Modification of Two-pawl Lever

A modified two-pawl lever is an oscillating ratchet driver that uses two spring-loaded pawls on a rocking arm to advance a ratchet wheel one tooth per half-stroke instead of one tooth per full cycle. Unlike the standard single-pawl ratchet, which idles on the return stroke, this version drives forward in both directions through clever pawl geometry. The result is doubled output rate per oscillation, which is why you find it inside stroke counters, fare meters, and tally registers where every input motion must produce a count.

Watch the Modification of Two-pawl Lever in motion
Video: Lever of two positions 2 by Nguyen Duc Thang (thang010146) on YouTube. Used here to complement the diagram below.

Inside the Modification of Two-pawl Lever

The mechanism sits on a single rocking lever pivoting beside a ratchet wheel. Two pawls ride on that lever, but they engage the wheel at different angular positions — typically 180° offset around the rim, or close to it. As the lever swings clockwise, pawl A pushes a tooth forward and pawl B slides over the rim of the wheel without engaging. On the return swing, pawl B picks up the next tooth and drives it forward while pawl A now does the sliding. A separate holding pawl (sometimes called a detent or click) prevents back-rotation between strokes. So you get two indexes per full oscillation instead of one — a single-tooth advance per stroke, every stroke.

The geometry has to be right or the thing skips. Pawl tip radius must match the tooth root radius within roughly 0.05 mm in a typical small counter wheel. Spring preload on each pawl is usually 0.2 to 0.5 N — too light and the pawl bounces off during a fast stroke, too heavy and the slide-over phase wears the rim flat in a few thousand cycles. Pawl-engagement angle should bisect the tooth flank within ±2°. If you cheap out on the pawl pivot bushings and they develop more than about 0.1 mm of radial slop, the drive pawl starts hunting between two adjacent teeth and you'll see double-counts on alternate strokes.

The most common failure modes you'll meet in the field: pawl spring fatigue causing missed counts, worn ratchet teeth (rounded crests below 0.3 mm peak height stop catching reliably), and holding-pawl drift letting the wheel back-rotate during the dead point of the input stroke. Anti-backlash relies on the holding pawl staying seated — if it lifts, the drive pawl will pull the wheel partway back during return.

Key Components

  • Rocking Lever (Rocker Arm): The oscillating input member that carries both drive pawls. Typical swing angle is 20° to 45°, sized so each pawl traverses exactly one tooth pitch. Pivot bushing clearance must stay under 0.1 mm or the pawls hunt between teeth.
  • Drive Pawl A: Engages the ratchet wheel during the forward swing of the lever. Spring-loaded at 0.2–0.5 N preload, with tip radius matched to the tooth root within 0.05 mm. Slides over the rim during the return stroke without engaging.
  • Drive Pawl B: Mounted on the same rocking lever but offset so it engages a tooth on the return stroke. Identical geometry to Pawl A but mirrored — together the pair gives one tooth advance per half-stroke, doubling indexing rate.
  • Ratchet Wheel: The driven member with asymmetric sawtooth teeth — typically 30 to 60 teeth on a small counter, tooth pitch 1.5–3 mm. Tooth flank angle around 60°, back angle 5°–10° to lock against the holding pawl. Crest wear below 0.3 mm height causes missed counts.
  • Holding Pawl (Detent Click): A separate spring-loaded pawl on the frame that prevents back-rotation between strokes. Critical during the dead-centre transition where neither drive pawl is engaged. Must stay seated under 0.3 N minimum to resist drive-pawl pull-back.
  • Pawl Springs: Light hairpin or torsion springs maintaining pawl-tooth contact. Preload 0.2–0.5 N for a typical small counter. Below 0.15 N the pawl bounces during fast input; above 0.6 N you accelerate rim wear during slide-over.

Real-World Applications of the Modification of Two-pawl Lever

You find the modified two-pawl lever wherever the input is an oscillation and the output must be a reliable count or step — every stroke matters. The bidirectional ratchet pawl design means a partial input still registers a count if it crosses the half-stroke threshold, which is why it dominates mechanical totalisers and stroke counters. Common search territory here covers fare registers, hand-tally counters, lever-operated counters, and double-acting pawl mechanisms in vintage instrumentation.

  • Mechanical Counters: Veeder-Root 1500-series stroke counters used on punch-press cycle logging, where each ram descent oscillates a sense lever through a fixed arc.
  • Public Transit: Ohmer fare register units in 1940s–1960s taxi meters, advancing the unit-fare wheel on each meter-throw of the driver's lever.
  • Industrial Machinery: Stroke counters on Bliss and Minster mechanical presses, totalising ram strokes via a rocker driven off the crankshaft.
  • Textile Machinery: Pick-and-pattern counters on Dornier rapier looms where each shed-change rocks an input lever and increments a yardage tally.
  • Laboratory & Medical: Hand-held lab tally counters and blood-cell differential counters where thumb depression oscillates a small lever and the bidirectional pawl advances the digit drum on every press without missing partial strokes.
  • Vending and Coin-Op: Older National Cash Register (NCR) totaliser wheels in pre-electronic registers, where the keystroke arm oscillated a two-pawl driver against the units wheel.

