Multi-stage telescoping lift where each stage extends a fixed multiple of the input. Used when the team needs maximum reach from a small collapsed footprint. Read Lift Mechanisms first if you have not chosen a lift type yet, and Climbing & Elevation if your goal is endgame elevation.
Read this first. Cascade lifts are not a beginner mechanism. If you are choosing your first lift, build a four-bar or DR4B instead. Cascades become worth it when the game requires more vertical reach than a parallelogram lift can provide in the available width — usually 3+ feet of extension from a collapsed height under 18 inches.
Why the Cascade Exists
A four-bar or DR4B has a fixed reach equal to roughly the bar length of its stages. A 24-inch four-bar reaches about 24 inches up. A DR4B with 18-inch bars reaches about 36 inches. To reach higher, you make the bars longer — but longer bars mean more weight, more torque demand at the pivot, and more tipping moment when extended forward.
The cascade solves this by stacking telescoping stages that each move proportionally. A two-stage cascade where the inner stage moves 2 units for every 1 unit the outer stage moves gives you total travel = (1 + 2) × outer stroke. A three-stage cascade with the same 2:1 multiplier per stage gives (1 + 2 + 4) × outer stroke. The collapsed height stays small; the extended height grows fast.
The tradeoff is mass and complexity. Each stage adds its own rails, slides, retention hardware, and string routing. The mass at the top of the lift grows linearly with stages, but the moment about the chassis grows by stage-height × stage-mass — so center of gravity climbs aggressively.
Single Stage
Reach = bar length Simple but limited
2-Stage Cascade
Reach = 3× stroke Most-common config
3-Stage Cascade
Reach = 7× stroke Maximum reach, max complexity
Where the Cascade Wins
Tower Takeover (2019–20) — cube stacks reached 7+ tall on tower-top scoring. Cascade lifts dominated tower stacking; the geometry favored vertical stroke over horizontal placement.
Tipping Point (2021–22) — high-platform mobile-goal placement on the platform during endgame. Top teams used cascades to lift mogos onto tall platforms.
Over Under (2023–24) — some elevation-bar grabs used cascades to extend a hook above the bar. Niche but effective when geometry forced it.
Modern descendants — whenever a game requires reach above ~5 feet from the floor in a robot that starts at 18 inches, the cascade becomes a primary candidate. Override may or may not call for it — the manual on Monday April 27 will tell us.
Where It Loses
Speed-of-cycle games — a cascade has more mass at the top than a four-bar, so acceleration limits are tighter. If the game scores by rate (cycles per minute), a faster simpler lift wins.
Low-ceiling games — if scoring caps below 3 feet, a four-bar or chain bar reaches it with less mass and less complexity. Cascade is overkill.
Forward-reach games — cascades extend mostly vertically. A six-bar or chain bar that places the load 12+ inches forward of the chassis is better for goals you reach over.
Beginner teams — the build is finicky. String routing, stage retention, and synchronized motion all have to be right. A first-year team will spend three weeks tuning instead of competing.
Override caveat. Override is being revealed Friday April 24 with the manual on Monday April 27. Do not commit to a cascade build before reading the manual. Cascades are high-leverage when the geometry calls for them and a serious mistake when it does not. Wait for the scoring heights, field dimensions, and endgame structure to be confirmed before building.
1 of 8
// Section 02
Cascade Geometry 📐
The math behind why each stage moves a fixed multiple of the input, and how to size stages for a target reach.
The Cascade Multiplier
A cascade works by anchoring a string to the base, looping it over a pulley on each stage, and tying it to the next stage. As the outer stage rises, the string's effective length on each side of every pulley changes — pulling the inner stages up in proportion.
The key equation is the stage multiplier. For a continuous-string cascade where each stage uses one pulley:
Stage k motion = k × outer-stage motion
Stage 1 (outer): moves 1 unit per unit input
Stage 2: moves 2 units per unit input
Stage 3: moves 3 units per unit input (3-stage)
Total reach = (1 + 2 + 3 + ... + n) × outer stroke
For a 2-stage cascade with 24-inch outer stroke: total reach = (1 + 2) × 24 = 72 inches above the base anchor. For 3-stage with 18-inch outer stroke: (1 + 2 + 3) × 18 = 108 inches. The reach grows fast with stage count, but so does mass.
Why the Multiplier Is What It Is
Picture a 2-stage cascade. Stage 1 (outer) sits in the base. Stage 2 (inner) sits inside stage 1. A string anchors at the chassis base, goes up over a pulley at the top of stage 1, and ties off to the bottom of stage 2.
