Reply 4 (by ShopOwner_Ontario): Practical reality check for anyone in this situation — get a quote for the ballscrew repair now, even if you're not ready to pull the trigger, because machine downtime planning is way easier when you've already done the vendor conversations and know your lead times. We deferred a Z-axis screw on our Okuma MU-500VA for about four months longer than we should have and when it finally got bad enough to park the machine we were scrambling to find a replacement screw, schedule the downtime around customer commitments, and find a technician all at the same time, and that chaos cost us more in overtime and expedite fees than the repair itself. In 2026 Okuma's OSP control also supports remote diagnostics through their Okuma App suite if you've set up the network connection, and their factory service team can actually monitor your axis servo data trends remotely and give you a data-driven timeline estimate for how long you realistically have before it becomes a hard problem.

Reply 3 (by QualityMgr_Aerospace): If you're doing any tolerance-critical work — aerospace, medical, automotive precision components — you really should put a go/no-go decision on this sooner rather than later because Z-axis backlash creep affects every operation involving plunge moves, drilling depth, peck cycles, and any tool path that reverses Z direction mid-feature. I've seen shops run with "compensated" ballscrew wear on Okuma vertical machining centers for months thinking parts were fine, then get a customer audit that pulls SPC data showing their Z-depth feature variation has been slowly trending out for weeks. Okuma's Thermal Friendly Concept on the MB-46VAE does a good job managing thermal growth but it cannot compensate for mechanical backlash, and the two error sources compound each other in a way that's hard to catch without real process monitoring data. Run a capability study on a Z-direction feature before you decide to defer the repair.

Reply 2 (by BallscrewRebuilder_WI): From a repair standpoint, the Okuma MB-46VAE Z-axis uses a 40mm diameter x 20mm pitch ballscrew that is readily available from both NSK and THK as a replacement, and if you get to a regrind or replace decision the current lead time from most North American distributors is running 6-8 weeks for a new screw versus 3-4 weeks for a professionally reground unit from a shop like Rockford Ball Screw or Steinmeyer. One thing worth noting — when Z-axis ballscrew wear is significant, always inspect the Z-axis linear guide rails and carriages at the same time because the wear patterns between ball screw and linear guide degradation are usually related and fixing one without the other often means doing the job twice within 18 months. On the Okuma MB series the Z-axis guideways are Fanuc-compatible roller types and they're not complicated to replace if you're doing a proper rebuild.

Reply 1 (by OkumaVet_Southeast): Running with growing Z-axis backlash on an Okuma MB-46VAE is something you can manage for a while depending on what you're making, but you need to be honest with yourself about what the machine is actually producing versus what you think it's producing. Okuma's OSP-P300 control does have backlash compensation built in and the service technician can update the compensation value to mask the mechanical wear temporarily, but compensation only corrects for a consistent, repeatable backlash value — if your screw has irregular wear along its length, compensation helps in one zone and hurts in another. The way to know your real situation is to run a ballbar test, not just a manual dial indicator check, because ballbar data will show you whether the backlash is consistent or variable and give you a proper diagnosis rather than a guess.

Reply 4 (by ProductionMgr_TN): To add some practical shop floor wisdom — any time you're dealing with mysterious data loss on a Fanuc control, pull up the alarm history and look specifically for 9001 or 9002 series alarms because those are SRAM and battery-related alarms that sometimes appear and clear so fast during power-up that the operator never sees them but they do leave traces in the alarm history log. Also check if your facility has any large machinery nearby that creates power quality issues because voltage spikes on the incoming power to a Fanuc control can cause SRAM write errors that look exactly like battery failure but happen even with a new battery. A UPS or a proper line conditioner ahead of the Fanuc control transformer is cheap insurance and I've seen it solve "unexplained" offset loss issues more than once in shops where the power quality was quietly terrible.

Reply 3 (by CNC_Maint_Foreman): Another angle that people miss — check whether your offsets are stored in G54-G59 or in the extended workpiece offset table G54.1 P1 through P300, because on some Fanuc configurations the extended table has separate battery-backed memory behavior and separate backup requirements. If whoever set up the machine originally programmed in G54.1 extended offsets and your backup/restore procedure only covers G54-G59, you'll lose the extended offsets every time and be completely confused because the standard offsets look fine. Also in 2026 it's worth knowing that Fanuc's FOCAS2 library for Ethernet-based data collection makes automated nightly offset backups very easy to implement, and several low-cost OEE tools now include this as a standard feature. If you're on a Kitamura with the optional Ethernet card, setting up automated offset backup is an afternoon project that pays for itself the first time you have a battery issue.