The Formula Behind the Modification of Two-pawl Lever

The core question with this mechanism is how many teeth advance per unit time given an oscillating input. At the low end of the typical operating range — say 30 strokes per minute on a hand-tally counter — the output advances slowly enough that pawl spring preload dominates reliability. At the nominal range of 60–120 strokes per minute the mechanism hits its sweet spot — pawl bounce isn't yet a problem and rim wear stays predictable. Push past 300 strokes per minute, like on a high-cycle press counter, and you start running into pawl-bounce missed counts even with proper spring sizing. The formula below gives you the theoretical advance rate; how close you come to it depends on whether you've stayed inside that envelope.

Nteeth = 2 × fosc × ηengage

Variables

Symbol Meaning Unit (SI) Unit (Imperial)
Nteeth Tooth advance rate of the ratchet wheel teeth / s teeth / s
fosc Oscillation frequency of the rocking lever (full cycles per second) Hz cycles / s
ηengage Engagement efficiency — fraction of half-strokes that successfully advance a tooth (0.95–1.00 for a healthy mechanism) dimensionless dimensionless
θswing Angular swing of the rocking lever per half-stroke — must equal one tooth pitch angle degrees degrees
ptooth Angular pitch between adjacent ratchet teeth degrees degrees

Worked Example: Modification of Two-pawl Lever in a hospital pharmacy pill-counting tray lever

You are designing the count register on a manual pill-counting tray for a hospital pharmacy — the kind a tech rocks with a thumb lever to sweep pills into a chute, where every rock must increment a digit drum by one count regardless of stroke direction. The rocking lever oscillates at a nominal 90 strokes per minute during steady counting. The ratchet wheel has 36 teeth, giving a tooth pitch of 10°. You want to verify the modified two-pawl lever delivers reliable counts across the realistic range of operator pace.

Given

  • fosc,nom = 90 strokes / min (1.5 Hz)
  • Teeth on wheel = 36 teeth
  • ptooth = 10 degrees
  • ηengage = 0.99 dimensionless (healthy build)
  • Pawl spring preload = 0.3 N

Solution

Step 1 — convert the nominal stroke rate from strokes per minute to Hz:

fosc,nom = 90 / 60 = 1.5 Hz

Step 2 — apply the formula at nominal pace. Each full oscillation advances 2 teeth (one per half-stroke), so:

Nteeth,nom = 2 × 1.5 × 0.99 = 2.97 teeth / s

That works out to about 178 counts per minute — comfortably above the 90 strokes per minute the operator delivers, because every stroke registers in both directions. The pharmacy tech sees a clean count-per-rock with no missed indexes.

Step 3 — at the low end of the typical operating range, a slow counting pace of 30 strokes per minute (0.5 Hz):

Nteeth,low = 2 × 0.5 × 0.99 = 0.99 teeth / s ≈ 60 counts / min

At this speed pawl bounce is a non-issue and the 0.3 N spring preload sits well inside its reliable range — you'd never see a missed count from a tired tech working slowly.

Step 4 — at the high end, an aggressive 240 strokes per minute (4 Hz) which a fast operator can hit during bulk counting:

Nteeth,high = 2 × 4.0 × ηengage

Here ηengage drops below 0.99. With a 0.3 N preload the pawl tip starts to lift off the rim during the slide-over phase at this speed — engagement efficiency falls to roughly 0.92, giving 7.36 teeth / s actual versus 7.92 teeth / s theoretical. The operator notices missed counts every 12 to 15 strokes, which is the practical ceiling for this build without bumping spring preload to 0.5 N.

Result

The mechanism advances 2. 97 teeth per second at the nominal 90 strokes per minute — equivalent to 178 counts per minute, comfortably tracking every operator stroke without lag. At 30 strokes per minute the mechanism cruises with full reliability and no risk of missed counts, while at 240 strokes per minute pawl bounce knocks engagement efficiency from 0.99 down to about 0.92 and you start seeing one missed count per dozen strokes — the sweet spot lives between 60 and 180 strokes per minute. If you measure fewer counts than predicted at nominal pace, look for these failure modes in order: (1) holding-pawl spring weakened below 0.2 N letting the wheel back-rotate during the dead point, (2) pawl pivot bushing wear over 0.1 mm radial slop causing the drive pawl to hunt between adjacent teeth, or (3) ratchet tooth crest wear below 0.3 mm peak height where the pawl tip skates over the tooth instead of catching the flank.