When stage 1 rises by 1 inch, the pulley at the top of stage 1 also rises by 1 inch. The string on the chassis-side of the pulley gets shorter by 1 inch (because the pulley moved up). The string is fixed length, so the stage-2 side of the pulley must also gain 1 inch — meaning stage 2 rises 1 inch relative to stage 1. Stage 2's absolute motion is its motion relative to stage 1 (1 inch) plus stage 1's motion (1 inch) = 2 inches per inch of input.
For 3 stages, the same logic compounds: stage 3 rises 1 inch relative to stage 2, which is rising 2 inches absolute, so stage 3 rises 3 inches per inch of input. Each stage adds 1 to its predecessor's multiplier.
Sizing Worksheet
Three numbers drive the design:
Target reach (R) — how high above the base anchor the top stage's effector needs to go. Read this from the game manual.
Collapsed height (H) — how tall the lift can be when stowed. Bounded by the 18-inch starting size and any in-match size limit.
Stage stroke (S) — how far each stage can extend before mechanical stops or string runs out. Roughly H minus pulley/hardware overhead, typically 0.6–0.7 of H.
Given those, the stage count needed is the smallest n such that (1 + 2 + ... + n) × S ≥ R, or equivalently n(n+1)/2 ≥ R/S.
Example: 60-Inch Reach From an 18-Inch Robot
1
Target R = 60 inches. Need to reach scoring height of 60 inches above the floor (a hypothetical 5-foot scoring zone).
2
Collapsed H = 17 inches. Leaves 1-inch margin under the 18-inch start size.
3
Stroke S = 12 inches. Roughly 0.7 of H after subtracting pulley housings and stage caps.
4
R/S = 60/12 = 5. Need n(n+1)/2 ≥ 5, so n = 3 (gives 6). A 3-stage cascade reaches (1+2+3) × 12 = 72 inches — 12 inches of headroom over the target.
5
Sanity check. Could a 2-stage do it? (1+2) × 12 = 36 inches. No. Could a 4-stage with shorter stroke do it? (1+2+3+4) × 6 = 60 inches exact. Yes, but with 4 stages worth of mass and complexity for the same reach. 3 stages is the sweet spot.
Always design with 10–20% reach headroom. Real-world stroke is shorter than spec by hardware overhead. Field tolerances, slight wheel wobble, and the load you're holding all eat into effective reach. If you need exactly 60 inches and your math says exactly 60 inches, you will not reach it on game day.
2 of 8
// Section 03
String Routing 🧶
The actual cable path that turns one motor into synchronized multi-stage extension. Get this wrong and the lift binds or stages stick.
Two Routing Patterns
VRC cascades use one of two string topologies. Pick before you start cutting parts.
Continuous-String (Single Loop)
One length of string anchors at the chassis, runs up the lift through pulleys at the top of each stage, and ties off at the top of the inner-most stage. The motor pulls the chassis end of the string. As the motor pulls, the string shortens between chassis and the first pulley, extending stage 1; meanwhile the path from stage 1's pulley to stage 2's anchor also shortens, extending stage 2; and so on. This is the most common VRC pattern.
Pros: one motor input drives all stages. No synchronization issues between stages. Mechanical guarantee that all stages move proportionally. Cons: the entire string is under tension during the entire stroke. String stretch compounds across stages. A snapped string drops the entire load.
Discrete-Pulley (Per-Stage)
Each stage has its own string and motor (or its own segment of a multi-output gearbox). Stage 1 lifts directly; stage 2 lifts off stage 1; stage 3 lifts off stage 2.
Pros: independent control. Easier to debug — if a stage is stuck, you know which one. Less compounded stretch. Cons: needs synchronization between motors or a multi-output gearbox. More motor budget. Heavier. Almost no VRC team uses this; it's an FRC pattern.
For VRC, use continuous-string. The motor budget alone rules out discrete-pulley unless you have a very specific reason. Spend your debug time on string anchoring, pulley alignment, and stage retention — not on synchronizing motors.
Continuous-String Path, Step by Step
For a 3-stage cascade:
1
Anchor 1: chassis base. Tie one end to a fixed point on the chassis frame, near the bottom of stage 1's rail. This is the load-bearing anchor — it holds the entire lift weight. Use a screw clamp or knot through a metal eye, not just a knot to plastic.
2
Up to pulley A: top of stage 1. String runs vertically up the inside of stage 1 to a pulley mounted at stage 1's top.
3
Down through stage 2. Over pulley A, string drops down the inside of stage 1, alongside stage 2.
