🔊 Signal-Path Design Tool

Coupling & Decoupling Capacitor Calculator — AC Coupling, Input/Output Cap, Emitter & Bypass Cap Size

Every series coupling cap and every parallel bypass cap reduces to one formula: C = 1/(2πfR). Enter the resistance and the corner frequency, get the exact capacitance, the next standard value to buy, and the corner you actually achieve with it.

Best for: audio stages, tube preamps, digital railsTotal procedure: ≈ 15 minDifficulty: beginner-friendly

Coupling & Decoupling Capacitor Calculator

C = 1/(2πfR) — size AC coupling caps and emitter/cathode bypass caps live.

Calculation mode

Source Resistance R1

Load Resistance R2

Cutoff Frequency fc (Hz)

Quick presets

Required capacitance C

3.98µF

Nearest standard (round up): 4.7 µF

Achieved fc with standard value: 16.9 Hz

Recommended type: Electrolytic (polarized)

Formula: C = 1/(2π × 20 Hz × 2 kΩ)

💡 Rule of thumb. A coupling capacitor blocks DC and passes AC. The rule of thumb: set fc one decade below your lowest frequency of interest. For audio (20 Hz–20 kHz), set fc = 2 Hz for coupling caps.

The three jobs

Coupling vs Decoupling vs Bypass — Three Jobs, One Formula

Capacitors that deal with signals and supplies divide into three overlapping jobs, and beginners mix them up constantly. A coupling capacitor sits in series with the signal path. It charges up to the DC bias difference between two stages and then simply refuses to pass that DC any further, while audio-frequency AC sails across it almost untouched. That is how a preamp running at 2 V DC bias hands a clean 0 dBV signal to a power amp running at a completely different bias point.

A decoupling capacitor (also called a rail bypass) does the opposite geometrically: it sits in parallel with the supply, from the power pin to ground. Its job is to be invisible at DC and nearly a short circuit at the frequencies where digital edges and switching regulators dump noise, so the trash takes the easy path to ground instead of wandering through the IC. It also acts as a tiny local battery that supplies the sudden current spikes a chip demands, saving the long wire back to the supply from drooping.

A bypass capacitor is the umbrella term: any capacitor that gives an unwanted signal an easier parallel path. Rail decoupling is bypassing; so is the emitter bypass on a transistor, which shorts out the emitter resistor for AC to restore gain while the resistor keeps doing its DC bias work. What surprises most newcomers is that all three jobs — series AC coupling, parallel rail bypass, parallel emitter/cathode bypass — are sized from the same relationship, C = 1/(2πfR), just with R found differently in each case. The two H3 blocks below spell out that difference.

Printed circuit board close-up — coupling and bypass capacitors share the board with the transistors they serve

Coupling and bypass caps on an audio stagePhoto: Unsplash

What the coupling capacitor calculator computes

In coupling mode, this tool treats the capacitor as the series element of an RC high-pass filter. The resistance it uses is the sum R_source + R_load, because the signal current leaving the source must cross both resistances and the capacitor on its way to the load. Given your target corner frequency, it returns the ideal capacitance, snaps it up to the next standard catalog value, and then tells you the corner frequency you actually get with the part you can really buy — because 3.98 µF exists only in mathematics, while 4.7 µF exists in the drawer.

What the decoupling capacitor calculator and bypass capacitor calculator compute

In bypass mode, the same formula runs in low-pass fashion with a single parallel resistance: the emitter resistor in parallel with r′e, the cathode resistor in parallel with the tube's internal resistance, or the rail impedance you are decoupling. The output is the capacitance that keeps that node at a low impedance down to your chosen frequency, plus the standard value to order. For pure digital supply decoupling, the answer is usually prescribed — 0.1 µF per power pin — and the section further down explains why that value won.