Choosing the Modification of Two-pawl Lever: Pros and Cons

Designers picking an intermittent indexer for a counter face a real trade-off between count rate, reliability, and complexity. Here's how the modified two-pawl lever stacks up against the single-pawl ratchet it competes with directly, and against a Geneva drive that comes from a different design space entirely.

Property Modified Two-Pawl Lever Single-Pawl Ratchet Geneva Drive
Indexes per input cycle 2 (one per half-stroke) 1 (forward stroke only) 1 per input revolution
Practical max stroke rate ~300 strokes / min ~400 strokes / min ~600 RPM input
Indexing accuracy ±0.5° at tooth flank ±0.5° at tooth flank ±0.05° geometric lock
Cost (small counter scale) Low — stamped pawls and springs Lowest — single pawl High — precision slot and pin
Maintenance interval 50,000–200,000 strokes 100,000–300,000 strokes 1M+ cycles
Tolerance to dirt and debris Good — pawls self-clear Good Poor — slot fouls easily
Best application fit Bidirectional stroke counters, fare meters One-way ratchets, cable winches High-speed indexing turrets
Failure mode at limit Pawl bounce, missed counts Pawl bounce, idle return wastes input Slot-pin pound-out, position drift

Frequently Asked Questions About Modification of Two-pawl Lever

This is almost always asymmetric pawl geometry. The two drive pawls must be mirror-images with engagement angles matched within ±2°. If one pawl engages 4° later than the other, the slow pawl misses the tooth flank during fast strokes while the fast pawl still catches — you get one count per full cycle instead of two.

Check by manually rocking the lever at quarter-speed and watching each pawl tip. Both should kiss the tooth flank at the same lever angle relative to mid-stroke. If one engages noticeably later, file the late pawl's stop or reposition its mounting hole.

Size for the high end and accept slightly faster wear. At 200 strokes per minute the pawl tip experiences inertial lift forces around 0.25–0.35 N during slide-over, so you need preload above that — typically 0.5 N. The same spring still works fine at 30 strokes per minute; the rim just sees about 30% more sliding friction than strictly necessary at slow speed.

If you size for the slow end (0.2 N) and run fast, the pawl bounces clear of the rim and skips counts. The asymmetry is intentional — slow operation tolerates a heavy spring far better than fast operation tolerates a light one.

Depends on whether your input lever returns under positive drive or under spring return. If the press crank rocks the input lever symmetrically — equal force in both directions — go two-pawl and double your count resolution per input cycle.

If the return is spring-loaded with low force (common on retrofitted sensor levers), go single-pawl. The return stroke on a spring system often doesn't deliver enough force to drive a tooth advance against the holding-pawl detent friction, and you'll get inconsistent counts on the return half. The two-pawl design only pays off when both half-strokes are mechanically driven.

You're catching the wheel at a mid-index position because the holding pawl isn't fully seating between strokes. The drive pawl pulls the wheel partway forward, then on the return half before the second drive pawl engages, the wheel can drift back if the holding pawl spring is weak or its tooth-flank contact angle is off.

Inspect the holding pawl: it should sit with at least 60% of its tip width against the tooth back-flank, with spring preload of 0.3 N or higher. If you see polish marks only on the tip corner, the pawl is rocking on its pivot and not locking the wheel — replace the bushing or shim out the slop.

Yes, but you need to manage the return stroke carefully. A solenoid with a spring return delivers asymmetric force — strong pull, weak push. With a modified two-pawl lever this means your forward count is rock-solid but your return count starts missing under any added friction.

Either use a push-pull solenoid (two coils, equal force both directions) or accept that you're running it as a single-pawl ratchet and remove the second drive pawl entirely. Running a half-functional two-pawl setup wears the unused pawl's slide-over surface for no benefit.

For long life, go larger pitch and harder material — counterintuitive because finer pitch gives better resolution. A 36-tooth wheel at 10° pitch in hardened tool steel (HRC 58–62) will outlast a 60-tooth wheel at 6° pitch in brass by roughly 4×, because tooth crest wear is proportional to contact stress and inversely to flank length.

Veeder-Root commercial counters typically use 30–40 teeth in case-hardened steel for high-cycle applications, and reserve fine-pitch brass wheels for low-cycle decorative or instrument work. Match the wheel hardness to the pawl tip — same hardness or pawl slightly softer, so wear shows up on the cheaper-to-replace pawl first.

References & Further Reading

  • Wikipedia contributors. Ratchet (device). Wikipedia

Building or designing a mechanism like this?

Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.

← Back to Mechanisms Index
Share This Article
Tags: