A double organ-blowing bellows is a paired pneumatic device that produces a continuous stream of pressurised air using two feeder chambers operating 180° out of phase. Pipe organ builders and traditional blacksmiths rely on it because a single bellows produces a pulsing output that ruins steady tone and inconsistent forge heat. While one chamber draws air through an inlet flap, the other discharges through a non-return valve into a reservoir or wind trunk, smoothing flow to within ±2% of nominal pressure at typical organ wind levels of 70–100 mm water column.
Double Organ-blowing Bellows Interactive Calculator
Vary feeder swept volume, pumping speed, and volumetric efficiency to see average wind-trunk airflow and losses.
Equation Used
The paired bellows flow is the swept volume of one feeder multiplied by its stroke rate, doubled for the second feeder, then reduced by volumetric efficiency. With Vs in L/stroke and N in strokes/min, the result is L/min.
- Both feeder bellows have the same swept volume and run 180 deg out of phase.
- Stroke rate is the rate of one feeder; the factor of 2 accounts for the paired feeder.
- Low-pressure organ wind is treated as incompressible for volumetric flow sizing.
- Volumetric efficiency lumps valve leakage, leather porosity, and reverse flow.
- The supplied worked-example excerpt is truncated before numeric inputs/results, so defaults are practical chamber-organ demonstration values.
The Double Organ-blowing Bellows in Action
The mechanism solves a problem you cannot dodge with a single bellows — air output stops the instant the top board reaches the bottom of its stroke. For a pipe organ that means the tone collapses every time the blower lifts the handle. For a forge it means the fire dies between strokes. Pair two feeders, drive them on opposite phases, and one chamber is always discharging while the other is filling. The output stream stays continuous, and a downstream reservoir bellows (the weighted top board you see resting above the wind trunk) absorbs whatever ripple remains.
Each feeder uses two leather-flapped non-return valves. The inlet valve sits in the bottom board and opens on the suction stroke. The outlet valve sits between the feeder and the reservoir and opens on the compression stroke. If you notice the flapper geometry on a Snetzler or Father Smith organ, the leather is skived to about 1.5 mm at the hinge edge and 3 mm at the seal — that taper is what lets the valve close in under 80 ms without slap. Get the leather too thick and the valve lags, dumping back-pressure into the feeder and starving the wind chest. Too thin and the flap flutters, audible as a low-frequency wheeze through the pipework.
Tolerances on the cuneiform (wedge-shaped) feeder ribs matter more than people expect. The hinge gap between rib and frame must hold below 0.5 mm with the bellows fully collapsed, otherwise air escapes sideways instead of through the outlet valve. Common failure modes are leather perishing at the fold (you'll see flour-like dust under the bellows), rib hinges loosening and letting the feeder rack out of square, and the inlet valve sticking open after long idle periods — that last one shows up as a sudden loss of reservoir pressure on the first pump of the day.
Key Components
- Feeder bellows (×2): The two wedge-shaped pumping chambers driven 180° out of phase by a crank, foot treadle, or electric blower drive. Typical stroke volume on a chamber organ runs 8–15 litres per stroke at 30–60 strokes per minute. The 180° phasing is critical — drift past 170° or 190° and you get an audible breath in the wind supply.
- Inlet (suction) flap valves: Leather-faced non-return valves in the bottom board that open during the expansion stroke to admit fresh air. Skived leather thickness 1.5–3 mm with a closing time under 80 ms keeps the valve quiet and responsive.
- Outlet (delivery) flap valves: One-way valves between each feeder and the wind trunk, sealing during the suction stroke so the partner chamber doesn't pull air backward through the idle feeder. Seat flatness must hold within 0.2 mm to prevent reverse leakage.
- Reservoir bellows: A weighted, parallel-rise top board that floats on the combined feeder output, absorbing the small residual pulsation and holding wind chest pressure steady to ±2%. Weights are sized for 70–100 mm water column on most stop organs, 50 mm on chamber organs.
- Wind trunk: The wooden duct carrying wind from the reservoir to the soundboard. Internal cross-section sized so velocity stays under 6 m/s — go higher and you'll hear hiss in quiet stops.
- Cuneiform ribs and leather gussets: The folding sides that let each feeder collapse without leakage. The leather is glued under tension and folded along a sharp crease — once that crease perishes, air leaks sideways and feeder efficiency collapses.
Industries That Rely on the Double Organ-blowing Bellows
Double feeder bellows show up wherever a process needs continuous low-pressure air without electric blowers. Most surviving examples sit in pipe organs, but historical metalworking, glass blowing, and even early chemistry labs used the same alternating-feeder principle. The defining characteristic is always two chambers, two sets of non-return valves, and a downstream smoothing volume — whether that's a reservoir bellows, a wind trunk, or a tuyère pipe.
