Crooked machine, everything is made of rubber, and the dude still splits a thou in 10" deep bores. You're one hell of a machinist, Pete. Not even joking.
Thanks, always wondered how accurate these big machines were vs how much you had to adjust your processes to make up for the inaccuracies. Very educational. One other note... couldn't help but think I was going to hear someone yell "cycle start" while you were inside that monster... Thanks for posting - always learn something sir.
Peter, you lucky dog! I would have bet money that thing was going to move on you when you tightened up all the clamp bolts after adjusting! It appears that it didn't even change your 0/90 positions a tenth or two! Thanks for showing this to all of us -- really fun to watch, and highly appreciated!
Masterpiece. I can watch you indicate things for 1/2 an hour. Not being the least bit facetious. Learned a tremendous amount about machine construction.
An interesting video thanks Peter. There is such a lot of effort gone into designing and building that machine.And the machines that built that machine. No wonder they cost so much and seem to hold a lot of their price especially the larger end like yours.
👍 Thanks again Peter for your time. Just a great insite into your skills again. Not just programing tool making and maintenance. Great. From a brit who only has 50 + year old machines to work with. 👍🔧🔩
Thank You again for showing us such a careful, thought out, and functional perspective from behind the weighs. This video was very informative. I purchased a used box of crayons, and now I see i need to look in the bottom of the box. My rainbow purple, and burnt sienna crayons may be at the end of their serviceable usage (huge smile).
What made you want to check it? Did you notice a finish cut not quite right.... a sound that didn't quite sound right.... or just a period check. I'm just curious what the temperature was on the test bar with lights and the surrounding temperature. Amazing what a teeni tiny difference makes on a bar that long. Now you think I'm mad I guess 😲😉
You mentioned that they should have made the tab for the adjustment stronger. I can tell you that you're right. The tab on our Integrex ended up snapping off during adjustment and now we have to do a convoluted setup with some clamps to adjust our B-axis. Hopefully they have a better or stronger setup on newer machines.
@@EdgePrecision Whenever we've needed to do an alignment, we've always needed the machine running again asap. One day if we've got the covers off and have some time, maybe I'll do the same.
The newer integrex (at least the I series) doesn't have the hard clamp every 45 degree. Instead it just hold the b axis with a "soft" brake. Not as strong, but it's really easy to adjust since it's done in the parameters.
@@Rasmus661 In my little experience, that is a real problem when doing heavy machining. It seems that now days a lot of machines can't handle that anymore. Of course there are modern adaptive toolpaths, but they don't work for everything. We do a lot of heavy milling at work in steel and also GS42CrMo4 castings and this stuff kills spindles and especially milling heads. And it seems to get harder to get machines with milling heads that can handle this kind of load, there are designed for different strategies now. I heard that we even did tests at a manufacture before buying a machine ten years ago and destroyed the milling head within two parts.
@@Rasmus661 I was wondering why one couldn't just dial in the B axis reference position with machine parameters, that explains it: in addition to the servo drive, the B axis has an indexable joint (curvic coupling?) which can be clamped hydraulically. Thanks.
The dreaded test bar. Shows about everything from spindle runout, bad spindle taper to error in machine geometry. I would check machine squareness with granite square or so, before adjusting milling head parallel to X (up down) axis, btw. Otherwise you might have no error on dial indicator, but a head that is parallel to an out of square axis. Same goes for Z (left right) axis. Good thing that capto taper is still good. Must be a nightmare to rework those, compared to ISO40&50.
Thats a hell of a machine man! Huge! I have changed out a ball screw once on an old okuma lathe. Lots of fun. Ours was only 1.75 inch. I would had never thought they built ballscrews with through coolant. That's crazy.
Hey Pete, I was wondering if there maybe a setting you can adjust so you can have zero in both directions. They would be some kind of encoder readings that tell the computer when it reaches position.
This is where lesser cost machines save money. But in the long run it costs the owner more. We had continual problems with the Quantum mills and their way covers that this would have corrected.
I don’t usually watch cnc machine work, but i watch yours,the best cnc machine work and channel in UA-cam .thank you and I really appreciate the effort you’re putting to share with us your works. Something is in my mind for a long time, how much did you paid for this machine?thank you once again
It's ideal to check the machine for twist like a piece of licorice before aligning the head. You have a nice machine. I'm a ~ retired Okuma tech. This is my favorite channel.
Thanks Robert! Yes you are correct. And because you know how involved that would be for a machine like this, I wont go into a explanation. But this machine is turning very straight and true. These Mazaks have a weakness in this area. And so it has to be watched and kept track off. Also this has to me running into something a little hard (The reason I checked it in the first place). I have never been into the B axis of this machine but I suspect something isn't exactly correct in there. But you know the amount of work it would take to do that. When I bought this machine I noticed the new looking paint in the B axis milling spindle compared to the rest of the machine. I asked the owners if it had been replaced. They said it had for some electrical problem with the spindle motor. But I suspect that someone maybe didn't quite do something the best way in there. To take that spindle motor off and disassemble the B axis is more than I want to do. I can work around these small insurances.
@@EdgePrecision If it was another time (Non Carona) and I wasn't here on the east coast in NJ, I'd be glad to help you out. I really enjoy this channel.
Wow, how much does a machine like that cost new and how often would you calibrate it? My guess is that it would be easier to get the machine company experts out once a year but fear that would be very expensive.
Sweet video and your skill set is mind blowing!! Not a lot of people like to share their knowledge, but you are quite the opposite! Would love to work under a machinist/programmer/maintenance tech like you!!
I have one question Instead of loosening the bolts and adjusting Is it not possible to adjust with CNC parameters? With Mitsubishi CNCs of this era, I feel that I was able to adjust the runout of all axes.
Not for this because what I'm doing here is adjusting the curvic coupling location. It has moved because I bumped into something a little hard. So I had reason to check it. This machine has both a full B axis and a coupling when it clamps the B axis at 45 degree intervals.
That’s very cool getting an in depth look inside a machine that large. Closest we have, in size only, is a Devlieg 4K-96. Been working on that lately. Nowhere near as sophisticated. Way cover accordion is the word you were looking for I believe. Although I know you know that already. Good video 👍
Edge Precision I don’t even need a camera around me to forget stuff I’ve known most my life. I first thought you were checking the front of the bar in the vertical position. I was trying to figure out how you were going to get that straight. Then I woke up. 1600 IPM damn that’s nice.
Nitrogen accumulators lose gas over time, so the lifting force changes, there is probably a range of gas pressure that Mazak specifies. Good thing is that most accumulators can be recharged, bad thing is that if you do it wrong they can be dangerous.
On OKUMA with OSP 300 control You can comp it out, but that's a cheap way of doing it. Get it mechanically as close as possible with adjustment bolts. Those are good machines. They do require competent technical assistance. Got 5 of them in my shop.
How often do you do the inspection? We do a ball bar, test bar, and RTCP check on a monthly basis on our machines but we have to to keep our part population tight. We do an annual laser calibration as well.
If you are running production style work. I could see the need. I only check things when I feel the need or something doesn’t seem to be cutting right. Because of the type of work I do and also no one else runs this machine but me. And I know the machine.
Hey Peter, Im Elect eng, worked on robots. There should be small offset adjustments that the service technician can do for each axis deep in the control. its like when you zero the machine to a job, but would be found deeper in maintenance section of the console. It would save you time pulling it to bits. fast, click click done. See if you can get the techo manual or ask a Mazak expert. The only mechanical adjustment should be in the axis of no movement. i.e. Same as what you were doing with the dial indicator but the front edge of the bar instead of the side of the bar. Side of bar adjustments (rotation of head), digital adjustments. Also, in an Okuma, the spindle will lock in 12 different positions when a lathe tool is in the spindle. Its a gear lock so the spindle wont turn. Your spindle was free to spin. It only makes sense that the Mazak has this lock as well, or else the servo will have to fight the tool to keep it still while the spindle is used for turning. If I were you, I would lock spindle in the down and also lock in the horizontal while doing those adjustments. Lastly, Thank you for showing us around inside the machine. I love your Mazak, nearly bought a single spindle version from auction and always wanted to see inside one. If you ever have the ways exposed on the 2nd spindle, pls film it for us. I love your stuff. Thank you from Australia.
I don't think I made this clear in the video. In this machine there are two things controlling the B axis. There is a cam roller rotary worm drive and also a curvic coupling. What I'm doing in this video is adjusting the mechanical curvic couplings zero location. This can't be done with parameters because it's a mechanical coupling that clamps the B axis every 5 degrees. This coupling has teeth like a gear on it's face that couple together for a very positive and strong clamping. This is more designed for turning tools and milling at the 0-90-180 degree locations. There is also a full B axis that can resolve to .0001 degree location. This would be adjusted as you say. I didn't show this in the video but the zero of the B axis has to be reset after adjusting this mechanical couplings zero location. So of course they are both the same. There is also pitch error compensation on the B axis Cam Roller drive but this has to be done by a Mazak technician.
