Key Takeaways

A DC-blocking capacitor couples changing signals but blocks steady DC in the ideal model. Real components also have leakage, losses and parasitic inductance.
The receiving circuit determines the output DC reference. A coupling capacitor does not automatically center a signal on zero volts, and amplifier inputs need a DC bias-current return path.
Select capacitance from the lowest required signal frequency and permitted attenuation, using the resistance seen by the capacitor rather than assuming the load resistance alone always applies.
For high-permittivity ceramic capacitors, use effective capacitance at the operating bias and temperature. Nominal capacitance and initial tolerance alone do not establish the actual cutoff frequency.
For RF applications, verify the mounted component’s impedance and transmission response across the entire band. More capacitance does not automatically provide better broadband performance.
Series signal coupling, supply bypassing and DC-link energy storage are different functions. Their capacitor requirements and circuit positions must not be confused.

What Systems Rely on Stable Waveforms?

AC coupling is useful when a changing signal must pass between circuit stages without transferring their DC operating levels. It is not mandatory in every AC or RF system: a DC-coupled design may be necessary when the wanted information includes DC or very low frequencies. [1, 2]

Audio amplifiers can use coupling capacitors between stages or at an output where the architecture requires separation of DC operating levels.
RF systems, including radio, broadcast television and microwave communications equipment, can use DC blocks between differently biased stages or at measurement interfaces.
Communication connectors and transmission interfaces may incorporate AC coupling where the interface permits it. The connector itself does not establish whether coupling is required.
Sensor interfaces can use AC coupling for changing signals, but it removes DC information and attenuates sufficiently slow changes.
RF filters and matching networks may include capacitors that also provide DC separation, provided their signal-band response meets the design requirements.
Power converters and power amplifiers can contain series capacitors with topology-specific functions. Their bypass, DC-link and energy-transfer capacitors must be assessed separately from small-signal coupling capacitors.

Why Are DC-Blocking Capacitors Necessary?

A circuit stage may intentionally operate around a nonzero DC bias, while the following stage requires a different reference level. A coupling capacitor separates those operating points so that the receiving stage can establish its own bias while accepting the changing signal. [2] Unwanted DC offsets can also arise from amplifier input offsets, supply-related operating-point shifts, component imbalance, faults or electrochemical potentials at a sensor or contact interface. These mechanisms are application-dependent; they should not all be treated as a single kind of supply noise. An unwanted offset can reduce available signal headroom or drive a high-gain amplifier toward saturation. [1, 2] AC coupling addresses the transfer of DC between stages. It does not remove supply ripple indiscriminately, correct an amplifier’s internal offset, guarantee low distortion or protect against every fault. Frequencies within the coupling network’s passband still pass, including unwanted interference at those frequencies.

How Blocking Capacitors Remove Unwanted DC Line Levels

A capacitor has conductive electrodes separated by a dielectric. Charge accumulates on the electrodes and energy is stored in the electric field; ordinary signal coupling does not require charge carriers to cross an ideal insulating dielectric. Figure 1 retains the original physical illustration. [1, 3]

Original illustration of capacitor electrode and dielectric arrangement and capacitor symbol — Figure 1. Capacitor physical diagram and symbol. Original image credited to Electronics Tutorials.

For an ideal linear capacitor, terminal current i is related to capacitance C and the rate of change of capacitor voltage v_C by Eq. (1), where t is time. At a constant voltage after charging, the derivative is zero and the ideal steady current is zero. A changing voltage produces charging and discharging current in the external circuit.

i(t) = C\frac{\mathrm{d}v_{C}(t)}{\mathrm{d}t} \tag{1}

A real capacitor has finite insulation resistance and therefore some leakage current. A sudden DC change also produces a transient response: blocking steady DC does not mean that a DC step, connection event or power-up disturbance produces no output. [2, 3]

The receiving circuit sets the DC reference

Figure 2 shows the original voltage-divider-biased amplifier example. The output coupling capacitor C2 separates the amplifier’s DC operating point from the load; the receiving-side DC path determines the load’s reference voltage. An output centered on zero volts requires an appropriate reference to ground, rather than following automatically from the presence of C2. [4]

Original voltage-divider-biased amplifier diagram identifying output coupling capacitor C2 — Figure 2. Voltage-divider-biased amplifier with output coupling capacitor C2. Source: Electronics Hub, reference [4].