4
Anchor 2: bottom of stage 2. String ties to the bottom of stage 2. As stage 1 rises, this anchor pulls stage 2 up relative to stage 1.
5
Up to pulley B: top of stage 2. A second length of string (or the same string continued) runs up the inside of stage 2 to a pulley at stage 2's top.
6
Down to anchor 3: bottom of stage 3. Same pattern. Each stage k has a pulley at its top and an anchor at its bottom on stage k+1.
7
Motor at the chassis pulls. The lifting motor pulls the chassis-anchor end of the string through a winch drum or pulley. As string spools onto the drum, all stages extend together.
Pulley and String Choice
Pulleys: use VEX 3-inch or 4-inch nylon pulleys. Larger pulleys reduce string bend radius, which extends string life dramatically. Avoid running string over a sharp edge or a small-radius shaft.
String: Spectra/Dyneema-equivalent braided line in 50–100 lb test. Avoid nylon — it stretches. Avoid kevlar — it cuts on bends. Spectra holds the load and resists stretch through 100+ cycles.
Anchoring: bowline knot or figure-8 follow-through, then heat-shrink the tail. Never glue the knot. Glue makes the string stiff at the load point and fails under cyclic load.
String stretch is real. Even spectra stretches 1–2% under load. On a 3-stage cascade with 4 feet of total string under tension, that's 0.5–1 inch of unwanted droop. Pre-tension the system at first build by extending fully and letting it sit overnight under static load. Re-tension (or replace) the string every 10–20 competition matches.
3 of 8
// Section 04
Build Order 🔨
Sequence the build so each stage is verified before stacking the next. Saves three days of debugging in week 4.
Build one stage at a time and test it. The single biggest cause of dead cascade builds is teams who build all 3 stages in parallel and try to integrate at the end. By the time they discover stage 1 binds, all the stage 2 work is wasted.
Phase 1 — Stage 1 Alone (Days 1–3)
1
Build the rails. Two parallel C-channel or U-channel rails, vertical, mounted to the chassis. The inner profile must accept stage 2's outer dimension with 1–2 mm clearance per side.
2
Add stage 1 slides. Linear bushings, slider blocks, or roller bearings on the inside of the chassis rails. Slide should glide smoothly through full stroke under finger-pressure with no bind.
3
String it up with no payload. Single string, anchor at chassis, pulley at top of stage 1, anchor at top of stage 1. Motor lifts stage 1 alone. Verify smooth motion, no bind, no overshoot.
4
Cycle 20 times. Up, down, up, down. If anything sticks, fix the rail/slide alignment now. Do not move on until stage 1 cycles 20 times reliably.
Phase 2 — Add Stage 2 (Days 4–6)
5
Build stage 2 rails. Same C-channel pattern, sized to fit inside stage 1 with clearance. Add slides to the outside of stage 2 that mate with the inside of stage 1.
6
Add the second pulley + string. Pulley at top of stage 1, second string anchored to chassis (or continued from stage-1 string), running up over pulley A, down to anchor at bottom of stage 2.
7
Test 2-stage motion. Lift the assembly. Stage 2 should extend twice as fast as stage 1 (geometry says so). Verify visually.
8
Cycle 20 times. Same drill. Watch for stage-2 binding, wobble, or string slipping off the pulley. Fix before stacking stage 3.
String stage 3. Pulley at top of stage 2, anchor on stage 3 bottom. Watch the total string path now — over 6 pulleys for a 3-stage continuous-string cascade. Friction adds up.
11
Test full extension. Up. All three stages extend together: stage 1 by S, stage 2 by 2S, stage 3 by 3S. Total reach = 6S.
Phase 4 — Effector and Tuning (Days 11–14)
12
Add the effector. Whatever picks up or scores at the top — a clamp, a flap, a passive hook. Mount to the top of stage 3.
13
Test under load. With effector and a representative payload (game element mass), cycle 30 times. Watch for string stretch, slide wear, payload tipping the lift.
14
Add stops. Mechanical hard stops at full extension and full retraction. The lift must not be able to pull a stage off its rails or drive past the string anchor.
4 of 8
// Section 05
Gear Ratios ⚙
Cascades demand torque, not speed. The geometry already multiplies position; the motor needs to provide the force.
The Force Multiplier Works Backward
Position multiplies up: outer stroke 12 inches becomes 36 inches at the top of a 3-stage cascade. But by the conservation of energy, force multiplies down. If the payload at the top weighs 2 lbs, the motor pulling the chassis-end string must provide 2 × (1+2+3) = 12 lbs of pull force.