No matter how good of a machinist you are some day you will crash a machine, especially cnc s. The fact that your machine crashed with the tailstock reminds me of an 1995 Hitachi Seiki lathe I worked with. It had a sub spindle and it was impossible to crash it. The machine always knew where the turret was in relation to the sub spindle-tailstock so it alarmed when you were to close, in jog or auto. Anyway you probably saved yourself a lot of money doing all this work to calibrate the machine. I imagine in the US this kind of labour is very expensive. Maybe you can give your opinion on linear guide ways versus box ways in heavy duty machining. Cheers!
Thanks Peter. Does the test bar deflect under it's own weight in the horizontal enough to show up on indicator? I just wonder if that accounts for some of the error there.
FWIW was wondering the same thing this morning. I ran some numbers for a bar 2.5" in dia and 14" long from the supported end the deflection is about 0.0001". For a bar 1.5" in dia it's about 0.0003".
I had the same issue with the squareness on my Integrex. The XZ and YX were not tracking straight so I learned how to build a comp table to adjust electronically. Now everything is .0001 square and I didn’t have to adjust manually. The Matrix controls allow this electronic straightness as well.
First Peter nice video, again thanx. From what i saw the B axis wasn't rotating a perfect 90 degrees from the directions of error. I have been programming and setting up Campbell rotary grinders where I'm at with a B axis. I program a B rotation with a backlash move to always come to my final angle in the same direction, seems to improve the rotate accuracy. Let me know if i was seeing that right about the B. Good job, I always look forward to the next video.
Cool video Peter! Did you measure the deviation in the Y-axis direction also? Not that it's easy to adjust but just to know. Used to work on an old (2005) mill/turn also and that machine had a error in the y-axis direction related to the testbar.
Yea I checked it in all directions. Along the total length of the test bar there was less then .001” deviation. That was the best I could get it. I also checked it a different B angles.
Great show, amazed at you and that mazak . I think it bieng an old machine.and seeing the maintnence to keep it in tolerance . Very fascinating . what kind of old machine would I buy, mills Bridgeport 9x49 K7T horizontal or the brown and sharp with the omniversanal table lathes hardinge dsm HVLV love the hardinge stuff. good to see ya agan
With the liquid cooled leadscrews, I'm guessing that this uses rotary encoders on the servos, rather than liner encoders? I would think that the encoder position error could be fixed in software, but not something they did back in 2006. Any idea what the expected lifespan is for a machine like that, before a rebuild? I know the iron will last, but all those hoses and cables are bound to fail with age someday.
Some machines of this spec have linear encoders for direct position measurement, looks like this one relies on the servo. All of those hoses are rated to last longer than your lifetime. Things that are going to fail are moving parts(rails, carriages, bearings) or the computer.
I am sure you did but I neveraw you check to be sure that the bar was exactly parallel to the spindle. You would wiggle the indicator to find center of the bar and then traverse, never saw you wiggle it at the other end to be sure it was still on center and that the bar wasn't pointing either towards the door or the wall. This could easily contribute to additional errors if you didn't check that as well. I would not be surprised for the head to be worn nodded down.
You are correct if the bar deviates from centerline in relation to the indicators stylist. But it has to be by a fair amount to cause a significant error. The bar is 40 mm in diameter or .787402" radius. Now lets say I was .001" off center from one end to the other when running the indicator down the bar. (You saw in the video I indicated the high point of the radius). Now at .001" off center the reading would vary to .787401 or only change .000001. From .787402 (20 mm Radius) to .787401 a .001" over on this radius. so yes there is a deviation but not much. the tip or nod of the spindle isn't this much. I checked that. But even if it was as you can see it not enough to worry about for this measurement.
Is there also an adjustment if the head is out In the Nod? If the Nod was out but the tilt was perfect, wouldn't you still see an error checking the tilt on a round bar as the indicator will be falling off the apex of the radius?
Thanks Bill. I’m waiting on something with inspection. At these times I check things like this and do Maintence issues for the machines. I don’t want to have to in the middle of a setup. Don’t ask how I know that. In reality this seems like a lot of error. But even this can be compensated almost away with fixture offsets. Remember that test bar is 14” long. Most machine work isn’t that long so the error isn’t that much in shorter lengths.
It is 300mm. long, not 500mm. Target is 0 when you align it, but usually I get it down to 0.01mm/300mm (10 micron/300mm). There are other factors in the picture too when you align the B-axis.
What about the tilt of the head if you measure on the bar frontside or backside when its vertical (straight down). Is it possible to adjust that if you got any errors?
Pete . You can go Bonkers you can get it right on specks . Then check it MUTCH later and get a. COMPLETE different reading the machine Temperature the floor Temp. All comes to play 😨😎
I'm 25, just did a spindle alignment and turret alignment on doosan puma today. Have you ever bumped in X- while turning and had to make the spindle parallel to Z again?
This runout I showed has to do with how the machine deals with lathe turning tools and clamping the spindle. It really wont affect milling tools. The turning tools can quite easily be compensated with wear offsets for this. I didn't show in the video about checking the front and back but I did check that. There was very little error that way.
Do you also do frequent kinematics measurements and adjustments? Would really love to see a video about that. I'm doing some kinematics checks and adjustments in fanuc, heidenhain and siemens controls lately. I'm interested how that works on a mazatrol controller.
I personally have never done anything but mechanical type of alignments on this machine. On my horizontal mill I have run a Ball Bar and adjusted the servo and backlash parameters. To improve it's circle roundness. Also a Ball Bar report will tell you about the squareness of the axes's being measured on the machine. So it will give you some idea of the machine's level, more than a bubble level can. I don't know if this is what you are referring to by Kinematics. I always thought of kinematics as something built into a machine that really could not be changed. With the exception of some servo parameters and the leveling of the machine for squareness. To it's original manufactured condition.
@@EdgePrecision thank you for the reaction. For the mill turn you can adjust the pivot point for the b-axis in x and z direction. If you put an ball in the spindle and verify that your machine X and Y are 0 on the main spindle with the b-axis on -90 degrees. (Think its zero on yours). Than have 2 indicators on the X and Z direction and move the b-axis around with active tcp. If you get any error you can adjust it by changing the kinematics parameters. I had problems using the same tool in different angles on my SMX2600S. Even worse on the subspindle side since the correction for main en sub is seperate.
Ok I see what you are referring to. Yes the Mazak has parameter settings for this. But my machine never needed this changed. I believe on the Mazatrol they are BA61 for X and BA62 for Y. My Cam software (Esprit TNG) also calls up the BA62 setting in the program so when rotating the B axis the machine doesn't over travel the machine.
If memory serves, yes there's a few parameters on the newer integrex that you can use. On the older ones like this, you must do it mechanically like this. They still reccomend you do it this way though.
Dumb question: Why the need to tram B-axis instead of offsetting the B-axis rotation abit electronically? Isnt it the same angle? I think I am missing something though
If this machine was just a 5 axis milling machine what you say would be correct. On a mill turn machine you can also do turning. So the B axis has two different methods of alignment so to speak. But you can actually use the B clamp feature when milling a 0,45,90,135,180 deg. The B axis has a coupling (Curvic coupling) and a servo drive that can position to a .0001 degree restitution. So if the curvic coupling is out of alignment it can only be aligned mechanically (It is a psychical coupling hydraulically clamped). This is necessary for running turning tools because you don't want the B servo to have to hold against say a long boring bars force. Also the coupling positions very accurately with no give that a servo system could have. I didn't show this in the video but once the coupling is aligned you do have to make sure the B servo drive coincides with the coupling. Or to put it another way they both have the same zero point. Also the mill spindle has a similar coupling to clamp at different angles (Every 15 degrees) for turning tools. Hope that makes sense and answers your question.
I love how Mazak's commercial videos are all clean CGI and techno music but we all know the reality of it is ear-splitting noise and a metric ton of grease and oil :-D
I'm not sure I follow what you are asking. When you say center rotation what are you referring to? The rotation of a work plane in a program? the tail center rotation? B axis rotation center?
Hey Peter - I've been watching you videos for a while now and have been totally enamored by your work. I really appreciate all the work you put into the videos and the knowledge you are giving to all of us. I am currently in the market for another machine to compliment my VF2SS. I am looking at adding a 4th/5th axis attachment to my mill or getting a live tool lathe. I have found a few early 2000 Integrex's. I was wondering what unknowns or complications do you see with getting an older Integrex up and running? How difficult/expensive is the initial tooling/CAM maintenance, etc? I don't know what I don't know as far as big issues with these machines that will cost me much more than the initial machine cost. Also how difficult is it to program? Sorry for the loaded question. Thank you so much for all your wisdom and knowledge. Cheers,
Hey Peter, I believe that Sandvik test bar would have its calibration data in microns. Wouldn’t you be much better off with a Millimess from Mahr or a Mitutoyo Hicator. That Intrepid is only reading 12-13 microns and with hysteresis its probably around 15 microns at least, so you’re not getting a good reading. Also, as you said, you can move it easily with your hand and working against gravity might not be helping when doing sideways calibration as the end of the calibration test adapter will be sagging a micron or two because of gravity. Wouldn’t it be best to calibrate the B axis when the spindle is facing down instead of left or right?