- Pipe organ building: The Father Smith organ at St Paul's Cathedral originally used hand-pumped double feeder bellows; many Snetzler chamber organs still operate with their original 18th-century double-feeder rigs.
- Traditional blacksmithing: Double-chamber great bellows used in Catalan forges and at the Sturbridge Village working blacksmith shop in Massachusetts deliver continuous air to the tuyère for steady fire temperature.
- Glass blowing: Historical Murano glass furnaces used paired feeder bellows to maintain consistent furnace temperature during long working sessions.
- Reed organ and harmonium repair: Mason & Hamlin parlour reed organs use a miniaturised double-feeder system foot-pumped by the player, feeding a small reservoir behind the reed pan.
- Heritage industrial sites: The waterwheel-driven double bellows at the Saugus Iron Works National Historic Site in Massachusetts feed the bloomery furnace continuously.
- Pump organ restoration: Estey Organ Company restorations frequently rebuild original double-feeder leather using 1.5 mm chrome-tanned sheepskin matched to the 1880s factory specification.
The Formula Behind the Double Organ-blowing Bellows
The useful number for a working organ builder or forge restorer is the average volumetric flow delivered by the pair of feeders. At the low end of typical operating range — slow hand-pumping at 20 strokes per minute on a small chamber organ — flow drops to roughly a third of nominal and the reservoir does most of the work absorbing the gap between strokes. At nominal 40 strokes per minute, the reservoir barely moves. Push past 60 strokes per minute and the leather starts heating, valves chatter, and you lose efficiency to backflow before each outlet valve fully seats.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Qavg | Average volumetric flow delivered to the wind trunk | m³/s | ft³/min (CFM) |
| Vs | Swept volume of one feeder per stroke | m³ | ft³ |
| N | Stroke rate of one feeder (the factor of 2 accounts for the second feeder operating 180° out of phase) | strokes/s | strokes/min |
| ηv | Volumetric efficiency — accounts for valve leakage, leather porosity, and reverse flow during valve closure | dimensionless (0–1) | dimensionless (0–1) |
Worked Example: Double Organ-blowing Bellows in a restored 18th-century chamber organ
You are restoring a 1768 Snetzler chamber organ at a heritage chapel in Bath. The instrument has two original cuneiform feeder bellows, each with a measured swept volume of 12 litres per stroke. You are sizing an electric blower replacement and need to know the equivalent average flow at the historical hand-pumping cadence so the new blower matches the wind chest's natural feel. Volumetric efficiency on the rebuilt sheepskin leather measures 0.88 on a static leak test.
Given
- Vs = 0.012 m³/stroke
- Nnominal = 40 strokes/min
- —v = 0.88 dimensionless
Solution
Step 1 — convert the nominal stroke rate from strokes per minute to strokes per second:
Step 2 — compute the nominal average flow with both feeders contributing:
Step 3 — at the low end of the typical hand-pumping range, 20 strokes per minute on a quietly-played chamber piece:
That is the threshold where the reservoir bellows starts visibly sinking between strokes — you would see the weighted top board drop 15–20 mm before the next compression catches it. The organ still speaks but a held chord on the principal stops will go flat in pitch by 2–3 cents as wind pressure droops. At the high end, an aggressive 60 strokes per minute during a full-organ passage:
In theory the flow scales linearly, but in practice ηv drops to around 0.78 above 55 strokes per minute because the outlet flap valves cannot fully reseat in the 0.5 second cycle window — air leaks backward into the just-expanding feeder, audible as a low huffing noise from the bellows frame.
Result
Nominal average flow is 0. 0141 m³/s, or about 30 CFM, which is exactly what a 1/4 HP centrifugal organ blower delivers against 75 mm water column — a clean match for the original feel of the instrument. At 20 strokes/min the system delivers 15 CFM and the reservoir does most of the work; at 60 strokes/min you reach 45 CFM in theory but valve reseating losses pull real-world output closer to 40 CFM and the bellows starts complaining audibly. If your measured flow comes in 20% below predicted, the most likely causes are: (1) outlet valve leather curled at the edges from age, lifting the seal off the seat by 1–2 mm, (2) the cuneiform rib hinge gap exceeding 0.5 mm and bleeding air sideways during compression, or (3) a perished gusset corner you cannot see without a borescope, typically at the rear bottom fold where moisture collects.
