A Modified Variable Alternating Traverse is a cam-driven guide mechanism that lays yarn, wire, or filament back and forth across a rotating spool while continuously varying its stroke length, traverse speed, and end dwell to prevent ribboning and edge build-up. The concept traces to Foster Needle and SACM-Schlumberger's textile-winding work in the mid-20th century. The cam profile shifts each pass slightly so layers never stack identically, breaking up resonance between spindle and traverse. The result is denser, square-edged packages that unwind cleanly at 1500+ m/min on modern texturising and copper magnet-wire lines.
Modified Variable Alternating Traverse Interactive Calculator
Vary stroke modulation, dwell, spindle speed, and modulation period to see the changing traverse stroke and anti-ribboning cam action.
Equation Used
The calculator models the article's modified variable alternating traverse by shifting both stroke endpoints by dL on a slow modulation cam cycle. The instantaneous stroke is S(n) = S0 + 2*dL*sin(2*pi*n/N), so the guide sweeps between S0 - 2*dL and S0 + 2*dL. The modulation rate is spindle rpm divided by the modulation period in spindle revolutions.
- Endpoint modulation is symmetric about the nominal stroke center.
- dL is the shift at each stroke endpoint, so total stroke changes by plus or minus 2*dL.
- Recommended end dwell band is 8 to 15 ms as stated in the article.
- Modulation period is measured in spindle revolutions per cam modulation cycle.
Operating Principle of the Modified Variable Alternating Traverse
The mechanism sits between a rotating package spindle and the material guide — the eyelet, pigtail, or roller that the yarn or wire passes through. A reverse-helix shaft, a barrel cam, or a programmable servo drives the guide left-right across the package face. In the standard fixed-stroke version, the guide travels the same distance every pass, which is exactly what causes ribboning: when the traverse-to-spindle ratio lands on a whole number or simple fraction, every layer drops on top of the previous one, building hard ridges that snag and break on payout. The modified variable alternating traverse fixes this by deliberately changing the stroke length, the traverse speed, or both, on a slow cycle that walks the lay pattern through the package over hundreds of revolutions.
The cam profile carries three modifications stacked on one drum. First, a stroke-length modulation — typically ±2 to ±5 mm of variation around the nominal stroke — shifts where each end of travel lands. Second, a velocity modification through the centre of the stroke flattens the lay angle and prevents the traverse ratio from ever locking onto an integer. Third, an end-dwell modification — usually 8 to 15 ms at each turnaround — controls the cheese-edge density. Get the dwell wrong by more than 3 ms and you either build a soft mushy edge that collapses on the pallet, or a hard knife-edge that flakes coatings.
If you measure ribboning on a finished package, the cam is rarely the cause. Most failures trace to backlash in the traverse drive train letting the modulation collapse, worn bushings on the guide carriage shifting the centre of stroke by 1-2 mm per revolution, or a phase-lock between the modulation cycle and the spindle drive's own resonance. The whole system only works as long as the modulation period stays prime relative to the spindle period.
Key Components
- Reverse-Helix Traverse Shaft: A grooved cylindrical shaft with crossing left and right helical grooves cut at typically 12° to 18° pitch. The follower picks one groove, rides it to the end, then transfers at the crossover. Groove depth tolerance must hold ±0.05 mm or the follower jumps prematurely.
- Modulation Cam: A secondary cam, often coaxial with the traverse shaft, that shifts the follower's pivot point on a slow cycle of 30 to 120 spindle revolutions. This is the part that converts a fixed-stroke traverse into a variable one. Cam lift typically 4 to 8 mm peak-to-peak.
- Guide Carriage and Yarn Eyelet: The carriage rides linear rails and carries the porcelain or tungsten-carbide eyelet that contacts the material. Carriage mass should stay below 200 g for traverses above 800 cycles/min — beyond that, inertia overwhelms the cam profile and the lay angle smears.
- End-Dwell Tappet: A short flat or radius on the cam at each stroke end that holds the guide stationary for 8 to 15 ms. This builds the package edge. A worn tappet — even 0.2 mm of wear — visibly softens the cheese edge within one shift.
- Phase-Lock Defeat Pinion: A small differential gear that adds a slow rotational offset between the spindle drive and the traverse drive, ensuring the traverse ratio never settles on a rational number. Typical offset 0.3 to 1.2 RPM relative drift.
- Package Spindle: Drives the bobbin, cone, or spool that receives the material. Surface speed at the package — not RPM — is what matters; on a copper magnet-wire line this typically holds 1200 to 1800 m/min within ±1%.
Real-World Applications of the Modified Variable Alternating Traverse
The mechanism shows up wherever you need to wind a long, continuous strand into a dense, stable package that pays off cleanly at high speed. Textile yarn, fine copper magnet wire, optical fibre, fishing line, and welding wire all share the same fundamental ribboning problem, and they all use some flavour of modified variable alternating traverse to solve it.