If the test bar is sagging from gravity than so are the tools in the horizontal position. I don't think there is enough error there to worry about it. I have inch tools so that what I use. I do have a higher resolution Interapid indicator reads in .0001". Inch or metric the indicators do the same thing. No one is any better than the other.
Some backlash in the B axis especially noticeable when the B axis clamps and unclamps and maybe the B axis bearing is not 100% if someone has crashed it in the past? Looks like backlash comp might need re setting in the control.
you may be able to calibrate it out in the servo driver. but that may cause more headache than its worth, assuming they have free tech support like i would give them a call.
Tramming.. is a complication issues....Also is the the matter of parallelism of two rails (or are they truly on a plane?any wear, clearance?) of the axis it is sliding on! Are they all of three pieces per axis!
Yes but if the coupling in the B axis of this machine gets knocked out. This is the only way to realign it. You think Mazak put those jack screws on there for a reason? I did run into something a little hard that's why I was doing this. I should have explained in the video.
Hmmm.... interesting. I would have thought that it would be possible to adjust the amount of rotation that the head makes between stops. I would think that the 45 degree tilt ( or 90 degree tilt - not sure) between stops would be handled by some sort of electrical device counting some sort of pulse. You'd need that to decide what angle the head was at anyway. I would think that it would do the same between the stop positions. A clever designer might even have different pulse counts for each specific segment of the arc the head rotates through. To tram it, you'd then set the bar straight in the horizontal location and then adjust the number of counts required for it to get the the vertical position. But I'm just a "back seat engineer". I'm sure the made the machine that way for a reason. Wow! if there is some sort of mechanical device that sets the angular rotation amount.... phew... I can't begin to imagine the accuracy required of such a device. Yowzers! Anyhow, thanks for sharing this. It is absolutely fascinating.
The B axis can position to .0001 degree position accuracy. But it also has a coupling that clamps every 45 degrees. Mostly for turning operations. But except for full 5 axis milling almost all machining is done at the 0 and 90 degree angles.
@@EdgePrecision So something similar to a hirth joint that gets clamped together to set the head at a given rotation? Just out of curiosity, can you rotate the head 180 degrees from the front facing horizontal position? And, if you can, does the alignment stay true? That could lead down a serious rabbit hole if it's not consistent with the other measurements.
it's was littlebit surprise for that the mazak control sofware runs on win2000. came littlebit nostalgic feeling because last time i have usy any windows is about 10-12 years ago, and question can you put some linux operstion system to mazak tool use it that way. i just means is ti posible at all, and then second thing came to my mind bcause in one vodeo you show you have ethernet cable going on mazak to your laptop and you have internet connection, is there any danger that virus or some else unwanted progmam find it's way to mazal machine, but i very much to like watch that you do parts that looks like they are imposible to do but seemingly they are not. and one question yet, how long it take you to came that good with mazak machine because i think if i start to learn how to use it it's take several years to learn to use it somewhat deasently so i can make some rather easy peaces on it. 👍👍 🙂 and hello from finland 🇫🇮 ( sorry if i ask stupid question but i have to because i have think these things almost always when i watch your videos. o and what was the name of mazak's controll software. was it mazatrol or something like that then i can look more info on internet when i know it.
I don't know about the linux operating system? The PC in the Mazak is a industrial one configures specially for the Mazak. All Mazaks actually run on a Mitsubishi Meldas control. All the electronics are Mitsubishi in the machine. The machine itself is not connected to the internet. Its on a separate network just to my laptop. The only files that get transferred to it are G code files. I don't know but I doubt a virus could be transferred in a G code file. Isn't a virus a program? The code file wouldn't have that kind of information. But what do I know? I just run machines. But I suppose its possible over the network cable from the laptop. We do have a server here that suppose to block that and my laptop has virus protection software. But anything is possible. The software that a Mazak control runs is called Mazatrol. But it is a proprietary software that Mazak uses set up for the machine it is running on. I don't think it can be purchased and set up on another machine. Like say Mach3.
On a say a vertical or horizontal mill the leveling of the machine makes a big difference. Also for this type of machine it comes into play to get it to turn straight. Witch this machine is doing fine. But this adjustment cant be controlled by just leveling. The squareness of the axis's to each other can but the indexing accuracy of this is probably tied in to something being damaged in the coupling in the B axis. I don't want to tear all that apart for this (To look at it another way. It would be very expensive). This small inaccuracy can be dealt with other ways. It just isn't worth it. What I'm trying to show here is for 99.999% of all machine work this is good enough. I bet if you really tested machines that most machinists us every day. They would all have these small inaccuracies. They just don't know it because they compensate for it in other ways. What do they say. You got to hunt with the dog you have.
@@EdgePrecision Thanks for the reply. I have Vertical CNCs and they sit on the shop floor that is not built for heavy equipment, but it is what it is so we have to relevel every now and then to keep things square. Love your channel BTW.
i wonder if you had some mechanical backlash on the B axis that could cause the symptoms You experience. i can not tell from the video but if You position without the backlash taken into consideration via using an unidirectional approach to establish the position of measurements than this can be what you get. also if it has a worm gear drive for the B axis than that should have some way of backlash adjustment.Tsudakoma on their rotary tables use evolving profiled wormscrews that are fine adjustable by changing the longitudinal position of the worm screw with its support bearing assembly (bearing + housing) altogether.
There are two different things here. By the way B axis is driven by a roller cam drive on this machine. No worm drive. But there is also a mechanical coupling that engages at larger angles. This clamps solid like a turret on a turret lathe. So in this coupling can get knocked out in relation to the B axis cam drive (witch can be zeroed anywhere like a worm drive). So if the two aren’t aligned to each other. When you couple/clamp the B axis it shifts a little. So first you align the coupling. Than re zero the cam drive.
@@EdgePrecision Yup Roller cam makes much more sense for the longevity and frictionless motion. obviously You have more insight to how this B axis is actually constructed. i highly appreciate your answer. made me curious enough tto dig around the internet how Mazak designed this B axis. however please don't ever hesitate to make videos on maintenance tasks, mastering maintenance saves a ton of money for the smart ones. i love to watch your channel mainly because You do challenging parts and take on difficult to tackle tasks. we in our shop are kinda leaning to the same direction with our endevours tho it is challenging with 30+ staff . doing challenging parts in smaller quantities really makes much more sense to get the most out of the machines in terms of profit while trying to keep maintenance costs and amortization low all be it one needs enormous knowledge to succeed on that road. for me it seems You really enjoy working on your own, we are pretty similar in that regard tbh. also i imagine the old days when You were running a whole shop did their number on You considering the stress factor involved just purely from the financial side of things. anyways ♥ from Hungary. oh and please dont hesitate to show us what is behind that cover on the B axis if You ever get to remove it for any reasons, i would love to have a sneak peak into the hearth of the B axis motion 😁
I am from the old school, before the days of CNC! In the early 60s with similar situations on a horizontal mill or when attaching a grinding attachment to a vertical lathe, etc. 1) We would indicate both spindles (zero, flat, concentric!), just like the spindle to the table on smaller Bridgeport! 2) It appears in the internal inside tour, there may be an "overbalance in equipment mass weight" towards the machining / tool / spindles area. I.E., overhang / overbalance on the ways from the backside towards the front. *** Needs an articulated counterweight similar to that of a pipe layer! * Might be overkill * C) Would it be feasible to do a complete range of motion test (X-Y-Z ), in a sectional comparative (ROM)? Old School would be looking for wear on load bearing carriage ways, etc (all 4 corners). Laser / target sight, main spindle to articulated spindle (gun sight adapted?). Consider: (14 years!) X (how many cycles?) X (tools changed?) X (tool bumps?) + (grit under the carriage?) = Too many questions! (way lube & wipes ain't perfect). D) That irregularities in the spindle lock brings a lot of other questions. Modern Girls need a little more, TLC!!!!
I'm used to 5 axis milling with FANAC controllers. One is a milling machine and the other is a large tilt/ rotate table attachment . In both cases I just read the angle deviation away from the homed position angle after tramming in, then deep dive into the controller and add or subtract that value to the existing one that sets the homing angle for that axis. I think from memory there is a +/- 5 degrees adjustability doing it this way, so its a bad day if you have to get the spanners out. Alternatively for the tilt/rotate table that makes the machine 5 axis only when it is mounted, I'll tram the table to horizontal and if it is say 0.001 degrees off the homing angle I make the 'machine position' offset angle in say G54 'B' axis= 0.001. When it runs a path with go to g54 with a 'B' angle of 0.0 degrees it will kick the table 0.001 degrees when it reads the g54 offset settings. (all other angle will be relative to the offset value for g54 'B' axis) Thinking out loud if this could be applied to your machine and something was supper critical, you could have B axis value spot on in the vertical using G54 and spot on in the horizontal using G55.