Double Organ-blowing Bellows vs Alternatives
Three pneumatic options can supply continuous low-pressure air to a pipe organ or forge. Each one buys you something different on flow steadiness, cost, and how the instrument responds to the player.
| Property | Double organ-blowing bellows | Single bellows with reservoir | Electric centrifugal blower |
|---|---|---|---|
| Pressure ripple at typical operating point | ±2% of nominal | ±8–15% without large reservoir | ±0.5% (electronically regulated) |
| Typical flow range | 15–60 CFM hand-pumped | 5–20 CFM hand-pumped | 20–500 CFM |
| Capital cost (restoration grade) | £3,000–£8,000 rebuild | £1,500–£3,500 rebuild | £600–£2,000 new unit |
| Service life of leather/diaphragms | 40–80 years on chrome-tanned sheepskin | 40–80 years on chrome-tanned sheepskin | 10,000+ hours on motor bearings |
| Power source flexibility | Hand, foot, water wheel, or electric | Hand or foot | Mains electricity only |
| Tonal authenticity for heritage organ | Authentic — preserves natural wind sag | Authentic but pulses unless reservoir is large | Sterile — too steady for pre-1900 voicing |
| Maintenance interval (active use) | Re-leather every 40–60 years | Re-leather every 40–60 years | Bearing service every 5–10 years |
Frequently Asked Questions About Double Organ-blowing Bellows
The phasing on the crank is rarely the problem — the issue is almost always asymmetric valve timing between the two feeders. If one outlet flap is slightly stiffer than the other (different leather batch, different glue cure, or one valve sat compressed in storage), it opens later in the compression stroke. That delay creates a brief moment where neither feeder is delivering wind, which the reservoir cannot fully smooth.
Quick diagnostic — pump slowly by hand and watch the reservoir top board. If it dips on every other stroke instead of staying level, your feeders are not delivering equal flow. Re-skive the slower outlet valve to match the faster one within 0.2 mm thickness.
For a working instrument, chrome-tanned sheepskin at 1.5–2 mm. It resists humidity swings far better than alum-tanned and holds its suppleness for 40–80 years. Alum-tanned leather is closer to the original 1700s material but goes brittle in 20–30 years in modern centrally-heated buildings.
The decision usually comes down to what kind of restoration you are doing. A museum-grade conservation rebuild that will not be played heavily can justify alum-tanned. A working chapel organ that gets used weekly should get chrome-tanned every time — the tonal difference is undetectable and the lifespan is doubled.
Rule of thumb is reservoir volume should equal at least 4–6× the swept volume of a single feeder. So if each feeder displaces 12 litres, the reservoir wants 50–75 litres of usable travel. Below 4× and you'll hear pulsation on quiet stops. Above 8× and the reservoir becomes sluggish — quick passages on full organ make the wind pressure droop because the reservoir cannot empty fast enough to call for more wind from the feeders.
The weighted top board mass is the other half of the equation. Size it for your target wind pressure first (typically 70–100 mm water column on a stop organ), then size the volume around it.
Calculated average flow is not the same as instantaneous demand. A full-organ chord with 16-foot pedal pipes can spike demand to 3–4× the average for the first 200–300 ms while pipes fill and stabilise. If your reservoir cannot supply that transient, pressure droops before the feeders can catch up.
Check the wind trunk cross-section first — if internal velocity exceeds 6 m/s during transients, the trunk itself is choking the flow regardless of feeder capacity. Then check reservoir rise distance; you want at least 80 mm of usable travel between the rest position and the top of stroke.
Yes, and many heritage organ shops do this routinely. The trick is the motor must drive a crank that replicates the natural stroke profile of a hand pump — roughly sinusoidal, with the compression stroke slightly faster than the suction stroke. A constant-speed AC motor through a simple eccentric works fine for chamber organs at 30–45 strokes per minute.
What you must avoid is replacing the feeders entirely with a centrifugal blower. The feeder-and-reservoir combination produces a tiny natural wind sag during attack that voicers spent the 18th and 19th centuries tuning the pipes around. A perfectly steady centrifugal blower removes that sag and the pipes sound sterile. Keep the feeders, motorise the crank.
Static leak tests measure leakage at full compression with valves seated. They miss two dynamic losses: valve reseating delay (the outlet flap takes 60–100 ms to fully close after the stroke reverses) and gusset flexing under cyclic load that opens micro-gaps you cannot detect at rest.
To verify, repeat the leak test at 50% compression with the bellows held mid-stroke. If efficiency at mid-stroke drops below 0.85, your gussets are flexing more than they should — usually because the leather is glued under insufficient tension or the rib hinges have loosened by more than 0.3 mm. Either fault drops dynamic efficiency by 10–15 points without showing on a static test.
Three cases justify full replacement. First, if the instrument was modified after 1920 with high-pressure stops (above 150 mm water column) — the original feeders were not designed for that load and will fatigue rapidly. Second, if the bellows frame timber is structurally compromised by woodworm or splits that re-leathering cannot fix. Third, if the building no longer has space for the original bellows housing.
For everything else, restore the original. A correctly rebuilt double feeder system gives 40–80 years of service, preserves the tonal character the pipes were voiced for, and on a heritage instrument is usually a listed-building requirement anyway.
References & Further Reading
- Wikipedia contributors. Bellows. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