- Textile Yarn Winding: SSM Giudici PSX precision package winders running polyester POY at 1200 m/min use a programmable variable traverse to build dye-package cheeses with uniform density for downstream package-dyeing autoclaves.
- Magnet Wire Manufacturing: Sampsistemi TWA double-twist bunchers and Niehoff MSM lines wind enamelled copper at 0.05 to 1.6 mm diameter onto DIN 250 to DIN 630 spools, where ribbon-free build is critical for downstream automatic motor-winding cells.
- Optical Fibre Spooling: Prysmian and Corning draw towers feed fibre at 25 to 60 m/s onto NATO-pattern shipping spools through servo-driven traverse heads that vary stroke to prevent micro-bend losses on inner layers.
- Fishing Line and Braid: Berkley and YGK braided line filling stations use a cam-driven variable traverse to prevent the dig-in that causes wind knots when an angler casts a high-speed reel.
- Welding Wire Packaging: Lincoln Electric and ESAB MIG wire spool lines wind 0.8 to 1.6 mm copper-coated steel at 1500 m/min onto B300 plastic spools, where any ribboning causes wire-feed stalls in the customer's robot cell.
- Synthetic Cordage: Samson Rope and other braided-cordage makers use slow-cycle variable traverse on creel-fed yarn packages so the braider draws evenly from every carrier.
The Formula Behind the Modified Variable Alternating Traverse
The single most useful number on a traverse-winding line is the traverse ratio — the number of spindle revolutions per complete double traverse stroke. When the ratio sits on an integer or a simple fraction like 2:1 or 3:2, ribboning is guaranteed. The whole point of the modified variable alternating traverse is to keep this ratio off the rational numbers, and the formula below tells you what ratio your machine is currently running and how far the modulation needs to shift it. At the low end of typical operating range — say 30 RPM spindle on a coarse welding wire — the ratio swings slowly and even small modulation matters. At the high end, 6000 RPM on a textured polyester package, the modulation has to track fast enough that the ratio drift outpaces the spindle's own jitter. The sweet spot for most textile and fine-wire work sits where the modulation period gives 60 to 120 spindle revs per traverse cycle.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Rt | Traverse ratio - spindle revolutions per double traverse stroke | dimensionless | dimensionless |
| Nspindle | Spindle rotational speed | rev/s | RPM |
| ftraverse | Traverse frequency (full strokes per second) | Hz | cycles/min |
| Lstroke | Nominal traverse stroke length | m | in |
| ΔL | Stroke modulation amplitude (peak-to-peak) | m | in |
Worked Example: Modified Variable Alternating Traverse in a fine copper magnet-wire spooler
Your team is commissioning a Niehoff MSM 85 magnet-wire spooler running 0.30 mm enamelled copper onto DIN 355 spools at a motor-winding contract supplier in Pune. The spindle runs at 3000 RPM nominal, the traverse cycles at 14 Hz, and the nominal stroke is 220 mm. You want to verify the traverse ratio sits clear of any integer and that the ±3 mm stroke modulation walks the lay across the package as designed.
Given
- Nspindle = 3000 RPM
- ftraverse = 14 Hz
- Lstroke = 220 mm
- ΔL = ±3 mm
Solution
Step 1 — convert the nominal spindle speed from RPM to rev/s so the units agree with the traverse frequency in Hz:
Step 2 — compute the nominal traverse ratio at 14 Hz traverse frequency:
That sits comfortably between 1.75 (7:4) and 2.0 (2:1), but 1.786 is close to 25:14 — a ratio that ribbons over a 14-revolution period. The modulation has to shift the effective ratio away from this rational neighbour.
Step 3 — check the low end of typical operating range. At spool start-up, the line typically runs at 1500 RPM while tension settles:
Below 1.0 the traverse moves faster than the spindle — the lay angle is steep, layers cross at sharp angles, and ribboning is physically impossible. Start-up is safe.
Step 4 — check the high end. The line tops out at 4500 RPM during the steady-state body of the spool:
This is the danger zone — 2.679 sits very close to 8:3 (2.667) and 19:7 (2.714). Without modulation the package will ribbon within 50 layers. With ±3 mm modulation on a 220 mm stroke, the effective ratio swings by roughly ±1.4% — enough to walk the lay through the rational neighbours rather than locking onto them.
Step 5 — verify modulation amplitude is sufficient:
Result
The nominal traverse ratio is 1. 786 spindle revolutions per double stroke, with a modulation swing of ±1.36% — enough to keep the lay pattern walking past the dangerous rational ratios at 1500, 3000, and 4500 RPM. At the 1500 RPM start-up the ratio of 0.893 means ribboning is geometrically impossible regardless of modulation; at the 4500 RPM steady state the 2.679 ratio sits dangerously close to 8:3 and the modulation is doing real work to keep the package clean. If you measure visible ribbons or hard ridges on the finished spool despite these numbers checking out, the most common causes are: (1) the modulation cam follower spring has lost preload and the modulation amplitude has collapsed below 1 mm peak-to-peak, (2) the phase-lock defeat pinion has seized and the traverse-to-spindle ratio is now fixed at a rational number, or (3) the traverse drive belt has stretched and lay-end position has drifted more than 5 mm, building one hard edge and one soft edge.