Do not leave an AC-coupled amplifier input floating. Its input bias current needs a DC return through a resistor, bias network or another suitable circuit path. Without that path, the coupling capacitor can charge until the input common-mode range is exceeded or the amplifier output saturates. The return may connect to ground in a suitable dual-supply design or to a defined bias reference in a single-supply design. [2]

RF and coaxial DC blocks

In coaxial systems, an inner DC block interrupts DC continuity in the center conductor, an outer DC block interrupts DC continuity in the outer conductor, and a combined block interrupts both. These arrangements are not interchangeable: the choice must account for the required signal return, shielding and grounding arrangement. A signal-path DC block should not be treated as a general safety-isolation solution. [1]

How to Select DC-Blocking Capacitors

Begin with the required passband, allowable signal loss and circuit impedances. A first-order RC model establishes the low-frequency requirement, but the selected physical component must also meet bias, voltage, loss and high-frequency requirements. [3, 5, 6]

Start with the RC high-pass model

Figure 3 demonstrates high-pass circuit, equation and response illustration. The capacitor is in series with the signal, and the output is taken across the receiving resistance. The resistor is a shunt element at the output, not a component connected in parallel with the series capacitor. [3]

Original composite illustration of a passive RC high-pass filter, its cutoff relationship and frequency-response plot — Figure 3. Passive RC high-pass filter illustration, including the circuit, cutoff relationship and response plot.

For a single series coupling capacitor feeding a resistive load, let R_S be the source’s Thevenin resistance and R_L the receiving resistance to an AC-stiff reference. The resistance seen between the capacitor terminals with independent sources set to zero is R_eq = R_S + R_L. With capacitance C in F and resistance in Ω, the cutoff frequency f_c in Hz follows Eq. (2). For negligible source resistance, R_eq reduces to R_L. [3]

f_{c} = \frac{1}{2\pi R_{\mathrm{eq}}C}, \qquad R_{\mathrm{eq}} = R_{S} + R_{L} \tag{2}

The resistive divider also establishes a passband voltage gain G_∞ = R_L/(R_S + R_L). Eq. (3) gives the magnitude of voltage gain H(f), normalised to that passband gain, at signal frequency f. This distinction prevents coupling loss from being confused with ordinary source/load division.

\frac{|H(f)|}{G_{\infty}} = \frac{1}{\sqrt{1+(f_{c}/f)^{2}}}, \qquad G_{\infty} = \frac{R_{L}}{R_{S}+R_{L}} \tag{3}

At f_c, voltage amplitude is approximately 0.707 of its passband value, corresponding to approximately −3.01 dB. The amplitude response falls at approximately 20 dB per decade sufficiently below cutoff. It is the gain that falls, not the frequency itself. These statements apply to the first-order response, not to every complete amplifier or RF network. [3]

Size for allowable attenuation, not just cutoff

Placing the cutoff at the lowest wanted frequency accepts approximately 3 dB of attenuation there, which may be excessive. Let A_max be the maximum permitted positive attenuation in dB relative to passband at the lowest required frequency f_min. Rearranging Eq. (3) gives the minimum effective capacitance C_min in Eq. (4).

C_{\min} = \frac{1}{2\pi f_{\min}R_{\mathrm{eq}}\sqrt{10^{A_{\max}/10}-1}} \tag{4}

As an illustrative design calculation, assume negligible source resistance, a 10 kΩ receiving resistance, a lowest wanted frequency of 20 Hz and maximum coupling attenuation of 0.1 dB. Eq. (4) gives a minimum effective capacitance of approximately 5.21 µF. An effective capacitance of 5.6 µF gives a cutoff of approximately 2.84 Hz and attenuation of approximately 0.087 dB at 20 Hz.

The 5.6 µF figure is an operating capacitance used for the calculation, not a claim that every nominal 5.6 µF component meets the requirement. Include resistance tolerance and the capacitor’s minimum effective capacitance under the intended conditions before selecting an ordering code.

Account for bias, temperature and dielectric behavior

Multilayer ceramic capacitors can provide compact coupling solutions, but dielectric choice matters. High-permittivity types such as X5R and X7R can lose capacitance under DC bias. C0G temperature-compensating ceramics do not exhibit that ferroelectric DC-bias capacitance reduction. Use the manufacturer’s data for the exact part and operating conditions, rather than applying one assumed reduction factor to every MLCC. [5]

Compare capacitance tolerance, temperature behavior, DC-bias dependence, leakage and loss alongside size and cost. Where available capacitance, linearity or voltage requirements make a ceramic part unsuitable, evaluate other technologies such as film or an appropriately specified electrolytic capacitor. A polarized component requires an operating waveform and polarity permitted by its specification. [5, 7]

Check the upper-frequency limit

Real capacitors have equivalent series resistance, ESR, and equivalent series inductance, L_ESL. Neglecting leakage in the relevant AC range, the simple lumped series model gives impedance Z in Eq. (5), where j is the imaginary unit and ω = 2πf is angular frequency in rad/s. [6]