The design procedure

How to Calculate Coupling Capacitor Value — the 5-Step Design Procedure

Follow these five steps in order for any coupling or bypass capacitor — an input coupling cap, an output cap feeding a speaker, an emitter bypass, a cathode bypass or a rail filter. The order matters: each step consumes the output of the one before it, so skipping ahead means recomputing everything. Total time from schematic to shopping list: about fifteen minutes.

Identify the resistance the capacitor sees

≈ 5 min Critical Step

Everything in this design procedure depends on one number: R, the total resistance in the capacitor's view. Get R wrong and every later step inherits the error, which is why this is the critical step of the whole procedure.

For a series coupling capacitor, the AC signal path runs from the source, through the capacitor, into the load — so the capacitor sees the sum: R = R_source + R_load. A DAC with a 100 Ω output impedance driving a 1 MΩ amplifier input gives R ≈ 1 000 100 Ω, which rounds to 1 MΩ for practical purposes. A 1 kΩ pot wiper feeding a 1 kΩ bias network gives R = 2 kΩ.

For a bypass capacitor — the emitter bypass on a transistor or the cathode bypass on a tube — the capacitor sits in parallel with a resistor, so the math flips to a parallel combination: R = R_bypass || R_internal. On a bipolar transistor, the internal part is the dynamic emitter resistance r′e ≈ 25 mV ÷ Ic (about 25 Ω at 1 mA of collector current). On a tube stage the same trick applies to the cathode resistor R_K in parallel with the tube's internal cathode resistance, roughly 1/gm.

Pitfall: Forgetting the source output impedance. A signal generator, pot wiper or previous stage adds real ohms in series — ignoring it undersizes the capacitor.
Pitfall: Using R_E alone for an emitter bypass. Ignoring r′e can overstate the true resistance by 10–40× and produce a capacitor far too small to do the job.

Choose fc one decade below your lowest signal frequency

≈ 2 min

The cutoff frequency fc is where the capacitor's effect becomes visible: a coupling cap passes −3 dB (about 71% of the voltage) at fc and rolls off steeply below it. If you set fc exactly at your lowest wanted frequency, that frequency is already down 3 dB and phase-shifted 45° — audible as thin, late bass.

The professional fix is margin: place fc one decade (ten times) below the lowest frequency you care about. Full-range audio starts at 20 Hz, so design for fc = 2 Hz. A voice-band intercom that only must pass 300 Hz and up can relax to fc = 30 Hz. A guitar pedal focused on 100 Hz fundamentals can use fc = 10 Hz.

For bypass capacitors on emitter and cathode resistors, the same fc is the frequency below which stage gain starts falling — so it obeys the same decade rule. Rail decoupling is different: there you are shunting noise, and the target is simply to keep impedance low at the frequencies your circuit switches at.

Compute C = 1/(2πfR)

≈ 1 min with the calculator

One formula covers both modes of this calculator. Rearranging the reactance relation Xc = 1/(2πfC) and setting |Xc| = R at the corner gives C = 1/(2π·fc·R). Feed it the R from Step 1 and the fc from Step 2.

Worked directly: R = 2 kΩ (1 kΩ source + 1 kΩ load) at fc = 20 Hz gives C = 1/(2π × 20 × 2000) ≈ 3.98 µF. R = 1 kΩ at fc = 100 Hz gives C ≈ 1.59 µF. The calculator above performs this arithmetic live as you type, including the unit conversion — the most common source of hand-calculation errors.

The same R × C product is the RC time constant τ, so this step and the time-constant view of the circuit are two faces of one design. If you can state τ = R × C, you can state fc = 1/(2πτ).

Round up to the next standard value and pick the right type

≈ 2 min

Capacitors come in standard values — the 1, 2.2, 4.7, 10, 22, 47, 100, 220, 470, 1000 µF ladder and its nF/pF siblings — and you always round up, never down. Rounding 3.98 µF down to 2.2 µF pushes the achieved corner up to 36 Hz, well above the 20 Hz target; rounding up to 4.7 µF keeps you safe at 16.9 Hz.