Reply 2 (by Kitamura_Owner_CA): I've had this exact thing happen on our Mycenter-3XG and while the battery is the first thing to check, there's another cause that caught us off guard — the parameter that protects offset write access. On the Fanuc 0i-MF there's a write-protect parameter and if it's set incorrectly or a parameter reset happened for any reason, the machine can accept offset entries during a session but not commit them to non-volatile memory, so they survive within a power session but vanish at shutdown. Check parameter 3290 bit 0 which is the offset write protect setting and also parameter 8900 series if your Kitamura has any builder-specific macro protection active. The Kitamura documentation supplements for the Mycenter line cover these parameters pretty well and Kitamura America's tech support in Wheeling is genuinely helpful if you call them.

Reply 1 (by FanucWizard_Midwest): Losing work offsets after power cycle on a Fanuc 0i-MF is a SRAM battery problem until proven otherwise — full stop. The 0i-MF uses a lithium battery to maintain the SRAM contents including work offsets, tool offsets, pitch error compensation, and macro variables, and when that battery gets low it doesn't always throw a clean low-battery alarm before data starts getting corrupted. The battery is a standard Fanuc A98L-0031-0007 and it costs about twelve dollars. If your machine is more than 3-4 years old and you've never replaced it, do it today regardless of what the battery voltage reads because these batteries often measure fine on a multimeter right up until they fail under load. Keep a full data backup of your SRAM on a USB or through Focas before you swap the battery and the data survives the swap fine.

Reply 4 (by 5AxisShop_Bavaria): For what it's worth, DMG Mori released a revised labyrinth seal design for the DMU 50 and DMU 65 monoBLOCK platforms in early 2025 that's supposed to address the chronic leak issue with a wider lip geometry and a different elastomer compound. It's sold as an upgrade kit rather than just a replacement part and the pricing from DMG is reasonable if you compare it to downtime cost. We installed it on two machines in our facility during scheduled maintenance and so far no leaks after 14 months of production use, which is better than any previous seal we've run on those platforms. If you're still on the original or first-replacement seal design, the upgrade kit is genuinely worth doing proactively at your next major PM rather than waiting for another leak event.

Reply 3 (by ManufacturingEngineer_IL): One thing worth adding — if you have a coolant leak at the rotary table and you've been running it while waiting for parts, pull the table apart and inspect the angular contact bearing races for rust staining before you just replace the seal and button it back up. Coolant that gets past a leaking seal sits in the bearing cavity and corrosion can start within weeks, especially with water-based coolants that have degraded past their recommended concentration. We had a DMU 50 where we did a seal-only repair and six months later had to do a full bearing replacement because we didn't catch the early corrosion, and a full trunnion bearing job is a much bigger number than a seal kit. A borescope inspection of the bearing cavity takes ten minutes and can save you a very expensive lesson.

Reply 2 (by Precision_Shop_NorCal): We replaced the trunnion seal on our DMU 50 in-house last year and I'll say it's a manageable job if you're comfortable with the machine and have DMG Mori's service documentation. The critical thing people mess up is the torque sequence on the trunnion bearing retaining ring — there's a very specific angular torque and preload spec that if you ignore it, you'll have table runout problems that you didn't have before. DMG Mori's local service team will tell you it must be done by them but honestly any competent maintenance tech with the documentation can do it. Budget about a full day for the job, order the Simrit seal kit and the O-ring set together because the secondary seals are cheap and foolish not to replace while you're in there. The seal kit part number has changed twice since 2020 so verify directly with the DMG spare parts portal.

Reply 1 (by DMGtech_Europa): The rotary table seal leak on the DMU 50 third generation is unfortunately something a lot of shops hit around the 5-7 year mark, especially in production environments running aggressive coolant concentrations. The culprit is almost always the Simrit labyrinth seal at the A-axis trunnion interface — DMG Mori switched to a slightly different seal compound around 2019 on the 3rd gen but the design is still vulnerable to certain biocide chemicals in modern synthetic coolants that swell and then crack the seal lips over time. Before you order parts, check what coolant you're running and whether it's compatible with the seal material — there's a compatibility chart in the machine maintenance section of the documentation portal and it's genuinely useful. If you're using a high-biocide semi-synthetic, switching to a full synthetic like Blaser Swisslube B-Cool 755 has solved the recurring seal issue for several shops I know.

Reply 4 (by VeteranCNC_Germany): For anyone running Mazak in 2026, make sure you're on the latest Smooth Technology firmware because there were documented ATC sequencing timing corrections released in late 2025 that significantly reduced false-positive arm faults on both the Variaxis and Integrex platforms. I manage a cell of three Variaxis i-700s in an automotive tier-1 shop and after the firmware update we went from maybe two ATC faults per week across the cell to essentially zero over four months. Also worth saying — the Mazak MTConnect data stream is much more useful than people realize for predicting ATC issues, because you can watch ATC motor current trends over time and see the degradation curve before it becomes a hard fault. If your shop has any kind of OEE software, map that parameter and set a threshold alert.

