c/cnc-operators

Old school tool offsets vs. today's presetters, I miss the feel

Last week I watched a kid on a new Haas just trust the presetter numbers and run, and it hit me that back in 2017 I was touching off with a .001 shim and a prayer in that shop outside Columbus. Anyone else think we lost something when we stopped double checking with a manual cut every single time?

The coolant mix warning from a retired machinist in Dayton

Old guy at the Dayton tool show told me to stop trusting the auto-mixers and check with a refractometer every morning. I blew it off for 3 months, then our Haas started smelling like rotten eggs and the sump turned slimy. Turns out the mix was at 3% concentration instead of 8%, and it cost us $400 in new filters and a weekend of downtime. Has anyone else had a coolant ratio creep up on them without the machine throwing a warning?

Brand new spindle bearing seized up after 40 hours

Put a fresh spindle bearing in my mill last Tuesday. Ran fine for two days, then started screaming at 8k rpm. Pulled it apart last night and the cage was shredded, balls were blue. Grease was there, preload was set per the manual. Turns out the spindle housing had a burr I missed, threw the whole alignment off. Cost me $180 and a full shift of downtime. Anyone else deal with a new bearing failing that fast?

Warning: my 3 jaw chuck almost wrecked a $400 job in Cleveland

Last Tuesday a guy named Mike at a shop in Cleveland told me he never trusts 3 jaw chucks over 4 inches without indicating the part first, and I laughed it off at the time. Then my repeat run on a 4 inch aluminum flange walked 15 thou on the bore because I got lazy and skipped the indicator, so he was right and my foreman was NOT happy. Anyone else got a habit their coworker warned them about that finally bit them in the ass?}

Unpopular opinion: setting tool offsets with a paper feeler is fine for most jobs

Last week I watched a guy spend 15 minutes with a Haimer trying to get a perfect Z offset on a Haas, while I just used a piece of copy paper and was cutting in 2 minutes. 3 years ago I bought a $200 probe setup and it sat in the drawer for a year before I sold it. For tight tolerance work I get it, but for basic aluminum parts with +- .005 tolerance, paper works. Anyone else skip the fancy touch probes?

Tried a cheap Chinese end mill from Amazon and it actually cut clean

Picked up a 10 pack of 1/4 inch carbide end mills for $18 off Amazon last month. Thought they'd be garbage but ran them on 1018 steel with a 0.02 doc and they held up fine for 3 parts. Anyone else had luck with those budget packs or just me?

Picked a Kennametal boring bar over a Sandvik on a tight ID job last week

I had to bore a 0.750 hole 4 inches deep in 4140 prehard. The Sandvik bar was shorter and stiffer but the Kennametal had a coolant-through option. I went with the Kennametal because the chips were piling up bad. It worked fine but the surface finish was a little rougher than I wanted. Has anyone else had to choose between rigidity and chip control like that?

Tried roughing my finish pass with a 2-flute instead of a 4-flute on 6061 aluminum and got chatter marks I couldn't explain

Switched to a 2-flute to clear chips faster at 12,000 RPM but the surface finish came out worse than my usual 4-flute setup, so has anyone else seen better chip evacuation actually make things worse with a certain material or spindle speed?

My old calipers were off by .002 for six months in Phoenix

At the shop I work at in Phoenix, I kept noticing some of my parts were coming out tighter than usual. Checked the machine, checked my offsets, everything looked fine. Then last week I grabbed a known good part and measured it with my old digital calipers I've had for like eight years. It read .502 when it should have been .500. I felt like an absolute fool. I grabbed a new pair of Mitutoyo calipers from the tool room and boom, everything lined up again. So yeah, I ran almost six months with bad measurements on hundreds of parts. Has anyone else had a tool drift like that without you realizing it?

Remember when we set up machines with paper tapes?

Last Tuesday I had to dig out a roll of paper tape from the 80s to troubleshoot an old lathe, and it took me back to my first week in the trade back in 1993. That tape had a tear in it and I spent two hours reprogramming the part by hand. Anyone else still have to deal with legacy machines that run on tape?

The trick I learned for dialing in feeds on an old VMC

After fighting with chatter on a 15 year-old Haas for months, I finally tried slowing the spindle by 20% and bumping the feed up by 10 bucks an inch, which my old boss always swore by but I ignored. Has anyone else had luck with messing with the ratio instead of just changing depth of cut?

Just found a stat that changed my mind about tool presetters

I read on Practical Machinist that 73% of setups with manual tool measuring have a .002" or greater error on the first cut. That number made me reconsider spending $400 on a presetter for our Haas after swearing they were overkill for years. Anyone else have data that flipped your opinion on something?