Yes but in this machine there is a curvic coupling in this B axis. And this can't be adjusted with compensation or parameters. It's a mechanical coupling that clamps the B axis at 45 degree intervals. When the B axis is unclamped than yes that can be controlled the way you describe. But for most operations it is better to clamp the B axis if possible.
When I was @ Mazak on the Apps probably about the same time this machine was made... we recommend you check all your axis and spindle orientation @ least once a week...minimum,,,,,,if it’s causing you that big a problem you could always replace the axis plural for new... but for a 2006 machine probably more cost effective to buy another machine......I recommend okuma 😂👍😃
The test bar has according to it's test sheet is in spec. It can have .005 mm of runout at 330 mm from the gauge line (This equals .000197"). At where I'm showing it in this video is 550 mm Out on the test bar (14"). So taking that the theoretical runout of the test bar at 550 mm would be .008 mm (.000328"). Also the roundness of the test bar as measured on its test sheet at .002 mm (.000079") Not enough to worry about. So the rest of the runout is in the spindle at the 14" out. This is Sandvik's certified test bar for this spindle. Everything has runout. No there is no more accurate test bar available. These are very small amounts. If you check most machines they wont be any better than this. Probably worse if anything.
@@EdgePrecision thank you for your detailed and comprehensive reply. I have been enjoying your videos for a long time and every one teaches me something new.
There is no backlash. These are the clamping angles of the B axis. There is a clamp coupling in the B axis that clamps every 45 degrees. The only way to align it is mechanically.
A few things that might help. A ; always consider test bar as perfect don't worry about manufacturing tolerance it is you reference so it is perfect. Most important I think you miss information. At B0 you must dial in both ZX plan and ZY plan and do the same at B90°. Even a new machine have a bit of run out at the end of the test bar. You must find middle of run out at the end and mesure on this line. at both the start and the end of the test bar you can adjust height on the test bar by moving Y axis. Not easy to do even after several years for me! 5 Axis machine are a bit tricky to set, in your example there is something more than just B orientation alignement defaut. Also what can help is to measue with and without B axis brake because sometimes B axis brake "clutch" don't know the exact word can be deformed. but when you see good alignment at B0 and not good at B90 that mean there is an other default soomewhere else.
There are two kind of alignment procedures for this. In this video I’m showing aligning the curvic coupling in the B axis. This is mechanical. Kind of like a turret in a lathe. It can clamp in every 15 degrees. After this witch I didn’t show in this video is to reset the B zero of the B axis to coincide with this alignment. Otherwise if you unclamp the coupling the B will shift slightly between clamped and unclamped. The coupling is important for turning tools so they repeat very precisely. It also comes into play milling square on the OD. I did rotate the test bar to average out any slight runout it had and indicate the high spot in the radius. There are people on the internet that think they can get these machines perfect. But they either don’t know or have never done it before. I’m pretty happy if I can get everything in at less than .001” all along the test bar at B zero and at B 45 and B 90. In all the axis. X, Y, Z. I didn’t really show all this in this video.
I wasn’t referring to you. It is obvious from your comments you are experienced. I was meaning in general. I need to be more clear in my comments. Sorry.
Is that the right way to put the indicator? It looks to me like it should be placed on the side, or top/bottom at 0⁰... AM I MISSING SOMETHING? It looks like you're using a long-range dial indicator. Something for a lathe, not necessarily a high precision milling machine such as this. Granted your lookong for variation, you'll have some regardless. You'll just have to work that into you're toolwork/programming.
I have no idea what you are referring to. It’s been a long time sense I made this video. I don’t know what you are referring to by a lathe indicator being different than a mill indicator. There are indeed different kinds of indicators. But they are all used on all types of machines. There are none specifically just for a lathe.
Intimately familiar with this fun project, just finished E670HII replacing mill spindle covers as we speak 🤣😂🤣😰 Operator-“I’ve never crashed it!!!”, Me-“Uh hunh, whatever helps you sleep at night 🙄🙄🙄”
It is somewhat easy to knock the B axis on these machines out of alignment. I have done it using a high feed mill when the inserts failed. I don’t know why that is. As you know after doing the job. How large the diameter of the flange is and how many large bolts hold it. Maybe there is something in the roller cam drive/coupling that slips I don’t know.
@@EdgePrecision this is second machine I’ve had to do this on, had to rehome B too droop was WAY up there @ like -430! Was dropping .020” at end of test bar when unclamped vs clamped. -21 after rehoming with .002” droop now
@@scottlampert1628 I didn’t show this in the video. But once you get the B axis aligned mechanically when clamped. Than you need to reestablish the B axis electronic zero to coincide with the new mechanical location. Otherwise when you unclamp the B axis it changes like you are describing. There is a procedure for this. It is explained in the manual. On this video I got many comments saying why don’t I just reset the B zero. But they don’t understand there is a mechanical curvic coupling and a electrical B zero. They both have to be the same. So to put it simply you can get rid of that sag when unclamped be readjusting the electronic B zero. If that makes any sense. I hope so. Thanks.
Its a combination of way oil and coolant residue that the water has evaporated out of. Even if I took the time to clean it . It would be back in a short time. I have never been in a gun turret but with all the gears and lub in them I could imagine.
Those thingy Are "Telescopic Rapid Linkage Mechanism" , I did a quick search commonly known as Telescopic Covers these high speed ones are Reinforced with such Links When something breaks down here , I think I should not open a precision machine let alone be loosening those bolts hehe , So I observe the experts who come to fix ours , but I think it can be done with experience. Great Video Sir , you now know your machine "Inside Out" , did Mazak advised / helped you with this ?
No Mazak didn't help me. This machine is obviously no longer in warranty so I either do the work myself or higher a guy here in Houston that is very good with Mazaks.
Gee whiz Peter, when you give people a tour of your machine you don't fool around! Thank you for all the climbing in and the insight on the process.
I second that, thanks Peter.
and I third that
litterally noone else will share such an process with us.
thanks for that.
Crooked machine, everything is made of rubber, and the dude still splits a thou in 10" deep bores.
You're one hell of a machinist, Pete. Not even joking.
Enjoyed Peter! What a beast of a machine.
ATB, Robin
Thanks, always wondered how accurate these big machines were vs how much you had to adjust your processes to make up for the inaccuracies. Very educational. One other note... couldn't help but think I was going to hear someone yell "cycle start" while you were inside that monster... Thanks for posting - always learn something sir.
A pantograph. That's what the zigzag things on the way covers are called. Awesome machine tour!
I feel your pain. I hate digging in and doing alignments like that.. Nice vid.
Peter, you lucky dog! I would have bet money that thing was going to move on you when you tightened up all the clamp bolts after adjusting! It appears that it didn't even change your 0/90 positions a tenth or two! Thanks for showing this to all of us -- really fun to watch, and highly appreciated!
Masterpiece. I can watch you indicate things for 1/2 an hour. Not being the least bit facetious. Learned a tremendous amount about machine construction.
An interesting video thanks Peter. There is such a lot of effort gone into designing and building that machine.And the machines that built that machine. No wonder they cost so much and seem to hold a lot of their price especially the larger end like yours.
👍 Thanks again Peter for your time. Just a great insite into your skills again. Not just programing tool making and maintenance. Great.
From a brit who only has 50 + year old machines to work with. 👍🔧🔩
Thank You again for showing us such a careful, thought out, and functional perspective from behind the weighs. This video was very informative. I purchased a used box of crayons, and now I see i need to look in the bottom of the box. My rainbow purple, and burnt sienna crayons may be at the end of their serviceable usage (huge smile).
Very interesting seeing inside the machine ~ slide ways, lubrication, hydraulics etc many thanks.
What made you want to check it? Did you notice a finish cut not quite right.... a sound that didn't quite sound right.... or just a period check. I'm just curious what the temperature was on the test bar with lights and the surrounding temperature. Amazing what a teeni tiny difference makes on a bar that long. Now you think I'm mad I guess 😲😉
It has a built in breakroom.
You mentioned that they should have made the tab for the adjustment stronger. I can tell you that you're right. The tab on our Integrex ended up snapping off during adjustment and now we have to do a convoluted setup with some clamps to adjust our B-axis. Hopefully they have a better or stronger setup on newer machines.
It wouldn’t be hard to machine one heavier and maybe a stronger material. If mine ever breaks that’s what I plan to do.
@@EdgePrecision Whenever we've needed to do an alignment, we've always needed the machine running again asap. One day if we've got the covers off and have some time, maybe I'll do the same.
The newer integrex (at least the I series) doesn't have the hard clamp every 45 degree.
Instead it just hold the b axis with a "soft" brake.