Then choose the dielectric to match the job. Values of 1 µF and above: aluminum electrolytic (bipolar types where the signal can reverse-bias the cap, such as speaker outputs). Signal-critical values below 1 µF: film (polypropylene/polyester) or C0G/NP0 ceramic. Rail decoupling: X7R ceramic, typically 0.1 µF per power pin. Electrolytics should be rated at least twice the working voltage across them.

Verify — recompute the achieved fc with the standard value

≈ 1 min Verification

The standard value you buy is not the value you computed, so the last step closes the loop: recompute the corner frequency using the real capacitor. With C = 4.7 µF and R = 2 kΩ, the achieved fc = 1/(2π × 4.7 µF × 2 kΩ) ≈ 16.9 Hz — comfortably below the 20 Hz target, so the design passes.

If the achieved corner lands above your target (common after rounding up a small capacitor), climb one more step up the standard ladder and re-verify. It takes ten seconds in the calculator and prevents the classic 'sounded fine on the bench, thin in the room' outcome.

As an extra cross-check, drop the capacitor into the reactance calculator and confirm that Xc at your lowest signal frequency is a small fraction of R — the smaller, the more solidly the signal passes.

💡 Tip. When a datasheet quotes the input impedance of an amplifier as "1 MΩ typical", treat it as 1 MΩ in Step 1, but remember that a volume pot feeding that input appears in parallel with it — a 100 kΩ pot drags the effective load resistance down to about 90.9 kΩ, which changes the answer by a factor of eleven. Recompute after every wiring decision.

⚠️ Warning. If Step 1 stalls you for more than a few minutes — an unknown source impedance, a transistor without a quoted operating point — stop and measure rather than guessing. Feed a 1 kHz sine through the stage and measure the gain drop with and without the capacitor in circuit; or assume the pessimistic (smaller) R so the error lands on the safe side. Guessing large is how undersized emitter bypasses sneak into production.

❌ Don't. Never round the computed capacitance down to the nearest standard value to "save cost". A coupling cap 20% small pushes your corner frequency up by 25%, and an emitter bypass that is half the needed size silences the gain boost you added the transistor for. Rounding up is nearly always free audibly; rounding down is not.

Interactive Design Checklist — Run It Before You Order Parts

0 of 5 · 0%

I identified the resistance the capacitor sees (series sum or parallel combination)
I set fc at least one decade below my lowest signal frequency
I computed C = 1/(2πfR) with the calculator
I rounded up to the next standard value and chose the right capacitor type
I recomputed the achieved fc with the standard value and confirmed the margin

Numbers you can reuse

AC Coupling Capacitor Calculation — Worked Audio Examples

Three examples cover most real requests. Each one follows the five-step procedure exactly: find R, choose fc, compute C, round up, verify. Run each case through the calculator at the top of the page — the presets load them in one click.

Input Coupling Capacitor Calculator Example — 1 MΩ Input at 10 Hz

A tube preamp input presents roughly 1 MΩ to the source, and the source (a measurement generator or a line-level DAC) has an output impedance near zero, so R ≈ 1 MΩ. Choosing fc = 10 Hz gives C = 1/(2π × 10 × 1 000 000) ≈ 15.92 nF. The next standard value up is 22 nF — a cheap, stable film capacitor. Verify: with 22 nF the achieved corner is 1/(2π × 22 nF × 1 MΩ) ≈ 7.2 Hz, safely below target.

C = 1/(2π · 10 Hz · 1 MΩ) ≈ 15.92 nF → use 22 nF film → achieved fc ≈ 7.2 Hz

Apply the stricter decade rule for full-fidelity audio — fc = 2 Hz — and the same 1 MΩ load wants C ≈ 79.6 nF, which lands on the ubiquitous 100 nF value. That is precisely why 100 nF film caps appear at the input of so many commercial amplifier boards: one decade of margin costs about one cent.