Three years running the same G-code and just noticed a typo in line 47 that was adding 0.01 to my Z every cycle

Check your tool holders more than once

I run a Haas VF-2 in a small shop near Akron. Started noticing chatter on finish passes that wasn't there before. Checked my tool holders and saw .002 runout on one I'd been using since June. Never bothered to remeasure after the initial setup. That slop killed my surface finish on a $800 job last week. Who else checks their holders more than once?

I finally blew a tool holder clean off the spindle last Tuesday

I was running a 3/4 rougher in a CAT40 holder at 10,000 RPM and 200 IPM on a piece of 4140. I had tightened the pull stud but I guess I didn't torque it right. The whole holder flew off and smacked the chip auger housing so hard it left a dent. Has anyone else had a pull stud come loose like that and did you find a better way to check it before hitting cycle start?

Went to a local job shop last Thursday and it changed my mind about manual offsets

I always thought manual offsets were a waste of time for production runs, but the old timer there showed me how he tweaks his Z by .002 every 50 parts to account for tool wear. He was holding +/- .0005 across a 200 piece run without any probe cycles. Has anyone else tried this instead of relying on tool setters?

Old timer shed guy told me I was cutting my tenons wrong

I had this guy who's been running a CNC for 30 years watch me set up a tenon on a job last month and he just said 'your shoulders are all wrong, too much stress on the tool'. He showed me how to offset the cuts by a few thou so the tool enters smoother and I haven't had a chatter issue since. Has anyone else had their whole approach flipped by one comment like that?

Hot take: that time I spent 3 months chasing a chatter issue on a Haas VF-2 and it was just the workholding

Last June I was pulling my hair out over a 40 thou finish on 6061 until a guy at a trade show in Chicago asked if I'd checked my parallels were even clamped - turns out one jaw had a 5 thou lift from a burr I never saw.

Spent 4 hours chasing a .001 inch offset that was just a chip in the vise jaw

I was running a batch of 50 aluminum brackets yesterday and kept getting a .0015 offset on my X axis after every part. Checked the spindle, trammed the head, even swapped tool holders. Four hours later I pulled the vise apart and found a tiny aluminum chip stuck under the jaw. Cleaned it out and everything was dead nuts again. Anyone else ever waste half a shift on something that dumb?

Talked to a old timer at the shop who changed how I set my feeds

Ngl I been running CNC for about 8 years now and always pushed my feeds hard to get parts out faster. Last week this guy Jerry who retired from Boeing came by to pick up a part. He watched me run a job and said 'slow down, your tool deflection is killing your surface finish.' I pulled up the finish pass and checked with a micrometer and he was right on. He showed me his old setup sheets where he'd drop feed rates by 30 percent for the last pass. Tried it on a stainless job the next day and got the best finish I've ever seen off that old Haas. Anyone else have a old timer give you advice that actually changed your numbers?

Took me 6 hours to figure out a G-code offset that was off by 0.002 inch

Spent 4 hours chasing a 0.002 inch offset on a Haas VF-2

I was dialing in a new part for a job last Tuesday and the boss kept telling me the bore was off. After 4 hours of messing with the work offset, I finally figured out I left a chip under the clamp. Has anyone else wasted a whole shift on something that dumb?

My first aluminum part looked terrible, 6 months later it's like a different machine

Back when I started running CNC at this shop near Toledo, I cut a little aluminum bracket that came out looking like a chewed-up soda can. Surface finish was garbage, tool marks everywhere, and I had chatter so bad the whole part vibrated off the vise. Fast forward 6 months and I just ran the same part tonight, same material batch and everything. The difference is mostly from learning proper speeds and feeds plus using a better rougher toolpath. I also switched to a 3 flute variable helix endmill and started using a mist coolant setup. Anyone else look back at their early parts and cringe at how bad they were?

Can we talk about switching from manual offsets to tool probes

I spent 2 years doing all my offsets by hand with a piece of paper and a calculator, but after getting a Renishaw probe last month I can set up a job in under 2 minutes instead of 15, so has anyone else made the jump and regretted the cost or was it worth it for you?

PSA: Hit 10,000 hours of spindle runtime on my Haas VF-2 today and the backlash was way worse than I thought

I finally rolled over to 10,000 hours on my machine last shift and decided to check the backlash as a routine thing (figured it would be a couple thou, you know). Turns out I was running with 0.008 inches of backlash on the X axis and had no idea. My parts were wandering around by like 0.005 over the last couple weeks and I just blamed it on tool wear or temperature swings. The ballscrew nut is shot, gonna cost me around $1,200 for a new one plus a day of downtime to swap it. Has anyone else had a machine sneak up on them like this with a big tolerance issue? How often do you guys actually check your backlash?