Most ball screw requests start the same way: "I need a 1605." Four digits — one diameter (16 mm), one lead (5 mm) — and almost none of the information needed to know whether that pair suits the machine it is going into. The part number says what someone has already decided. It does not say why. This article treats diameter and lead as what they are: two numbers chosen together, against one load case and one motion profile. The figures are illustrative and demonstrate the method; they are not drawn from any single customer's project. Where this stops, a companion piece takes over: if your question is specifically "5 mm or 10 mm lead?", that trade-off has its own detailed walk-through in why 5 mm and 10 mm leads behave differently. Here the lead is one half of a pair, not a standalone choice.
This is the question that comes up most often on long-stroke axes: "I have a 1,500 mm X-axis on a vertical machining center, and I want to run 2,500 rpm to hit my rapid-traverse target. The screw is a 16 mm rolled unit. Will it hold?" The figures used below are illustrative, not drawn from any single customer's project.
When someone asks me about an SFU1605, one of the first questions is usually: “How much load can it handle?” It sounds like a simple question, but there isn't one useful number that answers it. SFU1605 means a 16 mm nominal screw diameter with a 5 mm lead. It is widely used on compact CNC machines, small linear stages, automation equipment, and Z-axis mechanisms. The problem is that the load rating in a catalog is not the same thing as the payload of the machine. A ball screw may have a dynamic load rating of several kilonewtons, but that does not mean you should put several hundred kilograms on the axis and call the design finished. For an actual machine, I would look at the load, acceleration, screw length, speed, mounting direction, guide system, and required service life before deciding whether SFU1605 is suitable.
“There's about 0.05 mm of backlash on the X-axis. Should I replace the ball screw?” We get questions like this quite often. The first thing I would not do is recommend a new ball screw. Backlash can come from the ball nut, but it can also come from the bearing support, coupling, nut housing, guide system, or even the machine structure. If the actual problem is somewhere else, changing a C7 screw to a C5 screw—or replacing the whole screw assembly—won't fix much. Preload is part of the story, but it isn't the whole story. A properly preloaded ball nut can have very little internal clearance and good axial rigidity. Increase the preload too far, though, and you start paying for it with friction, heat and motor torque. That's why preload should be treated as a design parameter, not as a setting you simply turn up until the backlash disappears.
A while back, a customer ordered a batch of SFU1605 ball screws, 1000mm long, with 10mm lead. The specifications looked right on paper. The lead was correct for his speed requirement. The diameter was adequate for the load. But when he installed the first screw and ran the axis at higher speed, the screw started whining. Not a normal running noise—a high-pitched vibration that got worse as the RPM increased. He checked the alignment, checked the coupling, checked the motor tuning. Nothing fixed it. He sent me a video. I watched the screw spin and could see the vibration. Then I asked him one question: “What support bearings are you using on the floating end?” He said, “The standard ones that came with the kit.” That was the problem. The floating end bearing was too small for the screw length and operating speed. It allowed the screw to whip. The screw itself was fine. The end machining and support arrangement weren’t matched to the application. This happens more often than you’d think. People spend a lot of time choosing the diameter and lead, then treat the end machining and support bearings as an afterthought. But on a real machine, those details determine whether the screw runs smoothly or vibrates itself apart. So if you’ve already picked your lead and diameter, this is the next thing I’d check.
When someone asks for a ball screw, the lead is usually one of the first specifications we need to confirm. 5mm or 10mm? At first, the choice seems easy. A 10mm lead moves twice as far per revolution, so it should be better for high-speed motion. A 5mm lead gives more mechanical advantage, so it should be better for heavy loads. That's broadly true, but it doesn't tell you which one belongs on your machine. The lead affects speed, thrust, motor speed, resolution, and sometimes the way the whole axis has to be designed. A good choice starts with the machine requirements, not with the ball screw catalog.
A practical guide to choosing the right ball screw accuracy for CNC machines, automation equipment, and precision positioning systems. When a customer asks me for a ball screw, one of the first questions is usually: “Do you have C5?” My next question is usually: “Why do you need C5?” Sometimes there is a good reason. A precision CNC machine, inspection system, or positioning stage may genuinely benefit from the tighter lead accuracy. Other times, the customer is building a fairly ordinary CNC router and has simply heard that C5 is “better.” That is where things can get unnecessarily expensive. C5 and C7 are different accuracy grades, but choosing between them isn't really about buying the highest specification available. The more useful approach is to start with the machine's actual positioning requirement and work backward from there.
A practical look at screw diameter, load, stiffness, speed, installation space, and the real factors that should drive your selection. When someone asks me whether they should use an SFU1605 or an SFU2005, the first thing I look at is not the price. I look at the machine. What is the axis moving? How long is the screw? Is it horizontal or vertical? How much acceleration is required? Is the machine cutting metal, engraving wood, or simply transferring a fixture? Those questions usually tell you more than the model number does. SFU1605 and SFU2005 are both 5 mm lead ball screws. The obvious difference is the diameter: 16 mm versus 20 mm. That four-millimeter difference doesn't sound like much, but it changes the mechanical behavior of the axis. The larger screw generally gives you a stiffer mechanical element and more capacity, while the smaller one is easier to package and keeps the system lighter. So the question isn't really “Which one is better?” It's “Which one is big enough for the job without making the machine unnecessarily heavy and expensive?”