Output Coupling Capacitor Calculation — 8 Ω Speaker at 20 Hz

Feeding a speaker directly through a coupling cap (single-supply portable amps do this constantly) is where small values stop working. The speaker is 8 Ω, so even at an ambitious fc = 20 Hz the required capacitance is C = 1/(2π × 20 × 8) ≈ 994.7 µF — the 1000 µF standard value, an aluminum electrolytic rated for the full signal swing. Verify: 1000 µF into 8 Ω achieves fc ≈ 19.9 Hz. Notice the brutal scaling: resistance ten thousand times smaller means capacitance ten thousand times larger. This is also why the cap must be bipolar or correctly oriented — it sees the entire output waveform across it.

C = 1/(2π · 20 Hz · 8 Ω) ≈ 994.7 µF → use 1000 µF electrolytic → achieved fc ≈ 19.9 Hz

Practical designs usually step up to 2200 µF, because electrolytic tolerance runs −20% or worse and the bass damping improves as the corner drops. The charge inrush at power-on — the cap charging through the speaker — is the trade-off, which is why output muting circuits exist.

Emitter Bypass Capacitor Calculator Example — 1 kΩ Emitter Resistor at 100 Hz

A common-emitter stage with a 1 kΩ emitter resistor, biased at low collector current, wants its full gain back down to 100 Hz. In the naive view — treating the bypass as seeing the whole 1 kΩ — C = 1/(2π × 100 × 1000) ≈ 1.59 µF, rounding up to 2.2 µF. But the honest calculation in Step 1 includes the transistor's internal emitter resistance r′e ≈ 25 mV/Ic, which sits in parallel with the resistor and can pull the effective R down by a factor of ten or more. That is why real emitter bypasses land at 10 µF, 47 µF or even 100 µF even when the paper calculation says a couple of microfarads.

C = 1/(2π · 100 Hz · 1 kΩ) ≈ 1.59 µF → use 2.2 µF → achieved fc ≈ 72.3 Hz (before r′e correction)

The same logic explains the reference table's emitter row: the "1 kΩ" entry produces 10 µF once the transistor's internal resistance and electrolytic tolerance are priced in. Compute the naive value first, then apply judgment — or simply enter the parallel combination if you know the operating point.

How to Calculate Bypass Capacitor Value for Any Parallel Resistor

Whatever the parallel element — emitter resistor, cathode resistor, supply rail impedance, the bias network of a JFET — the procedure never changes. Collapse everything the capacitor sees into one equivalent resistance R (series elements add; parallel elements combine as R₁R₂/(R₁+R₂)), choose the frequency below which the node must stay low-impedance, and compute C = 1/(2πfR). Then round up and verify. The only domain-specific judgment is identifying R correctly, which is exactly why this page pins a red badge on Step 1. For a second opinion on any of these numbers, the same R and C can be pushed through the capacitive reactance calculator — at the corner frequency, |Xc| should equal R by definition.

🧯 Tube stage

Cathode Bypass Capacitor Calculator Math for Tube Preamps

Tube stages are where the bypass capacitor earns its keep most audibly. The cathode resistor R_K biases the tube, but every milliamp of signal current through it also creates degenerative feedback that slashes gain — often by 3 to 6 dB. The cathode bypass capacitor shunts R_K for audio, restoring gain, while leaving the DC bias point exactly where the designer put it.

