ADR-021: Dual-Rod Motion System (Reference Specification)

Status

Accepted

Supersedes

ADR-003: Precision Smooth Rods
ADR-020: Dual-8 Scavenger Variant

Context

Evolution of Thinking

ADR-003 established M10 smooth rods as the reference specification based on: - 2.4× stiffness advantage over M8 - Adequate rigidity for 220mm build volume at 330mm span (see ADR-024) - Single-rod-per-axis simplicity

ADR-020 proposed a “Dual-8 Scavenger” variant after analyzing the Voron Legacy, which achieves acceptable performance with 8mm rods at 200mm build volume. The key insight was that dual rods sharing load halve deflection, potentially enabling 8mm rods at Amalgam’s larger build volume.

The Physics Discovery

Beam deflection follows: δ = F × L³ / (48 × E × I)

When two parallel rods share the load equally: - Each rod carries F/2 - Deflection: δ_dual = (F/2) × L³ / (48 × E × I) = δ_single / 2

Dual 8mm rods achieve LESS deflection than single 10mm rods:

Configuration	Deflection @ 360mm, 280g	Verdict
Single 8mm	0.066mm	Excessive
Single 10mm	0.027mm	Acceptable
Dual 8mm	0.033mm	Acceptable

While dual 8mm has ~82% the stiffness of single 10mm, the load sharing more than compensates—resulting in better deflection performance.

The Scavengeability Argument

Amalgam already requires two donor printers for triple-Z (ADR-005): - 6-7 NEMA17 motors needed (X, Y×2, Z×3, extruder) - Most donor printers have 4-5 motors - Two donors are the practical minimum

Given two donors are required anyway: - Two Prusa clones provide 12× 8mm smooth rods - M10 smooth rods are rarely scavengeable - Buying M10 adds $60-80 AUD to build cost - Dual 8mm achieves equal or better performance at zero rod cost

The scavenger path became the optimal path.

Decision

We adopt Dual 8mm Smooth Rods with Vertical Stacking as the Amalgam reference specification for XY motion.

Reference Configuration

X-Axis: Two 8mm rods, vertically stacked

═══════════════════════════════════  Top rod
        [Plough carriage]
═══════════════════════════════════  Bottom rod
        ↑ ~30mm spacing ↑

Y-Axis: Two 8mm rods per side, vertically stacked

Left Y-rail:              Right Y-rail:
  ║══ 8mm ══║               ║══ 8mm ══║
  ║  belt   ║               ║  belt   ║
  ║══ 8mm ══║               ║══ 8mm ══║
    LM8LUU                    LM8LUU

Z-Axis: Two 8mm rods (unchanged from ADR-003) - Vertical orientation - Lower load than XY (bed weight distributed across triple-Z)

Why Vertical Stacking?

For pure gravitational sag, vertical and horizontal stacking are equivalent—both share the load equally.

The difference is moment resistance:

Orientation	Resists	Vulnerable to
Horizontal	Yaw (twist L↔︎R)	Pitch (nose-dive)
Vertical	Pitch (nose-dive)	Yaw (twist L↔︎R)

Vertical stacking is preferred because: 1. Pitch forces from Y-acceleration dominate toolhead dynamics 2. Belt runs cleanly between stacked rods 3. Validated by Voron Legacy design 4. X-end brackets are more compact vertically

Rod Count Summary

Axis	Previous (ADR-003)	New Reference
X	2× M10	2× M8 (vertical stack)
Y	2× M10	4× M8 (2 per side, stacked)
Z	2× M10	2× M8
Total XYZ	6 rods	8 rods

Bearing Summary

Location	Previous	New Reference
X carriage	4× LM10UU	4× LM8LUU (recommended)
Y gantry (total)	4× LM10UU	8× LM8UU (4 per side)
Z axis	4× LM10UU	4× LM8UU
Total	12× LM10UU	4× LM8LUU + 12× LM8UU

See ADR-022: Linear Bearing Selection for detailed bearing rationale, scavenged bearing protocol, and alternative options (IGUS, printed bearings).