Not as strong, but it's really easy to adjust since it's done in the parameters.
@@Rasmus661 In my little experience, that is a real problem when doing heavy machining.
It seems that now days a lot of machines can't handle that anymore. Of course there are modern adaptive toolpaths, but they don't work for everything.
We do a lot of heavy milling at work in steel and also GS42CrMo4 castings and this stuff kills spindles and especially milling heads.
And it seems to get harder to get machines with milling heads that can handle this kind of load, there are designed for different strategies now. I heard that we even did tests at a manufacture before buying a machine ten years ago and destroyed the milling head within two parts.
@@Rasmus661 I was wondering why one couldn't just dial in the B axis reference position with machine parameters, that explains it: in addition to the servo drive, the B axis has an indexable joint (curvic coupling?) which can be clamped hydraulically. Thanks.
The dreaded test bar. Shows about everything from spindle runout, bad spindle taper to error in machine geometry.
I would check machine squareness with granite square or so, before adjusting milling head parallel to X (up down) axis, btw. Otherwise you might have no error on dial indicator, but a head that is parallel to an out of square axis. Same goes for Z (left right) axis.
Good thing that capto taper is still good. Must be a nightmare to rework those, compared to ISO40&50.
Thats a hell of a machine man! Huge! I have changed out a ball screw once on an old okuma lathe. Lots of fun. Ours was only 1.75 inch. I would had never thought they built ballscrews with through coolant. That's crazy.
Longer the ball screw shop temperature can increase or decrease the length constant temperature equals no growth
Love your videos. I'm new to Integrex, only been running one for a year now. Please continue to make videos like this
Impressive machine tour, great insight into dimensions, proportions and design concepts. Thx. Stefan
This is The Inside Tour with all the bells ans whistles! Fantastic.
Hey Pete, I was wondering if there maybe a setting you can adjust so you can have zero in both directions. They would be some kind of encoder readings that tell the computer when it reaches position.
Never mind, I just seen you reply on how it mechanically clamps.
You can compensate in the plc for run out
I love those way cover scissor guides. I crunched a set on the Haas EC1600 I ran. Those are some high maintenance way covers.
This is where lesser cost machines save money. But in the long run it costs the owner more. We had continual problems with the Quantum mills and their way covers that this would have corrected.
I don’t usually watch cnc machine work, but i watch yours,the best cnc machine work and channel in UA-cam .thank you and I really appreciate the effort you’re putting to share with us your works. Something is in my mind for a long time, how much did you paid for this machine?thank you once again
I purchased this machine used when it was around 6 or 7 years old. New these cost between $900,000.00 to 1,200,000.00 depending on the options.
Quite a bit going on inside that thing. It really is a remarkable piece of engineering
It's ideal to check the machine for twist like a piece of licorice before aligning the head. You have a nice machine. I'm a ~ retired Okuma tech. This is my favorite channel.
Thanks Robert! Yes you are correct. And because you know how involved that would be for a machine like this, I wont go into a explanation. But this machine is turning very straight and true. These Mazaks have a weakness in this area. And so it has to be watched and kept track off. Also this has to me running into something a little hard (The reason I checked it in the first place). I have never been into the B axis of this machine but I suspect something isn't exactly correct in there. But you know the amount of work it would take to do that. When I bought this machine I noticed the new looking paint in the B axis milling spindle compared to the rest of the machine. I asked the owners if it had been replaced. They said it had for some electrical problem with the spindle motor. But I suspect that someone maybe didn't quite do something the best way in there. To take that spindle motor off and disassemble the B axis is more than I want to do. I can work around these small insurances.
@@EdgePrecision If it was another time (Non Carona) and I wasn't here on the east coast in NJ, I'd be glad to help you out. I really enjoy this channel.
Wow, how much does a machine like that cost new and how often would you calibrate it? My guess is that it would be easier to get the machine company experts out once a year but fear that would be very expensive.
Sweet video and your skill set is mind blowing!! Not a lot of people like to share their knowledge, but you are quite the opposite! Would love to work under a machinist/programmer/maintenance tech like you!!
so there's no difference with indicating from the chuck jaw compared to the face of the chuck?
Great video. Thanks a lot for showing mechanical side of your machine!
Very impressive machine Peter, thanks for sharing this!!
I have one question
Instead of loosening the bolts and adjusting
Is it not possible to adjust with CNC parameters?
With Mitsubishi CNCs of this era, I feel that I was able to adjust the runout of all axes.
Not for this because what I'm doing here is adjusting the curvic coupling location. It has moved because I bumped into something a little hard. So I had reason to check it. This machine has both a full B axis and a coupling when it clamps the B axis at 45 degree intervals.
That’s very cool getting an in depth look inside a machine that large. Closest we have, in size only, is a Devlieg 4K-96. Been working on that lately. Nowhere near as sophisticated. Way cover accordion is the word you were looking for I believe. Although I know you know that already. Good video 👍
Thanks Jason. It seems that when I turn on a video camera I can’t remember anything.
Edge Precision I don’t even need a camera around me to forget stuff I’ve known most my life. I first thought you were checking the front of the bar in the vertical position. I was trying to figure out how you were going to get that straight. Then I woke up. 1600 IPM damn that’s nice.
Is there any sort of maintenance or alignment that has to happen with the capto system?
There is nothing that can be aligned, no adjustment. Its either there or not.
Nitrogen accumulators lose gas over time, so the lifting force changes, there is probably a range of gas pressure that Mazak specifies.
Good thing is that most accumulators can be recharged, bad thing is that if you do it wrong they can be dangerous.
Thanks for the tour. That machine is amazing.
Mazak makes some badass machines. We have one of their 6kw Fiber lasers at the shop.
On OKUMA with OSP 300 control
You can comp it out, but that's a cheap way of doing it. Get it mechanically as close as possible with adjustment bolts.
Those are good machines. They do require competent technical assistance. Got 5 of them in my shop.
Could this have been caused by the little crash that you had a few months ago? Or just sagging / shift as a result of use and loading?
How often do you do the inspection? We do a ball bar, test bar, and RTCP check on a monthly basis on our machines but we have to to keep our part population tight. We do an annual laser calibration as well.
If you are running production style work. I could see the need. I only check things when I feel the need or something doesn’t seem to be cutting right. Because of the type of work I do and also no one else runs this machine but me. And I know the machine.
Hey Peter, Im Elect eng, worked on robots. There should be small offset adjustments that the service technician can do for each axis deep in the control. its like when you zero the machine to a job, but would be found deeper in maintenance section of the console. It would save you time pulling it to bits. fast, click click done. See if you can get the techo manual or ask a Mazak expert. The only mechanical adjustment should be in the axis of no movement. i.e. Same as what you were doing with the dial indicator but the front edge of the bar instead of the side of the bar. Side of bar adjustments (rotation of head), digital adjustments. Also, in an Okuma, the spindle will lock in 12 different positions when a lathe tool is in the spindle. Its a gear lock so the spindle wont turn. Your spindle was free to spin.
It only makes sense that the Mazak has this lock as well, or else the servo will have to fight the tool to keep it still while the spindle is used for turning. If I were you, I would lock spindle in the down and also lock in the horizontal while doing those adjustments. Lastly, Thank you for showing us around inside the machine. I love your Mazak, nearly bought a single spindle version from auction and always wanted to see inside one. If you ever have the ways exposed on the 2nd spindle, pls film it for us. I love your stuff. Thank you from Australia.
I don't think I made this clear in the video. In this machine there are two things controlling the B axis. There is a cam roller rotary worm drive and also a curvic coupling. What I'm doing in this video is adjusting the mechanical curvic couplings zero location. This can't be done with parameters because it's a mechanical coupling that clamps the B axis every 5 degrees. This coupling has teeth like a gear on it's face that couple together for a very positive and strong clamping. This is more designed for turning tools and milling at the 0-90-180 degree locations. There is also a full B axis that can resolve to .0001 degree location. This would be adjusted as you say. I didn't show this in the video but the zero of the B axis has to be reset after adjusting this mechanical couplings zero location. So of course they are both the same. There is also pitch error compensation on the B axis Cam Roller drive but this has to be done by a Mazak technician.
@@EdgePrecision Yes, this makes sense. Thanks for further explaining it.
Fantastic show and tell. Thank you!
Awesome Peter, really enjoy your videos. Regards from Norway.
No matter how good of a machinist you are some day you will crash a machine, especially cnc s. The fact that your machine crashed with the tailstock reminds me of an 1995 Hitachi Seiki lathe I worked with. It had a sub spindle and it was impossible to crash it. The machine always knew where the turret was in relation to the sub spindle-tailstock so it alarmed when you were to close, in jog or auto. Anyway you probably saved yourself a lot of money doing all this work to calibrate the machine. I imagine in the US this kind of labour is very expensive. Maybe you can give your opinion on linear guide ways versus box ways in heavy duty machining. Cheers!
Thanks for the tour
Awesome video, Peter. Your videos are the best.
Thanks Peter. Does the test bar deflect under it's own weight in the horizontal enough to show up on indicator? I just wonder if that accounts for some of the error there.
It is possible but I don't think the amount would be enough to worry about.
FWIW was wondering the same thing this morning. I ran some numbers for a bar 2.5" in dia and 14" long from the supported end the deflection is about 0.0001". For a bar 1.5" in dia it's about 0.0003".
@@billdlv Less than I thought, but I guess supported by it's own weight is one thing and doing any work at distance is whole different thing. Thx.
I had the same issue with the squareness on my Integrex. The XZ and YX were not tracking straight so I learned how to build a comp table to adjust electronically. Now everything is .0001 square and I didn’t have to adjust manually. The Matrix controls allow this electronic straightness as well.
First Peter nice video, again thanx. From what i saw the B axis wasn't rotating a perfect 90 degrees from the directions of error. I have been programming and setting up Campbell rotary grinders where I'm at with a B axis. I program a B rotation with a backlash move to always come to my final angle in the same direction, seems to improve the rotate accuracy. Let me know if i was seeing that right about the B. Good job, I always look forward to the next video.
Cool video Peter! Did you measure the deviation in the Y-axis direction also? Not that it's easy to adjust but just to know. Used to work on an old (2005) mill/turn also and that machine had a error in the y-axis direction related to the testbar.
Yea I checked it in all directions. Along the total length of the test bar there was less then .001” deviation. That was the best I could get it. I also checked it a different B angles.
how often do you have to do this?
As little as possible. But usually about twice a year or if I have a reason like crashing a tool into something that puts a lot of load in the B axis.
Great show, amazed at you and that mazak . I think it bieng an old machine.and seeing the maintnence to keep it in tolerance . Very fascinating . what kind of old machine would I buy, mills Bridgeport 9x49 K7T horizontal or the brown and sharp with the omniversanal table lathes hardinge dsm HVLV love the hardinge stuff. good to see ya agan
Are there counter weight for any of the axis.
Answer my own question after watching again. Yes a Strut. Thanks
With the liquid cooled leadscrews, I'm guessing that this uses rotary encoders on the servos, rather than liner encoders? I would think that the encoder position error could be fixed in software, but not something they did back in 2006. Any idea what the expected lifespan is for a machine like that, before a rebuild? I know the iron will last, but all those hoses and cables are bound to fail with age someday.
Some machines of this spec have linear encoders for direct position measurement, looks like this one relies on the servo.
All of those hoses are rated to last longer than your lifetime. Things that are going to fail are moving parts(rails, carriages, bearings) or the computer.
Good work I am sure glad I am not the one to do the maintenance on it. I can't imagine what Mazak charges for repair work.
You mention tolerance changes from cold to warm machine. Do you take that into account when checking tram?????
The machine has thermal comp built in.
we had some intergrex at Lockheed and it would get bumped now and then. Nice video
A gentleman never crashes, he "bumps".
@@tightmf I've got to remember this saying!!
I am sure you did but I neveraw you check to be sure that the bar was exactly parallel to the spindle. You would wiggle the indicator to find center of the bar and then traverse, never saw you wiggle it at the other end to be sure it was still on center and that the bar wasn't pointing either towards the door or the wall. This could easily contribute to additional errors if you didn't check that as well. I would not be surprised for the head to be worn nodded down.
You are correct if the bar deviates from centerline in relation to the indicators stylist. But it has to be by a fair amount to cause a significant error. The bar is 40 mm in diameter or .787402" radius. Now lets say I was .001" off center from one end to the other when running the indicator down the bar. (You saw in the video I indicated the high point of the radius). Now at .001" off center the reading would vary to .787401 or only change .000001. From .787402 (20 mm Radius) to .787401 a .001" over on this radius. so yes there is a deviation but not much. the tip or nod of the spindle isn't this much. I checked that. But even if it was as you can see it not enough to worry about for this measurement.
Is there also an adjustment if the head is out In the Nod? If the Nod was out but the tilt was perfect, wouldn't you still see an error checking the tilt on a round bar as the indicator will be falling off the apex of the radius?
Interesting look inside the machine as always. Did you have an issue making a tolerance that caused you to check the tram?
Thanks Bill. I’m waiting on something with inspection. At these times I check things like this and do Maintence issues for the machines. I don’t want to have to in the middle of a setup. Don’t ask how I know that. In reality this seems like a lot of error. But even this can be compensated almost away with fixture offsets. Remember that test bar is 14” long. Most machine work isn’t that long so the error isn’t that much in shorter lengths.
0.0028 degrees off parallel is pretty good. I am assuming it is a 500mm bar, I couldn't see the dimensions so I was eyeballing it.
It is 300mm. long, not 500mm. Target is 0 when you align it, but usually I get it down to 0.01mm/300mm (10 micron/300mm).
There are other factors in the picture too when you align the B-axis.
What about the tilt of the head if you measure on the bar frontside or backside when its vertical (straight down). Is it possible to adjust that if you got any errors?
The front to back had a little error but not that much. There is no way to adjust that direction. Except maybe adjusting the machines level.
Pete . You can go Bonkers you can get it right on specks . Then check it MUTCH later and get a. COMPLETE different reading the machine Temperature the floor Temp. All comes to play 😨😎
I'm 25, just did a spindle alignment and turret alignment on doosan puma today. Have you ever bumped in X- while turning and had to make the spindle parallel to Z again?
Yes I have done that before where a machine was crashed.
Such a beast of a machine..
That gets the rotational runout. But how about runout front to back at 180 and 270?
This runout I showed has to do with how the machine deals with lathe turning tools and clamping the spindle. It really wont affect milling tools. The turning tools can quite easily be compensated with wear offsets for this. I didn't show in the video about checking the front and back but I did check that. There was very little error that way.
Do you also do frequent kinematics measurements and adjustments? Would really love to see a video about that. I'm doing some kinematics checks and adjustments in fanuc, heidenhain and siemens controls lately. I'm interested how that works on a mazatrol controller.
I personally have never done anything but mechanical type of alignments on this machine. On my horizontal mill I have run a Ball Bar and adjusted the servo and backlash parameters. To improve it's circle roundness. Also a Ball Bar report will tell you about the squareness of the axes's being measured on the machine. So it will give you some idea of the machine's level, more than a bubble level can. I don't know if this is what you are referring to by Kinematics. I always thought of kinematics as something built into a machine that really could not be changed. With the exception of some servo parameters and the leveling of the machine for squareness. To it's original manufactured condition.
@@EdgePrecision thank you for the reaction. For the mill turn you can adjust the pivot point for the b-axis in x and z direction. If you put an ball in the spindle and verify that your machine X and Y are 0 on the main spindle with the b-axis on -90 degrees. (Think its zero on yours). Than have 2 indicators on the X and Z direction and move the b-axis around with active tcp. If you get any error you can adjust it by changing the kinematics parameters. I had problems using the same tool in different angles on my SMX2600S. Even worse on the subspindle side since the correction for main en sub is seperate.
Ok I see what you are referring to. Yes the Mazak has parameter settings for this. But my machine never needed this changed. I believe on the Mazatrol they are BA61 for X and BA62 for Y. My Cam software (Esprit TNG) also calls up the BA62 setting in the program so when rotating the B axis the machine doesn't over travel the machine.
Is there a override menu where you can add in the .0005" to the vertical axis to get it on zero. ?
If memory serves, yes there's a few parameters on the newer integrex that you can use. On the older ones like this, you must do it mechanically like this. They still reccomend you do it this way though.
that's one darn big machine.
I’m surprised you can’t adjust the runout in software.
Too bad you can't grind the tapers in those like you could a CAT or BT taper.
What version of Windows does the control run? 2000 Embedded?
Yes windows 2000.
Dumb question: Why the need to tram B-axis instead of offsetting the B-axis rotation abit electronically? Isnt it the same angle? I think I am missing something though
If this machine was just a 5 axis milling machine what you say would be correct. On a mill turn machine you can also do turning. So the B axis has two different methods of alignment so to speak. But you can actually use the B clamp feature when milling a 0,45,90,135,180 deg. The B axis has a coupling (Curvic coupling) and a servo drive that can position to a .0001 degree restitution. So if the curvic coupling is out of alignment it can only be aligned mechanically (It is a psychical coupling hydraulically clamped). This is necessary for running turning tools because you don't want the B servo to have to hold against say a long boring bars force. Also the coupling positions very accurately with no give that a servo system could have. I didn't show this in the video but once the coupling is aligned you do have to make sure the B servo drive coincides with the coupling. Or to put it another way they both have the same zero point. Also the mill spindle has a similar coupling to clamp at different angles (Every 15 degrees) for turning tools. Hope that makes sense and answers your question.
@@EdgePrecision wow, I have to read that answers couple times but looks like it's more complicated than I thought. Thanks a lot
I love how Mazak's commercial videos are all clean CGI and techno music but we all know the reality of it is ear-splitting noise and a metric ton of grease and oil :-D
Will you do and explain a center rotation one day?
I'm not sure I follow what you are asking. When you say center rotation what are you referring to? The rotation of a work plane in a program? the tail center rotation? B axis rotation center?
Hey Peter - I've been watching you videos for a while now and have been totally enamored by your work. I really appreciate all the work you put into the videos and the knowledge you are giving to all of us.
I am currently in the market for another machine to compliment my VF2SS. I am looking at adding a 4th/5th axis attachment to my mill or getting a live tool lathe. I have found a few early 2000 Integrex's. I was wondering what unknowns or complications do you see with getting an older Integrex up and running? How difficult/expensive is the initial tooling/CAM maintenance, etc? I don't know what I don't know as far as big issues with these machines that will cost me much more than the initial machine cost. Also how difficult is it to program? Sorry for the loaded question. Thank you so much for all your wisdom and knowledge.
Cheers,
You remind me of a friend named Conley Ownby. Great machinist.
Hey Peter, I believe that Sandvik test bar would have its calibration data in microns. Wouldn’t you be much better off with a Millimess from Mahr or a Mitutoyo Hicator. That Intrepid is only reading 12-13 microns and with hysteresis its probably around 15 microns at least, so you’re not getting a good reading. Also, as you said, you can move it easily with your hand and working against gravity might not be helping when doing sideways calibration as the end of the calibration test adapter will be sagging a micron or two because of gravity. Wouldn’t it be best to calibrate the B axis when the spindle is facing down instead of left or right?
If the test bar is sagging from gravity than so are the tools in the horizontal position. I don't think there is enough error there to worry about it. I have inch tools so that what I use. I do have a higher resolution Interapid indicator reads in .0001". Inch or metric the indicators do the same thing. No one is any better than the other.
Some backlash in the B axis especially noticeable when the B axis clamps and unclamps and maybe the B axis bearing is not 100% if someone has crashed it in the past? Looks like backlash comp might need re setting in the control.
you may be able to calibrate it out in the servo driver. but that may cause more headache than its worth, assuming they have free tech support like i would give them a call.
Look at my reply to Mathieu Fresco above. I enplane this in detail. Thanks.
Tramming.. is a complication issues....Also is the the matter of parallelism of two rails (or are they truly on a plane?any wear, clearance?) of the axis it is sliding on! Are they all of three pieces per axis!
Yes but if the coupling in the B axis of this machine gets knocked out. This is the only way to realign it. You think Mazak put those jack screws on there for a reason? I did run into something a little hard that's why I was doing this. I should have explained in the video.
literarely reading comments before watching the video to prepare myself on what should I focus on
Hi Peter you allways know what to do.
Hmmm.... interesting. I would have thought that it would be possible to adjust the amount of rotation that the head makes between stops. I would think that the 45 degree tilt ( or 90 degree tilt - not sure) between stops would be handled by some sort of electrical device counting some sort of pulse. You'd need that to decide what angle the head was at anyway. I would think that it would do the same between the stop positions. A clever designer might even have different pulse counts for each specific segment of the arc the head rotates through. To tram it, you'd then set the bar straight in the horizontal location and then adjust the number of counts required for it to get the the vertical position. But I'm just a "back seat engineer". I'm sure the made the machine that way for a reason.
Wow! if there is some sort of mechanical device that sets the angular rotation amount.... phew... I can't begin to imagine the accuracy required of such a device. Yowzers!
Anyhow, thanks for sharing this. It is absolutely fascinating.
The B axis can position to .0001 degree position accuracy. But it also has a coupling that clamps every 45 degrees. Mostly for turning operations. But except for full 5 axis milling almost all machining is done at the 0 and 90 degree angles.
@@EdgePrecision does adjusting the stops change where the incremental positioning ends up? Or is that except by an encoder or something?
@@EdgePrecision So something similar to a hirth joint that gets clamped together to set the head at a given rotation? Just out of curiosity, can you rotate the head 180 degrees from the front facing horizontal position? And, if you can, does the alignment stay true? That could lead down a serious rabbit hole if it's not consistent with the other measurements.
it's was littlebit surprise for that the mazak control sofware runs on win2000. came littlebit nostalgic feeling because last time i have usy any windows is about 10-12 years ago, and question can you put some linux operstion system to mazak tool use it that way. i just means is ti posible at all, and then second thing came to my mind bcause in one vodeo you show you have ethernet cable going on mazak to your laptop and you have internet connection, is there any danger that virus or some else unwanted progmam find it's way to mazal machine, but i very much to like watch that you do parts that looks like they are imposible to do but seemingly they are not. and one question yet, how long it take you to came that good with mazak machine because i think if i start to learn how to use it it's take several years to learn to use it somewhat deasently so i can make some rather easy peaces on it. 👍👍 🙂 and hello from finland 🇫🇮 ( sorry if i ask stupid question but i have to because i have think these things almost always when i watch your videos. o and what was the name of mazak's controll software. was it mazatrol or something like that then i can look more info on internet when i know it.
I don't know about the linux operating system? The PC in the Mazak is a industrial one configures specially for the Mazak. All Mazaks actually run on a Mitsubishi Meldas control. All the electronics are Mitsubishi in the machine. The machine itself is not connected to the internet. Its on a separate network just to my laptop. The only files that get transferred to it are G code files. I don't know but I doubt a virus could be transferred in a G code file. Isn't a virus a program? The code file wouldn't have that kind of information. But what do I know? I just run machines. But I suppose its possible over the network cable from the laptop. We do have a server here that suppose to block that and my laptop has virus protection software. But anything is possible. The software that a Mazak control runs is called Mazatrol. But it is a proprietary software that Mazak uses set up for the machine it is running on. I don't think it can be purchased and set up on another machine. Like say Mach3.
Can the level of the machine effect the squareness of the spindle. And how do you level a machine like that?
Wayne D
Yes it can, everything flexes. But leveling something that large takes a whole day id bet
On a say a vertical or horizontal mill the leveling of the machine makes a big difference. Also for this type of machine it comes into play to get it to turn straight. Witch this machine is doing fine. But this adjustment cant be controlled by just leveling. The squareness of the axis's to each other can but the indexing accuracy of this is probably tied in to something being damaged in the coupling in the B axis. I don't want to tear all that apart for this (To look at it another way. It would be very expensive). This small inaccuracy can be dealt with other ways. It just isn't worth it. What I'm trying to show here is for 99.999% of all machine work this is good enough. I bet if you really tested machines that most machinists us every day. They would all have these small inaccuracies. They just don't know it because they compensate for it in other ways. What do they say. You got to hunt with the dog you have.
@@EdgePrecision Thanks for the reply. I have Vertical CNCs and they sit on the shop floor that is not built for heavy equipment, but it is what it is so we have to relevel every now and then to keep things square. Love your channel BTW.
i wonder if you had some mechanical backlash on the B axis that could cause the symptoms You experience. i can not tell from the video but if You position without the backlash taken into consideration via using an unidirectional approach to establish the position of measurements than this can be what you get. also if it has a worm gear drive for the B axis than that should have some way of backlash adjustment.Tsudakoma on their rotary tables use evolving profiled wormscrews that are fine adjustable by changing the longitudinal position of the worm screw with its support bearing assembly (bearing + housing) altogether.
There are two different things here. By the way B axis is driven by a roller cam drive on this machine. No worm drive. But there is also a mechanical coupling that engages at larger angles. This clamps solid like a turret on a turret lathe. So in this coupling can get knocked out in relation to the B axis cam drive (witch can be zeroed anywhere like a worm drive). So if the two aren’t aligned to each other. When you couple/clamp the B axis it shifts a little. So first you align the coupling. Than re zero the cam drive.
@@EdgePrecision Yup Roller cam makes much more sense for the longevity and frictionless motion. obviously You have more insight to how this B axis is actually constructed. i highly appreciate your answer. made me curious enough tto dig around the internet how Mazak designed this B axis. however please don't ever hesitate to make videos on maintenance tasks, mastering maintenance saves a ton of money for the smart ones.
i love to watch your channel mainly because You do challenging parts and take on difficult to tackle tasks. we in our shop are kinda leaning to the same direction with our endevours tho it is challenging with 30+ staff . doing challenging parts in smaller quantities really makes much more sense to get the most out of the machines in terms of profit while trying to keep maintenance costs and amortization low all be it one needs enormous knowledge to succeed on that road. for me it seems You really enjoy working on your own, we are pretty similar in that regard tbh. also i imagine the old days when You were running a whole shop did their number on You considering the stress factor involved just purely from the financial side of things. anyways ♥ from Hungary.
oh and please dont hesitate to show us what is behind that cover on the B axis if You ever get to remove it for any reasons, i would love to have a sneak peak into the hearth of the B axis motion 😁
I am from the old school, before the days of CNC!
