Skip to content
Counterspringing.com

Selection considerations

What to weigh before you choose

Ten things your team has to establish and argue about, what each one does to each method, and — for the ones that can end the argument on their own — a warning saying so.

The considerations

Premise

Settle this before anything else. It doesn't favor a method — it decides whether you build one at all.

Premise

The problem you are actually solving

What to establish
Name the measured symptom, and the number that would tell you it's fixed. Not “the lift feels slow” — a cycle time, an amp draw, a stall.
Why its weight changes
This one doesn't compete with the others for weight; it sits above them. Every consideration below assumes the answer here is a real, measured problem. If it isn't, the rest of the page is a discussion about which unnecessary mechanism to add.

Can decide this alone. If nothing here is a measured bottleneck, this consideration ends the debate on its own — the answer is no counterspring, and none of the trade-offs below get a vote.

Lift speed is a bottleneck

Cycle time is limited by how fast the lift travels under load.

Pushes no method either way.

Current draw is starving other subsystems

The lift's draw browns out or slows the rest of the robot.

Pushes no method either way.

Holding position burns motor power

The lift spends match time held at height under power.

Pushes no method either way.

No measured problem — exploring

Worth doing as an offseason learning project. Not worth doing to a competition robot.

Pushes no method either way.

Givens

A fact about the lift you already have or have already committed to. It sets the shape of the load you must match.

Given

Stringing architecture

What to establish
Whether the lift is continuous or cascade. Trace one string from the spool to the carriage if you're unsure.
Why its weight changes
This decides the shape of the load — a staircase or a flat line — and springs are far better at flat lines than staircases. It is the consideration most likely to make a method that suits your team badly wrong for your lift anyway.

Continuous

Stages extend one after another. The most common arrangement.

  • near-decisively favors: Constant Force — Per-stage constant-force springs map directly onto a continuous lift's stepped gravity load.
  • mildly disfavors: Constant Torque — A single constant-torque unit gives one force level, which under- or over-assists individual stages of a continuous lift.
  • mildly disfavors: Sprung Linkage — Matching a stepped load is the hardest profile for a linkage to shape.

Cascade

All stages extend together at different speeds.

  • significantly favors: Constant Torque — Cascade's near-constant reflected load at the winch suits a single constant-force level.
  • mildly favors: Bungee — A near-constant target load is easier for a pretensioned cord to approximate.
  • mildly favors: Sprung Linkage — A near-constant target profile is the easiest for a lever + spring to approximate.
  • mildly favors: Constant Force — A single spring set at the actuation point works; per-stage matching is unnecessary.
Given

Moving mass

What to establish
The mass of the slides and everything riding on them, per stage. Weigh it — CAD mass is a starting estimate, not an answer.
Why its weight changes
Mass sets the magnitude of the load, but it rarely decides the method by itself: most methods scale to most robot-sized lifts. It matters most at the extremes, where it starts pushing methods out of their comfortable range.

Light (< ~1.5 kg)

Small claw or simple bucket.

  • mildly favors: Bungee — Light loads are within the comfortable range of one or two cords.
  • mildly disfavors: Constant Torque — Spring-pack hardware is heavy relative to the benefit on a light lift.

Medium (~1.5–3 kg)

The middle of the range doesn't discriminate — every method is comfortable here.

Pushes no method either way.

Heavy (> ~3 kg)

Arm, wrist, and multiple servos riding the lift.

  • mildly favors: Constant Force — Heavy assemblies benefit most from accurate, stage-matched compensation.
  • mildly disfavors: Bungee — Large loads push bungee toward many parallel cords, which compounds routing friction and wear.

Requirements

Something the counterspring has to do. Requirements are argued about — how much, how well, and whether the game actually asks for it.

Requirement

Unpowered holding

What to establish
Whether the game actually requires the lift to stay put without power, at how many positions, and for how long.
Why its weight changes
This is the requirement most often asserted and least often needed. If it is real, it is close to decisive — it eliminates any method whose force profile drifts. If it isn't, it should carry no weight at all, and teams routinely let it drive the choice anyway.