Consequences

Benefits

True scavengeability: All smooth rods from 2 donor printers
Cost reduction: Eliminates $60-80 AUD rod purchase
Equal or better stiffness: Dual 8mm outperforms single 10mm on deflection
Proven design: Vertical stacking validated by Voron Legacy
Parts availability: 8mm rods/bearings are ubiquitous
Aligned with triple-Z: Already need 2 donors for motors

Trade-offs

More parts: 8 rods + 16 bearings vs 6 rods + 12 bearings
Redesigned brackets: X-ends and Y-pucks need dual-rod mounts
Toolhead mass budget: Keep under 280g for optimal performance
Slightly more complex assembly: More rods to align

What This Enables

Sub-$250 AUD builds: When fully scavenging from 2 donors
$300 target easily achieved: Even with MKS SKIPR + CANbus upgrade
Lower barrier to entry: No need to source obscure M10 precision rods
Better alignment with RepRap philosophy: Use what you have

Alternative Paths (Still Supported)

The parametric CAD system supports alternative configurations:

M10 Single-Rod (Legacy Reference)

SMOOTH_ROD_DIA = 10.0
SMOOTH_ROD_CONFIG = "SINGLE"

For builders purchasing new rods
Fewer parts, simpler assembly
~$60-80 AUD additional cost

M8 Single-Rod (Budget/Compact)

SMOOTH_ROD_DIA = 8.0
SMOOTH_ROD_CONFIG = "SINGLE"

Limited to ~200mm build volume
Single donor may suffice
Marginal deflection at 235mm span

MGN12 Linear Rails (Tinker Path)

LINEAR_MOTION = "MGN12"

For high-speed experiments (200mm/s+)
Requires rail purchase (~$150-200 AUD)
Different bracket designs

BOM Implications

Reference Build: Two Prusa Clone Donors

Purchased: | Part | Cost (AUD) | |——|————| | M10 threaded rod (frame) | $45 | | MKS SKIPR + EBB36 + ADXL | $130 | | E3D V6 clone hotend | $15 | | Belts, pulleys, hardware | $30 | | Total | ~$220 |

Cost Comparison

Configuration	Rod Cost	Total Build
Dual 8mm (reference)	$0	~$220
Single M10	$70	~$290
MGN12 rails	$150	~$370

Implementation Notes

Deflection Thresholds

Based on layer height and quality requirements:

Deflection	Verdict	Suitable For
< 0.015mm	Excellent	All applications
< 0.025mm	Good	Standard printing
< 0.035mm	Acceptable	Most printing
< 0.050mm	Marginal	Draft quality
> 0.050mm	Excessive	Not recommended

Toolhead Mass Budget

For acceptable deflection (< 0.035mm) at 360mm span with dual 8mm:

Toolhead	Mass	Deflection	Verdict
Pitan + pancake NEMA17	~220g	0.025mm	Good
Pitan + standard NEMA17	~280g	0.033mm	Acceptable
Wade + standard NEMA17	~400g	0.047mm	Marginal

Recommendation: Use pancake NEMA17 on extruder for best results.

Span Calculations

For 220mm build volume (reference spec, see ADR-024):

X-rod length = BUILD_X + (2 × CARRIAGE_OVERHANG) + CLEARANCE
            = 220 + 80 + 30 = 330mm

Y-rod length = BUILD_Y + GANTRY_DEPTH + CLEARANCE
            = 220 + 80 + 30 = 330mm

CAD Parameters

# Reference specification (Dual 8mm)
SMOOTH_ROD_DIA = 8.0
SMOOTH_ROD_CONFIG = "DUAL_VERTICAL"
BEARING_TYPE = "LM8UU"  # or "LM8LUU" for long bearings

# Derived
X_ROD_COUNT = 2  # vertically stacked
Y_ROD_COUNT = 4  # 2 per side, stacked
Z_ROD_COUNT = 2
TOTAL_SMOOTH_RODS = 8

Assembly Sequence

Y-rails first: Install dual rods on each Y-rail assembly
Verify parallelism: Rods must be parallel within 0.1mm over length
X-gantry: Mount X-end brackets to Y-rails
X-rods: Install vertical stack, verify spacing
Plough carriage: Install with bearings on both rods
Belt routing: Run belts between stacked rods

Quality Verification

Rod Quality

Spec: Precision ground, hardened steel
Test: Roll LM8UU bearing along rod—should be smooth, no grinding
Measure: Verify 8.00mm ± 0.01mm diameter with calipers

Parallelism Check

Measure rod spacing at both ends
Should be equal within 0.1mm
Adjust bracket mounting if needed

Load Test

With toolhead at center of X travel
Measure deflection with dial indicator
Should be < 0.05mm under static load

References

ADR-003: Precision Smooth Rods (superseded—historical reference)
ADR-005: Triple-Z Independent Kinematic Leveling
ADR-020: Dual-8 Scavenger Variant (superseded—merged into this ADR)
ADR-022: Linear Bearing Selection (companion ADR for bearing choices)
ADR-024: Heated Bed Size Selection (defines 220mm reference build volume)
Voron Legacy: Vertical stacking validation
Beam Deflection Theory
../philosophy.md: “Tractor with a Racecar Brain”
Analysis conversation: “Voron Legacy Comparison” (2025-01-27)