The resistance the capacitor sees is R_K in parallel with the tube's internal cathode resistance, approximately 1/gm. Take a small-signal triode stage with a 1 kΩ cathode resistor whose internal resistance works out near 4.3 kΩ: the parallel combination is 1000 × 4300 / 5300 ≈ 811 Ω. Designing for full gain down to 20 Hz:

C = 1/(2π · 20 Hz · 811 Ω) ≈ 9.81 µF → use 10 µF electrolytic → achieved fc ≈ 19.6 Hz

A 10 µF, 25 V or 50 V electrolytic with its positive terminal to the cathode (the cathode sits at a positive DC potential relative to ground) is the canonical part. Two practical notes: first, because tube rolling changes gm stage to stage, builders often fit 22 µF or 47 µF to keep the corner low regardless of which tube is plugged in. Second, if you skip the bypass entirely, the stage still works — you simply trade gain for linearity, which is itself a legitimate design choice known as degeneration.

✅ Cross-check. The cathode bypass and the coupling capacitors around a tube stage interact: the cap that couples the previous stage into the grid also sees this stage's 1 MΩ grid resistor, and the τ = RC product involved is the same one explored by the RC time constant calculator. Compute fc from τ, or τ from fc — they are reciprocals: fc = 1/(2πτ).

Reference

Audio Stage — Standard Coupling & Bypass Capacitor Values

Four rows cover the overwhelming majority of audio-stage capacitor requests. Every value can be reproduced with the calculator above — enter the impedance and corner frequency from each row and compare. Notice how the two coupling rows differ by five orders of magnitude in capacitance purely because the impedances differ.

Function	Typical Impedance	fc (Hz)	Recommended Cap	Capacitor Type
Input Coupling (Amp)	1 MΩ	10	10 µF	Electrolytic
Output Coupling (Speaker 8 Ω)	8 Ω	20	1000 µF	Electrolytic
Emitter Bypass (CE Amp)	1 kΩ	100	10 µF	Electrolytic
Power Supply Bypass	Rail	10k	0.1 µF	Ceramic X7R

Reading the table critically: the 1000 µF output coupling value reflects the 8 Ω speaker's tiny impedance; the 0.1 µF power-supply bypass row is not derived from a 20 Hz corner at all but from the high-frequency impedance target described in the digital section below; and the emitter row's 10 µF includes the r′e correction discussed in the worked examples. Treat the table as a starting point, then run your own numbers — tolerance, temperature and layout all nudge the final value.

🔌 Digital rails

Digital Decoupling — Why 0.1 µF Decoupling Wins

For digital ICs the calculation changes character. You are no longer chasing a −3 dB corner for a signal band; you are keeping the supply impedance low from roughly 1 MHz to several hundred MHz, where logic edge energy lives. The famous answer — 0.1 µF ceramic per power pin — wins because it sits at the intersection of several constraints: large enough to deliver nanosecond-scale charge, small enough that its self-resonant frequency (with real PCB inductance of a few nanohenries) stays above the frequencies that matter, and cheap enough to sprinkle on every pin.

Placement matters as much as value. A decoupling cap four centimeters from the pin adds roughly 4 nH of loop inductance per 4 mm of trace — enough to turn your 0.1 µF into an open circuit above 50 MHz. Keep the loop short: capacitor next to the pin, via straight down to the ground plane. Layer a second tier of bulk capacitance — 10 µF ceramic or 47–100 µF electrolytic per board region — to resupply the local caps between switching bursts.

⚡ Placement rule. For digital ICs, place 0.1 µF ceramic decoupling caps within 5 mm of every power pin. Add one 10 µF bulk cap per 8 ICs on the same rail. This is non-negotiable for clean power delivery.

Breadboard power rails wired with bypass capacitors feeding the IC supply pins

0.1 µF decoupling caps at every power pinPhoto: Unsplash

Where the C = 1/(2πfR) view still helps

Even in the digital world the same formula explains the bulk caps. A voltage regulator and its wiring look like a few tens of milliohms; to keep the rail impedance flat down to 10 kHz, a region drawing 200 mA of bursty current wants on the order of hundreds of microfarads of bulk capacitance — which is why buck converters specify 22–100 µF of output ceramic plus electrolytic. The formula also explains why the 0.1 µF cap cannot do the bulk job alone: its impedance at 10 kHz is dominated by ESR and ESL, not by the ideal capacitance.