Z(\omega) = \mathrm{ESR} + j\left(\omega L_{\mathrm{ESL}}-\frac{1}{\omega C}\right), \qquad \omega = 2\pi f \tag{5}

In this model, capacitive and inductive reactance cancel at self-resonance, and the response becomes predominantly inductive above it. This does not make self-resonance a universal pass/fail bandwidth boundary: the total series impedance and resulting transmission response determine whether the installed part meets the specification. Terminations, pads and interconnects contribute to the mounted response. [6]

For RF coupling, compare insertion loss and return loss across the intended band using suitable manufacturer models or measured S-parameters. Check the reference impedance, fixture and mounting assumptions. Broadband components should be selected from their verified response rather than from nominal capacitance alone.

Consider startup and digital signals

Increasing capacitance lowers the high-pass cutoff but also increases the RC time constant. Check startup, reconnection and bias-settling behavior, especially where a transient could overload a following amplifier. A satisfactory steady-state sine-wave response does not establish acceptable behavior during these events. [2, 3]

For digital links, low-frequency content depends on the signal pattern and coding, not simply the bit rate. AC coupling can produce droop or baseline movement during sustained levels. Follow the particular transmitter, receiver and interface requirements rather than reusing an audio-coupling calculation as a digital-link specification. [3]

DC Blocking Versus Bypassing in Power Converters

Capacitors in power equipment are not all DC-blocking capacitors. Circuit position and current path determine whether a capacitor performs series coupling, supply bypassing, DC-link buffering or another topology-specific function. [4, 8]

A series signal-coupling capacitor interrupts the steady DC path between stages. A bypass capacitor is connected across a supply or another node and provides a low-impedance path for changing current; it does not interrupt DC delivery along the main power path. A DC-link capacitor supports the DC bus and its ripple-current demands. [4]

Figure 4 retains the original converter illustration as a contrast example. Its cited TI reference design is a bidirectional, isolated dual active bridge DC/DC converter for electric vehicle charging, not simply a DC/DC step-down converter. Capacitors around that converter must be identified from their actual connections and design documentation; their presence on either side is not evidence of series DC blocking. [8]

Original converter capacitor configuration illustration associated with the TI TIDA-010054 reference design — Figure 4. Original converter capacitor configuration illustration. Bypass and DC-link functions must be distinguished from series DC blocking. Cited design source: Texas Instruments TIDA-010054, reference [8].

Design and Selection Checklist

Approve the component against its installed function and operating conditions, not just its nominal capacitance and voltage marking. The following checks connect the circuit calculation to a purchasable and verifiable component. [2, 5, 6, 7]

Define the function and scope: audio or sensor coupling, RF DC blocking, digital-link coupling, or a specific power-converter role. Record whether DC information must be preserved.
Record the minimum and maximum signal frequencies, permitted attenuation, source impedance, receiving impedance and receiving-side DC reference. Include additional coupling stages where present.
Determine minimum effective capacitance, including tolerance and applicable bias, temperature and ageing behavior. Use the exact part’s characteristic data.
Check the voltage across the capacitor, including the difference between stage biases, the superimposed AC waveform and startup or fault conditions. Verify permitted polarity and separate DC, AC and pulse limits.
Compare leakage or insulation resistance for high-impedance circuits, and ESR, impedance, self-resonance and mounted response where signal loss or bandwidth matters.
Check heating and current capability for RF power or power-converter service. Do not transfer a small-signal coupling approval to a high-current application.
Confirm a DC bias-current return path at every AC-coupled active input. Check amplifier common-mode range, startup response, settling and overload recovery.
For catalog parts, approve the complete ordering code, dielectric, tolerance, package, termination, assembly requirements and required qualification evidence. Treat a different package or dielectric as an engineering change, not an automatically equivalent substitute.
For custom parts, agree the waveform, frequency band, bias, electrical and thermal limits, mounting geometry, acceptance tests and change-control requirements with the supplier. Low ESR or a custom package alone does not establish application compliance.
Where the circuit involves mains or a safety barrier, establish the applicable safety requirements separately. Do not assume that an ordinary DC-blocking capacitor or its DC voltage rating establishes safety isolation.
Validate the populated circuit. Use frequency-response measurements for low-frequency coupling and appropriate network measurements for RF; document bias, temperature, loading and fixture conditions.

Conclusion

A DC-blocking capacitor separates DC operating points while coupling a defined range of changing signals. Its performance depends on the entire circuit: the receiving-side bias path establishes the DC reference, the surrounding impedances determine low-frequency attenuation, and component losses, inductance and mounting determine the upper-frequency response.

