2.1 Machine Architectures
Machine architecture forms the foundation of any Surface Mount Technology (SMT) line. It establishes the critical balance between raw placement speed, component flexibility, and capital investment. The architecture you choose ultimately determines how effectively your line can adapt to changes in product mix and your ability to meet your target Takt time.
The Two Core Architectures: Gantry vs. Turret
Section titled “The Two Core Architectures: Gantry vs. Turret”Placement machinery is generally categorized by its primary motion system. It’s important to match the machine type to your product mix and production needs. For instance, running a high-mix product line on a machine designed for high-volume production will lead to excessive changeover times, which can significantly impact your margins.
| Feature | Modular/Gantry Systems (Flexible Mounters) | Rotary/Turret Systems (Chipshooters) |
|---|---|---|
| Motion System | X-Y gantry, moving over a stationary board. | Rotary head spins components past stationary vision. |
| Placement Speed | Moderate to High. (Up to 80k CPH per module) | Extremely High. (Often ≥ 100k CPH in pure chip mode) |
| Component Range | Excellent. Handles everything from 01005 passives to large, odd-form, heavy connectors and BGAs. | Limited. Best for small passives (01005 – 0603). Struggles with heavy/large ICs. |
| Placement Accuracy | Superior. Uses linear encoders and dedicated Z-axis control. Essential for ultra-fine pitch (≤ 0.4 mm). | Good, but generally lower for large components due to high speed and rotational inertia. |
| Changeover Time | Fast. Component swap is typically managed via smart feeder carts. | Slow. Fixed feeder banks mean a high penalty for swapping parts not currently loaded. |
| CapEx & Flexibility | High CapEx initially, but excellent scalability (add modules) and flexibility (high-mix capability). | Lower CapEx per placement, but fixed capability and poor high-mix performance. |
For High-Mix/Low-Volume (HMLV) production environments where changeovers happen daily, a Flexible Modular or Gantry system is usually the best choice. Conversely, for High-Volume/Low-Mix production that requires continuous stability, a Turret system delivers an optimal cost per placement.
Line Topologies: Structuring the Physical Flow
Section titled “Line Topologies: Structuring the Physical Flow”Arranging placement machines to match both the component distribution curve and the target Takt time is a critical step in line design.
Role Split (Chipshooter – Flexible)
Section titled “Role Split (Chipshooter – Flexible)”In a role split setup, a Chipshooter (Pick & Place 1) handles the massive volume of small passives. The board then moves to a Flexible Mounter (Pick & Place 2) for the large, complex ICs. This setup is ideal when the Bill of Materials (BOM) is overwhelmingly dominated by thousands of 0402 or 0201 passives (resistors, capacitors) with only a handful of large processors. It ensures the high-speed work is assigned to the fastest asset. However, careful load balancing is essential to avoid creating a bottleneck. If the Flexible Mounter becomes the constraint, the fast Chipshooter will sit idle waiting for the conveyor to clear.
Tandem Split (Flexible A – Flexible B)
Section titled “Tandem Split (Flexible A – Flexible B)”A tandem split uses two flexible modular mounters of identical capability running in series, splitting the component count evenly. This is a highly resilient default setup for HMLV production. It simplifies feeder staging since parts can run on either machine, and provides line redundancy if one head requires maintenance. The primary risk here is asymmetry. If all complex parts are placed on the second machine, the entire line is constrained by that single asset. Symmetric programming is highly recommended to maintain a smooth flow.
Dual-Lane / Parallel Processing
Section titled “Dual-Lane / Parallel Processing”In a dual-lane setup, the entire line processes two PCBs simultaneously on two separate conveyor tracks. This architecture is particularly useful for extremely short boards or highly aggressive Takt time requirements where physical placement time is the hard limit. However, this approach requires exactly twice the number of feeders, or meticulously mirrored feeder setups. This significantly increases the complexity of material tracking and kitting, as well as the capital expenditure.
Load Balancing and Throughput Management
Section titled “Load Balancing and Throughput Management”The line Takt time is always defined by the slowest process step, which becomes the manufacturing constraint. In any Pick & Place line, the machine with the highest placement time must be carefully managed.
Start by isolating the cycle time. Calculate the placement time per board for each machine, specifically excluding board travel and transfer time. This serves as the primary balancing metric. The goal is to balance the cycle times of all placement machines in the line to within ±10% of each other. If one machine is slower, you can re-allocate lower-complexity, higher-count components—like common pull-up resistors—back to the faster machine until the cycle times equalize.
To maximize uptime, prioritize permanent feeder banks. Keeping 80% of common components, such as standard 0402 passives and common diodes, in permanently fixed slots across all programs greatly reduces the time and risk involved in every product changeover.
Feeder Management and Traceability
Section titled “Feeder Management and Traceability”Feeder capacity and changeover efficiency are critical drivers of machine uptime.
The chosen machine architecture governs the total feeder capacity per footprint, which establishes a physical limit on the maximum component complexity a single board can possess without forcing a mid-run change.
Using intelligent feeders is highly recommended in modern operations. These feeders electronically communicate their Part ID and loaded slot position to the machine’s software, preventing common kitting errors like placing the wrong part. To minimize changeover downtime, utilize dedicated kitting carts for offline staging. The feeders for the next scheduled job can be prepared, verified, and loaded onto trolleys completely offline while the current job is still running on the machine.
Recap: Machine Architecture Selection
Section titled “Recap: Machine Architecture Selection”| Parameter | Requirement | Gantry/Modular System | Turret/Rotary System |
|---|---|---|---|
| Placement Speed (CPH) | Match target Takt time. | ≤ 80,000 per module. | ≥ 100,000 (chip mode). |
| Component Range | Must handle all BOM parts. | 01005 passives to large BGAs/connectors. | 01005 to 0603 passives only. |
| Placement Accuracy | ≤ 0.4 mm pitch requires superior accuracy. | Superior (linear encoders, Z-axis control). | Good (lower for large/heavy ICs). |
| Changeover Time | Minimize for high-mix. | Fast (smart feeder carts). | Slow (fixed feeder banks). |
| Line Architecture | Balance cycle times within ±10%. | Use for HMLV, tandem splits, or role split for ICs. | Use for HVLM, role split for passives. |