Analog sections on mixed-signal boards follow the audio rules from the earlier sections instead: coupling caps between stages sized from the impedances they see, plus generous rail bypass where the analog supply meets the noisy digital one. When a board has both worlds, decouple them separately and join the grounds at exactly one point.

± Installation rules

Polarized Capacitor Orientation — Getting the Plus and Minus Right

Once your calculation lands on an electrolytic — as it nearly always does at 1 µF and above — you inherit an installation rule the math does not show: aluminum electrolytics are polarized. The positive terminal must sit at the higher DC potential. In a coupling application that means the plus side faces the stage with the higher DC bias: for an input cap between a 0 V-biased source and a grid biased at −1 V, the plus side faces the source; for an output cap feeding a speaker from a single-supply amp sitting at half the rail, the plus side faces the amp.

Emitter and cathode bypasses are easier: the resistor's signal-side end sits at a positive DC voltage, so the capacitor's plus terminal goes there and the minus terminal goes to ground. Reverse-biasing an electrolytic by more than about a volt degrades the dielectric, generates gas, and eventually vents the part — audibly first as distortion, later as a small pop and a smell you will not forget.

❌ Don't install blind. If a coupling capacitor can see signal swing of both polarities around 0 V — speaker outputs, and any bipolar-supply circuit — a polarized electrolytic is the wrong part. Use a bipolar (non-polarized) electrolytic, or two equal electrolytics in series with plus ends outward. The static table's speaker row assumes exactly this consideration in single-supply designs.

Reality check

Real-World Corrections — X7R DC Bias and Electrolytic Tolerance

The formula assumes the capacitor is the value printed on it. Two common dielectrics break that assumption loudly. First, X7R and other class-II ceramics lose capacitance as DC voltage across them rises — a "10 µF" 0805 X7R rated 16 V can measure closer to 4 µF with 12 V across it, a 60% loss that no formula predicts unless you consult the manufacturer's DC-bias curves. This is why X7R belongs on rails (where the cap works around a fixed bias you can derate for) and not in the signal path (where each volt of signal silently re-tunes your corner frequency).

⚠️ Warning. Never copy a rail-decoupling value into a signal-path position without checking the DC-bias curve. A 0.1 µF X7R coupling cap carrying a 2 V bias may behave like 0.05 µF — shifting your carefully computed corner frequency by an octave and undoing Step 5's verification entirely.

Second, aluminum electrolytics carry tolerance ratings of ±20% as standard, and −20/+10% or even −10/+50% in some series — and they lose capacitance further as they age and dry out, especially in hot enclosures. Design margin absorbs this: if your achieved corner frequency sits at 16.9 Hz with a nominal part, a −20% batch still lands near 21 Hz, which is why the decade rule (fc = 2 Hz for 20 Hz audio) exists rather than "fc = 19 Hz". The extra margin is not luxury; it is the tolerance budget spent in advance.

Put together, the corrections produce a simple purchasing heuristic: compute with the formula, choose the standard value one step larger than the math demands whenever the part is an electrolytic or a class-II ceramic, and reserve exact-value designs for film and C0G parts whose tolerances run 5% or better. The checklist above encodes this judgment so it survives contact with a real parts bin.

Save this tool. Bookmark this coupling and decoupling capacitor calculator — it is one of sixteen free tools hosted at Capacitor Calculator, and the full index of every sizing, conversion and reference tool lives in the calculators hub. Whether you are sizing an input coupling cap for a tube preamp, an output cap for an 8 Ω speaker, an emitter bypass for a transistor stage or the 0.1 µF decoupling network for a microcontroller rail, the same five-step procedure and the same C = 1/(2πfR) formula apply — and this page keeps them in one place.

Coupling & Bypass Capacitor Questions

How do I calculate coupling capacitor value?▼

Use C = 1/(2π·fc·R). First find R, the total series resistance the capacitor sees: the source's output impedance plus the load's input impedance. Then choose fc at least one decade below your lowest frequency of interest — 2 Hz for full-range audio. A 1 kΩ source driving a 1 kΩ load at fc = 20 Hz needs C = 1/(2π × 20 × 2000) ≈ 3.98 µF, so you round up to the 4.7 µF standard value and re-verify that the achieved corner of about 16.9 Hz sits below your target.

What size bypass capacitor for an emitter resistor?▼

Treat it as a low-pass problem: the capacitor is in parallel with R_E and the transistor's internal emitter resistance r′e ≈ 25 mV ÷ Ic. Compute C = 1/(2π·fc·R) with R = R_E || r′e. On paper, a 1 kΩ resistor bypassed for 100 Hz needs only about 1.59 µF (round to 2.2 µF) — but because r′e drags the effective resistance down sharply, real audio stages typically land between 10 µF and 100 µF of electrolytic. When in doubt, go larger: the cost is only a slightly longer charge time at power-on.

What does a cathode bypass capacitor do?▼

In a tube (valve) stage, the cathode resistor sets the DC bias but also creates negative feedback that lowers gain. The cathode bypass capacitor shunts that resistor for AC signals only, restoring the full gain of the tube while leaving the DC operating point untouched. Size it with the same C = 1/(2π·fc·R) formula, using R equal to the cathode resistor in parallel with the tube's internal cathode resistance — the tube preamp worked example below walks through an 811 Ω case at 20 Hz.

Why 0.1 µF for decoupling?▼

A 0.1 µF (100 nF) ceramic sits at a sweet spot: large enough to supply local charge through nanosecond switching transients, yet small enough that its self-resonant frequency stays in the tens of MHz, so its impedance remains low exactly where fast logic edges put their energy. X7R dielectric holds that impedance across temperature and applied voltage far better than Y5V, which is why the 0.1 µF decoupling value became the industry default placed next to every power pin.

Coupling vs decoupling vs bypass — what's the difference?▼

A coupling capacitor sits in series with the signal path: it blocks DC and passes AC from one stage to the next. A decoupling capacitor sits in parallel with the supply, from rail to ground: it shunts high-frequency noise away from the IC and supplies transient current locally, so the two names are used interchangeably for the rail job. 'Bypass' covers both senses — a bypass capacitor bypasses something: a rail's noise to ground, or an emitter/cathode resistor for AC.

What happens if the coupling capacitor is too small?▼

The high-pass corner fc = 1/(2πRC) climbs into your signal band. In an audio system that means bass rolls off early, low frequencies arrive phase-shifted relative to the rest of the spectrum, and the sound turns thin and 'late'. In measurement gear it means droop on square waves. The fix is the verification step: recompute the achieved corner with the standard value you actually plan to buy, and step up the ladder until it sits below your lowest wanted frequency.

Electrolytic or ceramic for coupling?▼

Follow the value. Anything 1 µF and above is usually aluminum electrolytic — bipolar (non-polarized) types wherever reverse voltage can appear, such as speaker coupling. Below 1 µF, prefer film (polypropylene or polyester) or C0G/NP0 ceramic in the signal path. Avoid high-K ceramics such as X7R and especially Y5V for coupling: their capacitance falls with applied DC bias, they can be microphonic, and their wide tolerance can shift your corner frequency by tens of percent.

Does the 50 Hz mains frequency change the calculation?▼

No. Signal coupling uses the lowest frequency inside your signal band, not the mains frequency — audio coupling math never references 50 or 60 Hz. Mains matters only where it leaks in as hum: a 50 Hz hum complaint calls for better rail decoupling, bulk capacitance and layout hygiene, and rectifier ripple sums at 100 Hz (or 120 Hz) for full-wave supplies. Those are power-supply sizing problems, a separate calculation from the signal coupling done here.

Keep calculating