Calculate the minimum effective capacitance from allowable signal loss, then verify the selected component’s bias behavior, voltage stress and installed response. Keep signal coupling distinct from supply bypassing and DC-link service, and validate startup as well as steady-state operation.

FAQ

These answers summarise the main operating principles and selection limits for signal and RF coupling.

What does a DC-blocking capacitor do?

A DC-blocking capacitor is connected in series to separate the DC operating levels of circuit stages while coupling changing signals. It blocks steady DC in the ideal model but passes charging transients. The useful signal band depends on the surrounding impedances and the real component’s capacitance, losses and inductance. It is also called an AC-coupling capacitor.

How do I calculate a DC-blocking capacitor value?

Start with the lowest required frequency, allowable attenuation and resistance seen by the capacitor. For a simple resistive source and load, that resistance is their sum. Calculate the minimum effective capacitance from the first-order high-pass response, then allow for component and resistance variation. Placing cutoff at the lowest wanted frequency accepts approximately 3 dB of attenuation there.

Does a coupling capacitor center the output on zero volts?

Not automatically. The receiving circuit’s DC reference and return path establish the output operating level. A suitable connection to ground can produce a ground-referenced signal, while a single-supply amplifier may require a nonzero bias. Leaving an active input floating can allow input bias current to charge the capacitor until the amplifier exceeds its operating range.

Can I use an X7R MLCC for DC blocking?

An X7R MLCC can be suitable if its effective capacitance and other characteristics meet the circuit requirements. Its capacitance can decrease under DC bias, so nominal value alone does not establish the cutoff frequency. Compare the exact part’s operating data with alternatives, including C0G where appropriate, and verify voltage stress, losses and installed frequency response.

How is a DC-blocking capacitor different from a bypass capacitor?

A DC-blocking capacitor is placed in series with a signal path to separate DC operating levels. A bypass capacitor is connected across a supply or another node to provide a path for changing current. The two functions have different current paths and selection requirements. Capacitors around a DC/DC converter should therefore be classified from their connections, not their proximity to the converter.

How to Select and Validate a DC-Blocking Capacitor

Use this procedure to turn the signal requirements into a component specification and an in-circuit verification plan.

Select a series coupling capacitor from signal requirements and verify its installed performance.

Define the application and signal band
Identify the signal-coupling function, lowest and highest required frequencies, permitted attenuation, source impedance and receiving impedance. Confirm that removing DC information is acceptable.
Establish bias and calculate capacitance
Provide a receiving-side DC reference and a return path for input bias current. Calculate minimum effective capacitance from the high-pass response and allowed attenuation, including the resistance seen by the capacitor.
Select the component and ratings
Compare effective capacitance, tolerance, dielectric, leakage, voltage limits, polarity, losses and package. Use exact-part bias and temperature data. Confirm current capability and required qualification evidence for the application.
Check mounted frequency response
Include capacitor parasitics and the intended mounting geometry. For RF coupling, assess insertion loss and return loss across the full band using suitable models or measurements rather than nominal capacitance alone.
Validate steady-state and transient operation
Measure the populated circuit under representative bias, loading and temperature. Confirm passband attenuation, startup, reconnection and settling behavior. For digital signals, check relevant patterns and baseline movement, then document the approved component and test conditions.

References

The following sources support the operating principles, component limitations and application distinctions discussed above. Undated web documents were accessed in November 2026.

Jordan Yates, What Are DC-Blocking Capacitors, and Why Are They Important?, Knowles Precision Devices, 2025. Original article basis.
Charles Kitchin, AN-937: Designing Amplifier Circuits: How to Avoid Common Problems, Analog Devices, publication year not stated on the accessed page.
Electronics Tutorials, Passive High Pass Filter, Electronics Tutorials, publication year not stated on the accessed page.
Electronics Hub, Bypass Capacitor Basics: Keeping Your Circuits Clean, Electronics Hub, publication year not stated on the accessed page.
Murata Manufacturing, Does the Capacitance Change When a DC Voltage Is Applied to Ceramic Capacitors?, Capacitors FAQ, Murata Manufacturing, publication year not stated.
Luke Murphy, Choosing Blocking Capacitors: It’s More Than Just Values, Knowles Precision Devices, 2018. Used for real-component impedance and mounting limitations, not as a series-coupling product recommendation.
Passive Components Blog, Capacitor Symbols, Passive Components Blog, 2026. Background on polarized and non-polar component distinctions.
Texas Instruments, TIDA-010054: Bidirectional, Dual Active Bridge Reference Design for Level 3 Electric Vehicle Charging Stations, Texas Instruments, design guide Revision F listed in 2026.