Must float anywhere without power

  • significantly favors: Constant Force — Near-constant, stage-matched force is what lets a lift float unpowered at any height.
  • significantly disfavors: Bungee — A linear force profile cannot hold every position; the lift drifts wherever the match is off.

Briefly, or at a few positions

A weaker version of the requirement, and one most methods can satisfy. Establish which positions before letting it rule anything out.

Pushes no method either way.

No — the motor can hold it

  • mildly favors: Bungee — If the motor holds position anyway, bungee's approximate profile costs little.
Requirement

Force-matching accuracy

What to establish
How closely the counterspring has to track gravity across the stroke — and what specifically goes wrong if it doesn't.
Why its weight changes
Accuracy is the requirement teams inflate most. Near-perfect matching is a real need for a floating lift and pure cost for everything else; demanding it by default disqualifies the simplest methods for no return. Argue for the accuracy your symptom needs, not the accuracy that sounds rigorous.

Rough assist is fine

The goal is just to reduce current draw.

  • significantly favors: Bungee — If rough assist is the goal, the simplest method wins.
  • significantly disfavors: Sprung Linkage — Custom geometry is wasted effort for a rough assist.

Good match

Small residual force is acceptable.

Pushes no method either way.

Near-perfect

The lift should float at any position.

  • significantly favors: Constant Force — Highest practical accuracy from COTS parts (~1–2% within a spring).
  • mildly favors: Sprung Linkage — Theoretically can match gravity exactly, if the team can execute it.
  • significantly disfavors: Bungee — A linear profile cannot deliver near-perfect matching across travel.

Constraints

A limit on what you can spend to get it: parts, time, skill, space. Constraints don't negotiate, and they rule methods out rather than rank them.

Constraint

Manufacturing capability

What to establish
What your team can actually make this season — not what it could make in principle, or what a mentor could make for you.
Why its weight changes
The hardest constraint on the page, and the one most likely to end the argument early. It does not trade off against anything: no amount of wanting a sprung linkage produces the links. Establish it honestly and first, because it deletes options rather than ranking them.

Can decide this alone. COTS-only is close to a single-answer constraint. It rules out three of the four methods regardless of how well they'd suit the robot.

COTS / hand tools only

  • significantly favors: Bungee — Fully COTS: cord, anchors, and existing holes.
  • significantly disfavors: Constant Force — Spools and shaft mounts realistically require 3D printing.
  • significantly disfavors: Constant Torque — Drums and retention hardware need fabrication.
  • near-decisively disfavors: Sprung Linkage — Custom links and pivots cannot be sourced off the shelf.

3D printing

  • significantly favors: Constant Force — Printed two-part spools are the documented, proven approach.

3D printing + machining / laser cutting

  • mildly favors: Constant Force — Full fabrication access removes the main installation obstacle.
  • mildly favors: Sprung Linkage — Machined or laser-cut links hold geometry far better than printed prototypes.
  • mildly favors: Constant Torque — Accurate drum bores and retention are within reach.

What each method needs before you can build it

Bungee
Hand tools; elastic cord; existing anchor holes or a drill.
Constant Force
3D printer for spools; COTS strip springs, shafts, and mounts; a luggage scale for verification.
Constant Torque
Accurate drums (printed or machined), fully supported shafts, and manufacturer spring data.
Sprung Linkage
CAD linkage design, custom links (printed, machined, or laser-cut), precise pivots.
Constraint

Timeline

What to establish
Whether this lands on a competition robot this season, or is an offseason project with room to fail.
Why its weight changes
In season this is nearly a constraint in its own right, because build time is competing with drive practice and reliability work. In the offseason its weight drops to almost nothing — the same method that is reckless in January is the right project in June.

Can decide this alone. In season, this can rule out the two least-documented methods on its own, whatever the robot wants.

During the competition season

  • near-decisively disfavors: Sprung Linkage — Linkage geometry iteration is hard to bound; this is a classic offseason project.
  • significantly disfavors: Constant Torque — Thin documentation makes in-season debugging risky.
  • mildly favors: Bungee — Fastest to install and fastest to fix at competition.