Choosing the Modified Variable Alternating Traverse: Pros and Cons
The variable alternating traverse is one of three families of solution to ribboning. The choice between them comes down to package speed, material cost, and how much electronics you are willing to put on a winding head.
| Property | Modified Variable Alternating Traverse (cam-based) | Fixed-Stroke Reverse Helix Traverse | Servo-Driven Programmable Traverse |
|---|---|---|---|
| Maximum traverse frequency | 20-25 Hz | 30 Hz+ | 50 Hz+ |
| Anti-ribboning effectiveness | High — modulation walks ratio off rational numbers | Low — relies on lucky ratio choice | Highest — software can dither continuously |
| Capital cost per head (USD, 2024) | $3,000-$8,000 | $1,200-$3,000 | $15,000-$40,000 |
| Lay-pattern flexibility | Fixed by cam profile — one product per cam | Fixed by helix pitch | Fully programmable per recipe |
| Mean time between failures | 8,000-15,000 hours | 15,000-25,000 hours | 20,000+ hours (electronics dominate) |
| Best application fit | Textile yarn, magnet wire, welding wire | Coarse cordage, low-speed packaging | Optical fibre, premium technical yarns, multi-product lines |
| Tolerance to drive backlash | Sensitive — backlash collapses modulation | Moderate | Self-corrects via encoder feedback |
Frequently Asked Questions About Modified Variable Alternating Traverse
You have hit a phase-lock between the modulation period itself and the spindle. The modulation cam is supposed to walk the traverse ratio through the rational neighbours, but if the modulation period in spindle revolutions divides evenly into the lay-repeat number, the cam shifts the lay onto the same set of dangerous ratios over and over.
The fix is to change the phase-lock defeat pinion ratio, or on a programmable head, add a second slower dither cycle on top of the modulation. As a rule of thumb, the modulation period in spindle revs should be prime relative to the integer part of the traverse ratio at your nominal speed.
The decision is driven by package edge requirements, not by ribboning. More modulation amplitude widens the effective lay zone at each end, which softens the cheese edge — fine for dye packages where you want dye penetration, bad for magnet-wire spools that have to slot into automatic feeders with hard square edges.
For magnet wire and welding wire, stay at ±2 to ±3 mm. For textile dye packages, ±4 to ±6 mm is standard. For optical fibre, you generally go below ±1 mm and rely on velocity modulation instead because edge density matters less than micro-bend control on inner layers.
0.5% drift over a full spool build is at the low end of what a healthy modulation system should produce. Typical designs run 1 to 1.5% peak-to-peak swing over the modulation period, which is usually 30 to 120 spindle revolutions, and the swing should repeat consistently spool to spool.
If the drift is monotonic — always increasing or always decreasing through the spool — that is not modulation, that is something stretching or wearing. Check the traverse drive belt tension first, then the differential pinion bearings.
You can retrofit if the head has a separable traverse drive — meaning the helix shaft is driven by its own gear train you can break into. The retrofit involves adding a modulation cam coaxial with the existing traverse shaft and a phase-lock defeat pinion in the drive train. Budget around $2,000 to $4,000 in parts per head plus a day of fitter time.
If the traverse is driven by the same shaft as the spindle through a hard mechanical link — common on older Schlafhorst and Savio machines — a retrofit is rarely worthwhile. At that point a servo-driven programmable head is the better path.
This is almost always a lay-angle problem, not a ribboning problem. The traverse ratio can be ideal, the package can look square and clean, but if the lay angle at the package surface is too steep (above about 18° on fine wire), adjacent turns dig into each other under tension and lock the payout.
Check the velocity-modification segment of the cam — that is the section that flattens the lay angle through the centre of the stroke. A worn or wrongly-profiled velocity segment lets the centre-of-stroke lay angle climb. Measure the helix angle at three points across the package face; if the centre is more than 2° steeper than the ends, the velocity cam needs replacement.
Massively, but in opposite directions. On a soft textile package, end dwell builds the cheese edge that resists collapse on the pallet — too little dwell (under 6 ms) and the package edges sag during transport. On a hard wire spool, too much dwell (over 12 ms on fine magnet wire) builds a knife-edge that flakes the enamel coating during payout, and you'll see the customer reject spools for insulation faults.
The 8 to 15 ms range in the article is the universal envelope; within that, soft yarn lives at 12 to 15 ms and fine magnet wire lives at 6 to 9 ms. Adjust by changing the dwell tappet — never by changing traverse frequency, because that breaks the ratio.
References & Further Reading
- Wikipedia contributors. Bobbin winding. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
