Not serious, just messing around :-) There is some underlying frustration in my comments though that it really sucks that nothing can be trusted anymore, you can't take ANYTHING to the bank that it is good without having a system or process to verify it. We bought a spindle test bar from Command recently and I specifically requested that it be calibrated, certified, and all said documents provided. Well I had a minor spindle bump so I put in the test bar to verify the spindle and it showed excessive runout. Just to verify, I put it in the spindle of my two week old makino and it said excessive runout. I sent the test bar back to Command and they inspected and said it was out of spec and that they would send me another. I WAS ROYALLY PISSED OFF and I told them to keep it and give me a full refund. Ordered a test bar from Parlec and it said both spindles were just fine... It really sucks when we get to the point where we need master gages to verify our gages, etc... This story also reminds me when I measured a calibrated pin gage with my calibrated micrometer and the pin gage measured 0.124" and it was a 0.125" gage pin... WTF??? I am 99.9% confident in the mic. Maybe I over torqued the mic. How many ft-lbs are mics supposed to be torqued to when taking measurements anyways? Should I need pliers? At the end of the day all of this stuff is produced by humans and we know how fallible humans are. Unfortunately no process will ever be perfect because it is implemented by people.

LOL! I'll just take my chances that the tap is good. At some point one needs to just push forward to avoid analysis paralysis and actually get some work done. The time wasted checking every end mill, every tap, every drill will FAR outweigh the time lost by actually getting one that was engraved wrong. Of course that will all change once we are AS9100 certified...

Right, but how does one know if the gage is engraved correctly?

That really sucks. Any way the plates can be fixed through thread milling? Inserts?

Cmon, lets be realistic... $100 per hour in a Haas shop can EASILY be turned up to $200+ per hour (without increasing the part value or cost) on the setup you are describing and you know I'm not joking or exaggerating.

I bought the parts and put them together myself.

I should also add that the operating system resides on its own disk and the storage resides on different disks so in total each box has four SSDs. Two for the OS in RAID1 and two for storage in RAID1. Nothing like redundancy.

ALL of my computers are set up in RAID1, even the server. If a drive goes down it takes a few minutes to replace it. Otherwise it takes a fully day by the time all is said and done...

I'm no expert in this area but I would anneal, then machine hard. At RC45 the stuff still cuts like butter.

I have to pay $$$ for it... I liken it to throwing $$$ in the fire pit and lighting a match.

The main reason I really like Vericut being the master is it acts somewhat like a firewall, only the NC file is passed through. The other day I posted a program from Mastercam for Solidworks and it showed as perfect in Mastrercam verify but Vericut clearly showed that the program had been shifted by .17" in X. I checked and rechecked everything and after a full regen in Mastercam it posted just fine. IMHO there have been too many reliability issues with Mastercam to make it the master. I would worry that Mastercam might push bad data (screwed up tool number, etc...) into Vericut, which it would then pass, only to cause a major issue at the machine. With only the NC file coming from Mastercam that will never happen. It might be a little more work but it makes the process bulletproof. Just my opinion and a big reason Vericut has been the master so far.

Leaving the lead screw engaged is kind of like leaving the chuck key in the chuck IMO. The way we run things here it takes two screw ups to cause a disaster and that is exactly what happened in this case. New guy shouldn't have left the lead screw engaged and the guy that stepped up to the machine should have done a quick overview to make sure it was safe to power up. I would have chewed them both...

James, have a closer look at your event log. You hit regen at 18:51:14 and the process completed at 18:55:13 for a time of 3:59, at least that is how I have been calculating the times. I believe the time you posted is for operation three alone, not the entire process. I have a few questions. What is vaporware? Also, you seem to be pretty computer saavy, do you know anything about overclocking or where I could look into it? We are running i7 5960 processors here (eight core @ 3.9GHz currently). I re-ran the benchmark and came in at 4:25 so a fair bit slower than your system. Before switching my graphics card setting to 'graphics only' I was at 5:00 so the graphics card appears to play a huge part in the performance of the benchmark results. I am running a Quadro 4200. When I ran the tests before (previously posted results) I was overclocked, but in a different configuration (ram was running faster). The 5960 processor can be overclocked a ton but one has to know what they are doing. Where do I find this stuff out? There has to be a deliberate process that yields good results. I have a fair amount invested in these computers and I want to get every ounce of performance out of them that I can.

If I was looking to get a 5-axis machine Mazak wouldn't even be in the running. I would look at Matsuura or Makino, though my feeling is that Matsuura has the edge in the 5-axis realm.

Vericut should be 100% accurate (in a perfect world) but getting it there involves a fair amount of work correcting issues to make the simulation match the machine. Unfortunately you found one of the issues the hard (and expensive) way but I wouldn't let this go until you can repeat the issue, fix it in Vericut, and confirm that it is no longer an issue. I am very confident that Vericut matches my machines but if I was aware of an issue where Vericut was off I would make addressing the issue my top priority before moving forward. Also, anytime I start implementing a new technology I watch everything with an eagle eye to make sure Vericut is doing what it should be doing. Once running for a few weeks like that I consider it good and move on.

My only guess would be that maybe the G5P0 or the G64 turned off the G43.4 in Vericut. Have you run this, or similar programs in the past? I ran into an issue some time ago where G53 canceled the tool length offset in the machine but Vericut didn't cancel it. The result was almost a full rapid crash into my tombstone but I caught it at the control. I fixed the issue in Vericut (simple control mod) so it will catch it going forward. When we implement G43.4 here we will do a thorough shakedown via testing at the machine to make sure Vericut and the machine control are on the same page.

From what I have heard I think I'd go crazy.

Funny how the bean counters, who are supposedly business and fiscal experts, can screw things up so badly. Say I have a shop rate of $100 per hour and I am running that tool at 150 ipm and getting INFINITE tool life. How much is there to gain if I push that tool to 500 ipm and only get a mere hour of tool life? Well, when that tool is in the machine running my effective shop rate has just climbed to $333 per hour (500/150*100). Not because the shop rate was adjusted but because that much more work is getting done in the same amount of time. The tool lasts for one hour in the cut so it is effectively making me $333 instead of $100 at the slower speeds and feeds. After the hour in the cut and a spent tool, at $40 per tool it just paid for itself and put $193 extra $$$ in my pocket, just by pushing it harder... That is how you make money with a machine shop. Do that for every tool, every job, and buy machines that can run them like that. Then your shop time becomes so valuable you start spending $$$$$ on quick change EVERYTHING because you don't want the machines sitting idle, EVER.

Somehow RFID gets the values directly into the Makino side of the control and I'm curious how they do it. I got quotes for RFID and its in the ballpark of $25k per machine... Too expensive at this point but valuable technology. I'm trying to figure out a cheaper way... Getting the offsets into the Fanuc side of the control is easy and there are turn key systems to do so that are reasonably priced. Getting it into the Makino side is a different matter though. What I envision is tool holder data stored permanently on the RFID chip and tool data read and written when tools are loaded. When the tool is read into the control the holder data would be stored as an eight digit number in macro variables (60 tool magazine would use macro variables 700-759 for example) allowing deriving holder length and holder ID via macro program (12505123 = holder #125 with a 5.123" nose length, or similar) and the tool data would go into the offsets. This would allow on-machine holder checking routines against misc reals posted from Mastercam (mr3= minimum cutter stick out) or Vericut could query the machine's current tool configuration via FOCAS or MTConnect for real time verification based on LOADED tool data. Trying to figure out how to make it work. I'm about 95% sure it could be done.

We have a Speroni with the Magis (base version) control software.

I have a few questions regarding the communication between tool presetters and machine tools. When the presetter is connected to the machine via Ethernet, how does it work when the tool offsets are sent directly to the machine? Are these values set immediately? Are they set to the Fanuc side or in my case the Makino side of the control? Are they posted to the machine as a program that needs to be executed at a later time? Currently I either print the offsets and manually enter them at the machine or I will post a program to a USB stick, load it into the machine and run the program to enter the offsets. I would like to know exactly what the process is when connected to the machine via Ethernet and sending the values directly to the control. I use the Makino side of the control exclusively for tool management in all of my machines.

When I posted my original Haas vs. Makino videos I got a lot of flack because I was comparing apples to oranges. Comparing a Makino PS95 to a Haas VM3 isn't apples to oranges but comparing an A51nx HMC to a Haas really is comparing apples to oranges so for this video I decided to leave Haas out of the picture. This is a short video of the machine that recently replaced my Haas VM3 "Mold Machine". I figured some folks would enjoy it. http://youtu.be/9QX09YOQdpg

Are there any that dehumidify as well?

I currently have Royal mist collectors on two of my three machines. The new horizontal mill doesn't yet have one but it isn't too much of an issue because we have primarily run steel in that machine and we run steel dry for the most part. The motor bearings on one of the units are making some noise and probably need to be replaced so I was planning on buying a new unit to replace the noisy unit, then rebuild the noisy unit and install it on the new horizontal machine.. Before I go out and buy a new unit I was curious if there is something better available. One The Filtermist works okay but the shop gets really humid when running hard. Any thoughts?

Got it, thanks for sharing. I'll try to have a look at it in the next few days, it has been a crazy week.