A customer once sent me a photo of a linear guide block sitting on a workbench. The email said: “We need this exact size. Send us a compatible replacement.” There was no ruler in the photo. No original part number. No machine information. I asked what machine it came from. He said it was a CNC router. I asked if he was replacing the rail too. He said no, just the block. At that point I knew the project could go two ways. Either it would be a quick fix, or it would turn into one of those jobs where the machine runs worse after the “repair.” Unfortunately, it was the second one. He ordered a compatible carriage. It fit the rail. But the axis felt tight after installation. The servo load went up. The operator complained that the machine didn’t sound the same. The new block wasn’t defective. The problem was a combination of things: the old rail was already worn, the replacement preload was different, and the block height was slightly off. None of those things showed up in a product photo. That experience, and several others like it, shaped how I now look at THK / HIWIN compatible linear guide sliders. It’s not about whether a compatible part can work. It’s about whether you’re comparing the right things before you buy.
A Practical Troubleshooting Guide for Wear, Preload, Alignment, Lubrication, and Thermal Drift When a CNC machine starts missing its tolerance, most engineers don't look at the linear guide first. The usual suspects are the ball screw, servo tuning, encoder, or coupling. That's reasonable. But after working through enough motion-system problems, I've found that the guide often gets blamed too late—or replaced too early. A linear guide doesn't have to completely fail before it affects machining accuracy. A small increase in friction, a change in preload, contamination in the raceway, or a rail that was installed slightly out of alignment can change how the whole axis behaves. The difficult part is figuring out which mechanical problem is actually causing the accuracy change. Here's how I would approach it.
What Engineers Should Know About Rigidity, Friction, Heat, and Long-Term Guide Performance When a CNC axis starts behaving differently after months of operation, the first things people usually check are the ball screw, servo parameters, and encoder. The linear guide preload is often much further down the list. That is understandable. Preload isn't something you can see during normal operation. There is no value on the HMI showing that the guide has become too tight or too loose. But you can see its effects. A carriage that develops higher running resistance, a servo motor that starts drawing more current, a machine that becomes less stable at low speed, or a guide that feels noticeably different from one end of the rail to the other can all point toward a mechanical problem involving preload, alignment, or both. This is why I don't treat preload as just another catalog specification. For a CNC machine, it is part of the mechanical balance between rigidity and running resistance.
An Engineer’s Practical Guide to Load, Rigidity, Speed, Installation, and Service Life When a CNC axis starts making noise, drawing more servo current, or losing repeatability, the linear guide is not always the first component engineers suspect. Sometimes the problem is the ball screw. Sometimes it is the mounting surface. Sometimes the machine frame is simply not stiff enough. But quite often, the guide system was underspecified from the beginning. This usually happens because the selection was made from one number: rail size. A 15 mm guide, for example, does not automatically tell you whether the system can handle a particular CNC axis. The actual load, moment, acceleration, rail spacing, preload, mounting accuracy, and operating environment all matter. When I select a linear guide for an automation machine, I normally work backward from the load conditions rather than starting with the catalog.
When designing an automation machine, choosing a linear guide is often underestimated. Many engineers initially focus on: Motor size Servo performance Controller accuracy But after the machine starts running, problems such as vibration, uneven movement, and positioning drift often come back to the mechanical guide system. I have seen many cases where the guide was selected only based on rail size, without considering: Load direction Moment force Mounting space Required rigidity HGH15, HGR15, and MGN15 are all 15 mm class linear guide solutions, but they are designed for very different working conditions. The right choice depends on how the machine actually moves.
When a linear stage is first installed, most machines can easily pass the initial accuracy inspection. The difficult part comes later. After months of operation, engineers may start seeing problems that were not visible during commissioning: Positioning deviation after warm-up Increased servo load Longer settling time Different machining results between cycles In many cases, the linear stage itself is not completely damaged. Instead, small changes in preload, lubrication, alignment, or thermal behavior gradually affect motion stability. From CNC machines to semiconductor equipment, maintaining long-term positioning performance requires more than selecting a high-accuracy stage. It requires understanding how the mechanical system changes during real operation.
When engineers design semiconductor equipment, the focus is usually placed on process accuracy, throughput, and reliability. However, the motion system supporting the process is equally important. A wafer inspection camera, bonding head, or laser processing unit can only perform accurately when the linear stage provides stable and repeatable movement. In semiconductor equipment, motion problems are usually not caused by a single positioning error. They often come from: Mechanical vibration Thermal drift Insufficient rigidity Long-term repeatability changes This is why precision linear stages are widely used in semiconductor automation systems where stable motion performance is required.
+8615622924499
+8615622924499