In the early 60s with similar situations on a horizontal mill or when attaching a grinding attachment to a vertical lathe, etc.
1) We would indicate both spindles (zero, flat, concentric!), just like the spindle to the table on smaller Bridgeport!
2) It appears in the internal inside tour, there may be an "overbalance in equipment mass weight" towards the machining / tool / spindles area. I.E., overhang / overbalance on the ways from the backside towards the front.
*** Needs an articulated counterweight similar to that of a pipe layer! * Might be overkill *
C) Would it be feasible to do a complete range of motion test (X-Y-Z ), in a sectional comparative (ROM)? Old School would be looking for wear on load bearing carriage ways, etc (all 4 corners). Laser / target sight, main spindle to articulated spindle (gun sight adapted?). Consider: (14 years!) X (how many cycles?) X (tools changed?) X (tool bumps?) + (grit under the carriage?) = Too many questions! (way lube & wipes ain't perfect).
D) That irregularities in the spindle lock brings a lot of other questions.
Modern Girls need a little more, TLC!!!!
very interesting for me in my machine tool world
I'm used to 5 axis milling with FANAC controllers. One is a milling machine and the other is a large tilt/ rotate table attachment .
In both cases I just read the angle deviation away from the homed position angle after tramming in, then deep dive into the controller and add or subtract that value to the existing one that sets the homing angle for that axis. I think from memory there is a +/- 5 degrees adjustability doing it this way, so its a bad day if you have to get the spanners out.
Alternatively for the tilt/rotate table that makes the machine 5 axis only when it is mounted, I'll tram the table to horizontal and if it is say 0.001 degrees off the homing angle I make the 'machine position' offset angle in say G54 'B' axis= 0.001. When it runs a path with go to g54 with a 'B' angle of 0.0 degrees it will kick the table 0.001 degrees when it reads the g54 offset settings. (all other angle will be relative to the offset value for g54 'B' axis)
Thinking out loud if this could be applied to your machine and something was supper critical, you could have B axis value spot on in the vertical using G54 and spot on in the horizontal using G55.
Yes but in this machine there is a curvic coupling in this B axis. And this can't be adjusted with compensation or parameters. It's a mechanical coupling that clamps the B axis at 45 degree intervals. When the B axis is unclamped than yes that can be controlled the way you describe. But for most operations it is better to clamp the B axis if possible.
@@EdgePrecision Ok I understand. Best of both worlds in most situations having the coupling.
Welcome to the machine.
When I was @ Mazak on the Apps probably about the same time this machine was made... we recommend you check all your axis and spindle orientation @ least once a week...minimum,,,,,,if it’s causing you that big a problem you could always replace the axis plural for new... but for a 2006 machine probably more cost effective to buy another machine......I recommend okuma 😂👍😃
HI PETER,, I WOLD NOT FANCY CLIMING IN THERE WITH IT POWERED UP!!!!!!!!!!!!!!!!!!!!!!!!!!!!REGARDS REALY GOOD TOUR.. THANK YOU...
Could you make a better test bar? One without the run-out?
The test bar has according to it's test sheet is in spec. It can have .005 mm of runout at 330 mm from the gauge line (This equals .000197"). At where I'm showing it in this video is 550 mm Out on the test bar (14"). So taking that the theoretical runout of the test bar at 550 mm would be .008 mm (.000328"). Also the roundness of the test bar as measured on its test sheet at .002 mm (.000079") Not enough to worry about. So the rest of the runout is in the spindle at the 14" out. This is Sandvik's certified test bar for this spindle. Everything has runout. No there is no more accurate test bar available. These are very small amounts. If you check most machines they wont be any better than this. Probably worse if anything.
@@EdgePrecision thank you for your detailed and comprehensive reply. I have been enjoying your videos for a long time and every one teaches me something new.
You can't tram it in the parameters and with backlash comp?
There is no backlash. These are the clamping angles of the B axis. There is a clamp coupling in the B axis that clamps every 45 degrees. The only way to align it is mechanically.
@@EdgePrecision That sucks, I'm surprised there aren't tapered pins. Is it a curvic coupling inside there?
@@speartoolmachine7420 Tapered pins are a waste of time
Thanks to Covid 19, I finally have time to overhaul my machines in my shop, beats quarantine.
K. Michael Rademacher
Yep doing the same, unfortunately I need parts I have to wait on for a lathe clutch.
A few things that might help. A ; always consider test bar as perfect don't worry about manufacturing tolerance it is you reference so it is perfect. Most important I think you miss information. At B0 you must dial in both ZX plan and ZY plan and do the same at B90°. Even a new machine have a bit of run out at the end of the test bar. You must find middle of run out at the end and mesure on this line. at both the start and the end of the test bar you can adjust height on the test bar by moving Y axis. Not easy to do even after several years for me! 5 Axis machine are a bit tricky to set, in your example there is something more than just B orientation alignement defaut. Also what can help is to measue with and without B axis brake because sometimes B axis brake "clutch" don't know the exact word can be deformed. but when you see good alignment at B0 and not good at B90 that mean there is an other default soomewhere else.
There are two kind of alignment procedures for this. In this video I’m showing aligning the curvic coupling in the B axis. This is mechanical. Kind of like a turret in a lathe. It can clamp in every 15 degrees. After this witch I didn’t show in this video is to reset the B zero of the B axis to coincide with this alignment. Otherwise if you unclamp the coupling the B will shift slightly between clamped and unclamped. The coupling is important for turning tools so they repeat very precisely. It also comes into play milling square on the OD. I did rotate the test bar to average out any slight runout it had and indicate the high spot in the radius. There are people on the internet that think they can get these machines perfect. But they either don’t know or have never done it before. I’m pretty happy if I can get everything in at less than .001” all along the test bar at B zero and at B 45 and B 90. In all the axis. X, Y, Z. I didn’t really show all this in this video.
@@EdgePrecision well this is my job. I am not thinking but doing it! perfect does not exist sure!
I wasn’t referring to you. It is obvious from your comments you are experienced. I was meaning in general. I need to be more clear in my comments. Sorry.
Is that the right way to put the indicator?
It looks to me like it should be placed on the side, or top/bottom at 0⁰...
AM I MISSING SOMETHING?
It looks like you're using a long-range dial indicator. Something for a lathe, not necessarily a high precision milling machine such as this.
Granted your lookong for variation, you'll have some regardless. You'll just have to work that into you're toolwork/programming.
I have no idea what you are referring to. It’s been a long time sense I made this video. I don’t know what you are referring to by a lathe indicator being different than a mill indicator. There are indeed different kinds of indicators. But they are all used on all types of machines. There are none specifically just for a lathe.
Intimately familiar with this fun project, just finished E670HII replacing mill spindle covers as we speak 🤣😂🤣😰 Operator-“I’ve never crashed it!!!”, Me-“Uh hunh, whatever helps you sleep at night 🙄🙄🙄”
It is somewhat easy to knock the B axis on these machines out of alignment. I have done it using a high feed mill when the inserts failed. I don’t know why that is. As you know after doing the job. How large the diameter of the flange is and how many large bolts hold it. Maybe there is something in the roller cam drive/coupling that slips I don’t know.
@@EdgePrecision this is second machine I’ve had to do this on, had to rehome B too droop was WAY up there @ like -430! Was dropping .020” at end of test bar when unclamped vs clamped. -21 after rehoming with .002” droop now
@@EdgePrecision approx 24” 10 bolt flange on 670
@@scottlampert1628 I didn’t show this in the video. But once you get the B axis aligned mechanically when clamped. Than you need to reestablish the B axis electronic zero to coincide with the new mechanical location. Otherwise when you unclamp the B axis it changes like you are describing. There is a procedure for this. It is explained in the manual. On this video I got many comments saying why don’t I just reset the B zero. But they don’t understand there is a mechanical curvic coupling and a electrical B zero. They both have to be the same. So to put it simply you can get rid of that sag when unclamped be readjusting the electronic B zero. If that makes any sense. I hope so. Thanks.
@@EdgePrecision tsz0 reset mode I believe boss showed me how then gotta power cycle to save new home it’s following error comp
Last time I saw that much oil sprayed around was inside a 6" gun turret. 😊
Its a combination of way oil and coolant residue that the water has evaporated out of. Even if I took the time to clean it . It would be back in a short time. I have never been in a gun turret but with all the gears and lub in them I could imagine.
Those thingy Are "Telescopic Rapid Linkage Mechanism" , I did a quick search commonly known as Telescopic Covers these high speed ones are Reinforced with such Links
When something breaks down here , I think I should not open a precision machine let alone be loosening those bolts hehe , So I observe the experts who come to fix ours , but I think it can be done with experience. Great Video Sir , you now know your machine "Inside Out" , did Mazak advised / helped you with this ?
No Mazak didn't help me. This machine is obviously no longer in warranty so I either do the work myself or higher a guy here in Houston that is very good with Mazaks.