Offseason project

  • significantly favors: Sprung Linkage — Offseason time is exactly what geometry design and tuning need.
  • mildly favors: Constant Torque — Time to research turn budgets and drum sizing without schedule pressure.
Constraint

Team experience

What to establish
Whether your build team has designed custom mechanisms before — and whether the students, not the mentors, will do it.
Why its weight changes
Experience is a constraint that is easy to overstate in both directions. It does not make an unsuitable method suitable, and inexperience does not forbid an ambitious one in the offseason. It mostly governs how much debugging risk you can absorb when the mechanism misbehaves.

Rookie / first custom mechanisms

  • significantly favors: Bungee — Simplest mental model and assembly; failures are visible and cheap.
  • significantly disfavors: Sprung Linkage — Requires linkage synthesis and structural design experience.
  • mildly disfavors: Constant Torque — Coupled drum/spring design decisions with little reference material.

Comfortable with custom lifts

The middle of this range doesn't push any method.

Pushes no method either way.

Advanced — custom mechanisms are routine

  • mildly favors: Sprung Linkage — An experienced team can realistically execute custom geometry.
Constraint

Packaging space

What to establish
The volume actually free around the slides and at the winch, with the intake, wiring, and the rest of the robot already in the CAD.
Why its weight changes
Packaging is worth little until it's worth everything. A lift with room to spare should ignore it; a lift where the envelope is already full lets it delete methods outright, and no force-profile argument survives a part that doesn't fit.

Tight — every millimeter is spoken for

  • significantly disfavors: Sprung Linkage — Swept linkage volume is the first casualty of tight packaging.
  • mildly favors: Bungee — Cord routed along existing structure adds almost no envelope.
  • mildly favors: Constant Torque — Concentrating hardware at the winch can free the slide envelope — if the winch area has room.

Some room along or at the base of the slides

Pushes no method either way.

Open — space is not a constraint

Then this consideration should carry no weight in your debate.

Pushes no method either way.

Constraint

Tuning and maintenance tolerance

What to establish
How much recurring adjustment the team will actually do — measured by what it already does, not what it intends to do.
Why its weight changes
The consideration teams weigh most optimistically. Tuning appetite in the design review is not tuning appetite in week six, and a method whose accuracy depends on maintenance that never happens is less accurate than the simpler method you'd have tolerated.

Set it and forget it

  • mildly favors: Constant Force — Once the right springs are verified, there is nothing to re-tune.
  • mildly disfavors: Bungee — Cord creep means periodic re-tensioning and replacement.
  • significantly disfavors: Sprung Linkage — Geometry tuning is the method's defining cost.

Occasional adjustment is fine

Pushes no method either way.

We enjoy iterating

  • mildly favors: Sprung Linkage — A team that enjoys iterating can extract this method's potential.

Once you’ve argued these through, the comparison matrix rates every method against the same criteria — and each method guide’s “Avoid this when” column is the fastest way to eliminate one. Compare methods

Requirements and constraints, at a glance

The same content as above, transposed for a design review: your situation down the side, the methods across the top. Find the rows that describe your robot and your team — most won’t — and argue about those.

This is the opposite axis from the comparison matrix, which rates the methods against fixed criteria. Here the rows are facts about you, and the cells are what those facts do to each method.