Reply 3 (by AdvancedMfg_Consultant): Something that doesn't get discussed enough on Mazak 5-axis platforms is hydraulic pressure creep in the tool clamp system. The i-500 uses a combination of Belleville springs and hydraulic unclamp pressure, and if your hydraulic unit has a worn relief valve, the unclamp pressure can bleed off slightly during rapid ATC cycles leaving the arm trying to pull a tool that isn't fully released. You'll see this as a torque overload alarm on the ATC motor or sometimes just an ambiguous position fault, and it gets worse as the hydraulic fluid heats up through the shift. Check your hydraulic pressure at the manifold during a live ATC cycle, not just at rest, and compare it to the spec in the Mazak maintenance manual section on the tool clamp unit. If it's drifting more than 5 bar low under load, rebuild or replace the relief valve assembly.

Reply 2 (by 5AxisForum_Member): Random M6 errors that only show up on repeat cycles and not on first tool call are almost always a thermal issue in my experience. The Variaxis i-500 has a cast iron ATC housing that expands slightly after 2-3 hours of operation, and if your tool pockets are at the tight end of spec, they'll bind intermittently when hot but work fine during cold morning startup. Mazak sets a pretty tight spec on pocket key orientation and if any of your retention knobs are worn or the pull stud specification is even slightly off — we're talking mixed ISO and MAS standards in the same magazine — you'll get random grab failures that look like arm faults in the alarm log but are really retention issues. Audit every tool holder retention knob in the magazine, they're cheap and it's an afternoon job.

Reply 1 (by JobShop_Kenji): The Variaxis i-500 ATC arm fault during M6 is a well-known pain point and almost always comes down to one of three things on the Mazatrol Matrix or Smooth controls: a dirty or worn cam follower roller on the ATC arm, a degraded proximity switch that's not confirming pocket seating, or a Z-axis macro timing drift that happens after years of thermal cycling. Start with the simple stuff — clean the ATC magazine thoroughly, check the pocket confirmation switches with a multimeter while manually cycling the magazine, and look for swarf accumulation in the arm pivot area because the Variaxis machines in particular collect chips in that zone due to the 5-axis tilt geometry. Mazak's own service bulletins from 2024 actually addressed a firmware patch for intermittent M6 faults on the i-500 series that's worth applying if you haven't already.

Reply 4 (by SpindleSpecialist_PNW): Genuine advice from someone who has rebuilt probably 200 Haas spindles over 15 years — the grinding at 3k hours almost always means the grease in the bearings has emulsified or dried out, and on the VF-2 the lubrication interval from the factory is honestly too conservative for shops running 3-shift operations. If you're running this machine hard, you should be doing a spindle health check at 2000 hours using vibration analysis, not waiting for noise. Also look at your tool holder balance ratings — if anyone on your floor is running unbalanced holders above 12,000 RPM you are destroying bearings faster than any other single factor. G2.5 balance at max RPM isn't optional on a high-speed spindle, it's just physics.

Reply 3 (by CNCFloor_Supervisor): Check your Renishaw tool setter and spindle orientation parameters in the parameter list before you assume it's purely mechanical. A drifting orientation encoder can create a harmonic loading condition that feels and sounds like bearing wear but is actually a control feedback loop issue. We chased a grinding noise on a 2019 VF-2 for two weeks before finding a loose Heidenhain encoder coupling behind the spindle motor. Also worth noting — if your machine is under Haas Warranty Plus or you have a local HFO nearby, get them in for a spindle vibration analysis with their diagnostic tool before you disassemble anything, because they can tell the difference between bearing frequencies and resonance issues using spectrum analysis.

Reply 2 (by MachineShopMike_TX): One thing nobody tells you upfront is that Haas spindle taper fretting wear is often happening at the same time as the bearing noise and people miss it completely. Pull your test bar and check runout before you do anything, because if you're seeing more than 0.0003" at 6 inches, you've got compound problems. I've seen shops replace bearings twice without fixing the actual root cause, which was a slightly damaged BT40 taper from years of improperly torqued tool holders. In 2026 the Haas HRT and UMC platforms have better sealing than the older VFs, but on legacy VF-2 machines like yours, budget for a full spindle exchange from a reman house like Northland Tool rather than a field repair if you're doing high-accuracy work.

Reply 1 (by ToolroomVet_Ohio): Had the exact same issue on our VF-2SS back in 2022, and honestly the grinding noise at around 3000 spindle hours is almost a rite of passage with that machine. What you're hearing is most likely the angular contact bearings starting to fail, specifically the upper pair. Haas uses 7008 CDUP4 bearings in the VF series spindles and they're not cheap but they're widely available from NSK or FAG if you don't want to pay Haas prices. Before you pull the spindle, check your coolant-through-spindle system because contaminated coolant that bypasses the labyrinth seal is usually what kills those bearings prematurely. If you're running flood coolant only and never CTS, you might still have contamination migrating up through worn seals, so don't skip that inspection step.