Why We Keep Coming Back to PSG Linear Stages in Semiconductor Tools – Notes from an Applications Engineer

Why We Keep Coming Back to PSG Linear Stages in Semiconductor Tools – Notes from an Applications Engineer

If you’ve ever watched a wafer inspection system slowly lose its edge after a few thousand cycles, you know the linear stage isn’t just a commodity part. I learned this firsthand when a die bonder we were qualifying started showing placement drift in the final 10% of each shift. The root cause wasn’t the vision system or the controller tuning – it was the stage’s inability to maintain repeatability under continuous start-stop loads.

That experience shaped how I look at motion components today. In semiconductor equipment, you typically can’t fix a mediocre stage with clever software. The mechanical backbone either holds up or it doesn’t. Here’s what I’ve found using PSG precision linear stages across several machines, from wafer AOI stations to laser scribing tools.

What Actually Matters in Fab Motion

Forget the datasheet wars for a moment. Most semiconductor applications I deal with share three real-world requirements:

• Repeatability over an entire shift, not just a 10-minute test. Wafer inspection stages need to hit the same focus point at cycle 10,000 as they did at cycle 10. With PSG ball screw stages, we typically see under ±2 µm drift in our in-situ checks, provided the mounting is right. The ground screws and preloaded nuts do their job.

• Vibration that doesn’t mess up the optics. In one AOI tool, we traced image blur at certain speeds to a resonant peak from the stage base. Switching to a PSG stage with a stiffer extrusion and tighter rail preload cleaned it up without needing to slow the scan. That’s when I started paying more attention to the cross-section rigidity, not just the travel specs.

• Sanity in a cleanroom. I don’t want to worry about outgassing or grease creeping onto a wafer. PSG’s standard lubrication options have been compatible with our ISO 6 areas without modification, which is one fewer thing to negotiate with the facilities team.

Where I’ve Put PSG Stages (and Why)

• Wafer inspection / metrology: The key here is smooth, stumble-free motion at slow scan speeds. A ball screw stage can sometimes show a subtle ripple if the screw isn’t good enough, but PSG’s ground screw assemblies have given us clean profiles even at sub-mm/s velocities. For step-and-repeat alignment, the stage’s settling time is predictable, which makes our vision triggers reliable.

• Die bonding and pick-and-place: Speed matters, but I’ve learned the hard way that chasing acceleration numbers without enough frame stiffness just moves the settling problem somewhere else. PSG stages with their solid aluminum base hold the payload rigidly, so we get crisper moves without exciting the gantry.

• Laser dicing / marking: Synchronization with the galvo is everything. Any random position jitter from the stage becomes kerf width variation. Using a direct-coupled ball screw stage instead of a belt drive eliminated an intermittent drift we saw on warm days, simply because the screw’s stiffness doesn’t change with temperature as much as a tensioned belt’s does.

• Test handlers: These machines run literally millions of cycles. After evaluating samples, we chose PSG modules because the rail and screw combination held preload longer, which delayed the inevitable wear-related backlash growth. Maintenance now is mostly re-lubrication, not emergency screw replacement.

Ball Screw vs. Linear Motor – When the Old School Wins

I get asked a lot why we don’t just spec linear motors everywhere. For ultra-high-speed, light-load scanning they’re great. But many of our stations have a heavy Z-axis or a multi-kg chuck, and we need holding force at power-off and resistance to back-driving during E-stops. A ball screw gives us that naturally. PSG’s screw stages also provide a self-locking characteristic at rest that simplifies vertical axis design. Is the top speed lower? Yes. But in 80% of my applications, repeatability and vibration under load matter more than velocity, and the screw wins on both.

Lessons from Selecting and Integrating Stages

If you’re about to specify a stage for a semiconductor tool, here’s what I’d suggest based on mistakes I’ve made:

• Load is more than the payload mass. Calculate the moment created by your fixture plus the wafer, especially if the stage is mounted vertically. A 3 kg load with a 150 mm offset arm turns into a serious moment that eats into the stage’s rated life. PSG stages have good moment capacity data in their specs – use it.

• Don’t chase the smallest encoder resolution. Repeatability often depends more on mechanical stiffness and mounting flatness. I’d rather have a rigid ball screw stage with a 1 µm encoder than a less stiff design with a 0.01 µm encoder that only works on a metrology granite.

• Environment sneaks up on you. Even in a temperature-controlled bay, localized heat from the motor can warp a long travel stage if it’s not mounted properly. We’ve started using a brief warm-up cycle in the morning to stabilize the stage before production, especially on the 400 mm stroke modules.

• Ask about customization early. Standard strokes didn’t fit our new gantry layout, and PSG was willing to build a custom-length stage without an absurd NRE charge. Having the same screw grade and rail preload in a non-standard frame saved us a major redesign.

Bottom Line

At this point, I see the linear stage as part of the measurement loop, not just a transport device. A well-built ball screw stage like PSG’s makes life easier: fewer unexplained drift tickets, simpler servo tuning, and predictable behavior across thousands of cycles. If your semiconductor equipment needs repeatable, low-vibration motion and you can tolerate a moderate top speed, this kind of stage is one of the safest bets I know.

After all, the best motion system is the one you forget is there – because it just works.