Requirements and constraints reference. Each row is a situation your team may be in; each column is a counterspringing method. Cells show whether the situation favors or disfavors the method, and how strongly.
Your situationBungeeConstant ForceConstant TorqueSprung Linkage
PremiseThe problem you are actually solving
Lift speed is a bottleneckPushes no method either way — this one doesn’t discriminate.
Current draw is starving other subsystemsPushes no method either way — this one doesn’t discriminate.
Holding position burns motor powerPushes no method either way — this one doesn’t discriminate.
No measured problem — exploringPushes no method either way — this one doesn’t discriminate.
GivenStringing architecture
Continuousno effectnear-decisively favors: One rating per stage tracks the staircasemildly disfavors: One force level can't track a staircasemildly disfavors: Steps are the hardest profile to shape
Cascademildly favors: Pretension-dominant cord approximates a flat loadmildly favors: One rating at the actuation point covers itsignificantly favors: Flat load at the winch suits one force levelmildly favors: A flat target is the easiest to approximate
GivenMoving mass
Light (< ~1.5 kg)mildly favors: Within range of one or two cordsno effectmildly disfavors: Hardware weighs more than it savesno effect
Medium (~1.5–3 kg)Pushes no method either way — this one doesn’t discriminate.
Heavy (> ~3 kg)mildly disfavors: Many parallel cords; friction and wear compoundmildly favors: Gains most from stage-matched compensationno effectno effect
RequirementUnpowered holding
Must float anywhere without powersignificantly disfavors: A linear profile drifts wherever the match is offsignificantly favors: Near-constant force is what lets a lift floatno effectno effect
Briefly, or at a few positionsPushes no method either way — this one doesn’t discriminate.
No — the motor can hold itmildly favors: Approximate profile costs nothing if the motor holdsno effectno effectno effect
RequirementForce-matching accuracy
Rough assist is finesignificantly favors: Simplest method wins outrightno effectno effectsignificantly disfavors: Custom geometry is wasted effort
Good matchPushes no method either way — this one doesn’t discriminate.
Near-perfectsignificantly disfavors: Linear profile can't match across travelsignificantly favors: Highest practical COTS accuracy (~1–2%)no effectmildly favors: Can match exactly — if you can execute it
ConstraintManufacturing capability
COTS / hand tools onlysignificantly favors: Cord, anchors, existing holessignificantly disfavors: Spools realistically need printingsignificantly disfavors: Drums need fabricationnear-decisively disfavors: Custom links can't be sourced
3D printingno effectsignificantly favors: Printed two-part spools are the proven pathno effectno effect
3D printing + machining / laser cuttingno effectmildly favors: Removes the main installation obstaclemildly favors: Accurate bores and retention in reachmildly favors: Machined links hold geometry
ConstraintTimeline
During the competition seasonmildly favors: Fastest to install and fix at competitionno effectsignificantly disfavors: Thin documentation makes debugging riskynear-decisively disfavors: Geometry iteration is hard to bound
Offseason projectno effectno effectmildly favors: Time to research drums without pressuresignificantly favors: Exactly what geometry tuning needs
ConstraintTeam experience
Rookie / first custom mechanismssignificantly favors: Simplest model; failures visible and cheapno effectmildly disfavors: Coupled decisions, little reference materialsignificantly disfavors: Needs linkage synthesis experience
Comfortable with custom liftsPushes no method either way — this one doesn’t discriminate.
Advanced — custom mechanisms are routineno effectno effectno effectmildly favors: Can realistically execute custom geometry
ConstraintPackaging space
Tight — every millimeter is spoken formildly favors: Adds almost no envelopeno effectmildly favors: Frees the slide envelope — if the winch has roomsignificantly disfavors: Swept volume is the first casualty
Some room along or at the base of the slidesPushes no method either way — this one doesn’t discriminate.
Open — space is not a constraintPushes no method either way — this one doesn’t discriminate.
ConstraintTuning and maintenance tolerance
Set it and forget itmildly disfavors: Creep means periodic re-tensioningmildly favors: Nothing to re-tune once verifiedno effectsignificantly disfavors: Geometry tuning is the defining cost
Occasional adjustment is finePushes no method either way — this one doesn’t discriminate.
We enjoy iteratingno effectno effectno effectmildly favors: Iteration extracts the method's potential

+ favors a method, − counts against it. More symbols means the situation pushes harder: + mild, ++ significant, +++ near-decisive.

These do not add up. They are not a score with the arithmetic left as an exercise. A single −−− in a row that applies to you outranks any number of +elsewhere, and a row that doesn’t describe your robot is worth zero no matter what it says.

Where to take the argument next