This is the cascade's biggest design constraint. A four-bar lifting 2 lbs at the end effector needs a motor providing only the torque to overcome that 2 lbs of moment. A 3-stage cascade lifting the same 2 lbs needs the motor to provide the equivalent of 12 lbs of payload — six times more.
Force at motor = total payload × sum-of-multipliers. For an n-stage cascade lifting weight W:
F_motor = W × n(n+1)/2
2-stage: F = 3W | 3-stage: F = 6W | 4-stage: F = 10W
The V5 200-RPM cartridge produces about 2.1 N·m of stall torque. With a winch drum of 0.75-inch radius (1.5-inch diameter), that's ~22 lbs of pull force at stall — enough to lift a 3-lb cascade payload through 6× multiplication with margin.
7:1
2-stage cascade, light payload. Motor input 7× reduced. Plenty of speed; lift extends in under 1 second.
15:1
3-stage cascade, light payload. The sweet spot for most VRC cascades. Reasonable speed (~1.5 sec full extension) with torque headroom.
25:1
3-stage cascade, heavy payload. Use when the load + lift mass approaches motor stall. Slower (~2.5 sec) but reliable.
36:1
4-stage or unusually heavy. Last-resort ratio. If you need this, reconsider whether 4 stages was the right call.
Two-Motor Configurations
Cascades over 4 lbs of payload usually run two motors geared together to share the load. Both motors drive the same winch drum through a synced gear train. This doubles available torque without increasing motor count beyond what the team can spare.
If you run two motors, they must be on opposite sides of the chassis to balance moment, and they must be wired through the V5 brain's motor-pair group so they stay synchronized when one pulls harder than the other.
Test stall current early. Run the lift at full payload at slow speed and watch the motor current draw on the V5 brain screen. If you're seeing >2 A continuous, the motor is working too hard and will overheat in match conditions. Either gear down further or shed weight at the top.
5 of 8
// Section 06
Integration with the Robot 🔗
A cascade lift constrains the rest of the robot more than most mechanisms. Plan the chassis around it.
Center of Gravity Climbs Fast
At full extension, the cascade puts a meaningful fraction of its mass at the highest point of the robot. A 3-stage cascade with 1 lb per stage and a 2-lb effector has 5 lbs total, with effective center of gravity at roughly 60% of the extended reach. For a 60-inch extension, that's ~36 inches above the chassis — well above the wheelbase.
The chassis must be wide enough that the moment from the extended lift cannot tip the robot. Practical rule: wheelbase width ≥ (extended height) / 2. For a 60-inch extension, that means a wheelbase at least 30 inches wide. Most VRC chassis are 17–18 inches wide, so cascades over 36 inches of reach are tipping-prone unless the team adds outriggers or a counterweight.
Drive Speed Limits
Driving a fully-extended cascade at full speed will tip the robot. Most cascade-equipped teams cap drive speed (in code) when the lift is above a threshold height. A simple proportional cap — max drive output decreases linearly as lift height increases — gives drivers control without surprises.
Pseudocode pattern (do not paste C++ in your notebook per EN4):
Read lift height from sensor (rotation sensor or pot).
If height > threshold, reduce drive_max by (height − threshold) / (max_height − threshold).
Apply drive_max in the joystick-to-motor mapping.
Result: drivers get full speed when lift is down, tapered speed when lift is up.
Position Sensing
The cascade needs a sensor for two reasons: closed-loop position control to scoring heights, and the drive-speed cap above. Three options:
Rotation sensor on the winch shaft. Most common. Cheap, accurate, and the rotation count maps cleanly to lift height through the gear ratio. Calibrate by extending fully, recording the count, retracting fully, recording the count, and storing those as height endpoints.
Potentiometer on a stage pivot. Works if a pivot rotates as the cascade extends (rare for pure cascades, common for hybrid mechanisms). 333-degree V2 pot gives 12-bit resolution.
Limit switches at extension endpoints. Always include these regardless of which primary sensor you use. Limit switches are the safety stop that prevents the lift from driving past its mechanical limits if the rotation sensor reads wrong.
Fits With What Else
Intake mechanisms. The cascade carries the effector up; the intake feeds it. Most cascades pair with a roller intake at the chassis or a transfer mechanism that hands payload up to the effector.
Drivetrain. 4-motor or 6-motor drives are fine; reserve 2 motors for the cascade unless payload is light.
Other lifts. Cascades occupy the vertical space at the chassis center. A second lift on the same robot is rare and almost never necessary.
The five failures every cascade hits in week 3. What causes them, and what to change.
Stage Binding
One stage drags or stops mid-stroke. Often appears at full extension first, then earlier in the stroke as wear progresses.
Fix: rail-to-slide alignment. Check that rails are parallel within 1 mm over the full stroke. Re-shim or re-mount slides. Apply dry PTFE lubricant — never grease (catches dust).
String Slip
String walks off the pulley mid-stroke. Lift stalls or a stage stops moving. Can also slip off the winch drum and unspool catastrophically.
Fix: add side flanges to all pulleys (most VEX pulleys ship with these — do not remove). Add a string keeper bracket. Use a deeper-groove winch drum.
String Stretch
Lift droops at full extension. Cycle-to-cycle position becomes inconsistent. Effector positions wrong by 1–2 inches.
Fix: use Spectra/Dyneema, not nylon. Pre-tension the system overnight at first build. Re-tension every 10–20 matches. Replace string entirely every 50 matches.
Tipping at Full Extension
Robot leans or tips when lift is fully up. Drivers cannot turn or accelerate without losing the load.
Fix: drive-speed cap in code (see Integration section). Add counterweight low on the chassis if mass budget allows. Consider whether the lift needs to be this tall — if a 2-stage works, drop a stage.
Asymmetric Extension
Stages extend at different rates than the geometry predicts. Stage 2 lags or stage 3 outruns it. Lift bows mid-stroke.
Fix: string anchor and pulley locations. Each pulley center must be on the geometric centerline of its stage; each anchor must be at the bottom of the next stage. A 2-mm offset accumulates by the top of a 3-stage.
Tuning Sequence at the Field
Visual stroke check (no payload). Lift up. Watch each stage move. Are all three stages moving at their expected ratios? If not, debug string routing before payload tuning.
Stall current at midstroke (no payload). Hold lift mid-extension under closed-loop control. Read motor current. Should be <1 A. Higher means binding.
Position repeatability (with payload). Drive to scoring height 5 times. Measure where the effector lands each time. Spread should be <0.5 inches. Higher means string stretch or sensor drift.
Endurance test. 50 cycles back-to-back at competition speed under representative payload. Watch for any new noise, wear, or droop. If anything changes between cycle 1 and cycle 50, the build is not match-ready.
Document everything in the notebook. Cascade tuning is methodical — record string tension, sensor calibration values, and stall current readings every time you re-tune. See Sensor Notebook Templates for the calibration log pattern.
7 of 8
// Section 08
CAD & Verification 🖌
Cascade design lives or dies on tolerances. CAD before cutting saves you from week-3 rebuilds.
What to CAD Before Building
Rail cross-sections, all stages. The clearance between rails and slides at each stage interface is the critical dimension. CAD it; verify before machining; do not eyeball it.
Pulley positions. Each pulley centerline relative to the stage centerline. Pulley height relative to stage top.
String path simulation. Onshape can simulate string motion through pulleys. Verify the string does not bind, double back, or wrap incorrectly at any stroke position.
Stop locations. Where do the mechanical hard stops sit at full extension and full retraction? CAD them in.
Center of gravity sweep. Plot CoG height vs lift extension. The slope of this curve is what determines the drive-speed cap.
Why a cascade: "We chose a cascade because the game required reaching X inches above our chassis, and a parallelogram lift would have needed bars longer than our chassis is wide. The cascade gives us 6× mechanical advantage in stroke length per inch of collapsed height."
How we tuned it: "Stage-by-stage. We built and verified stage 1 alone for 20 cycles before stacking stage 2, then verified stage 2 alone before stacking stage 3. Each stage's string and pulley alignment was the failure point; debugging at the integration step would have been three times slower."
Tradeoffs we accepted: "The cascade has more mass at the top than a four-bar, which limits our acceleration when the lift is extended. We compensate by capping drive output proportionally to lift height in code. We also accept slower full extension — ~1.5 seconds with our 15:1 ratio — in exchange for the reach."
Check for Understanding
A 3-stage cascade extends correctly when running unloaded but bows mid-stroke under payload. Stage 2 lags behind where the geometry predicts. Most likely cause?
Motor torque is insufficient for the load
Gear ratio is too high
String stretch is asymmetric or one anchor is slipping under load
Stage 2 rails need more lubricant
Why: Motion under payload but not unloaded points to a load-bearing failure, not a friction or torque problem. Motor torque insufficiency would show as the lift not reaching full extension at all (it does — just bowed). The geometry is mechanical, not driven by gear ratio. Stage rails being clean was already verified by the unloaded test. The remaining cause is something that fails only under tension — which means the string side: an anchor slipping, or one segment of the string stretching more than the others. Replace the string and re-tension.