Have any of you worked with these yet?? PMT tools Click on the "Nano Tools" link....

Mark.... I can't get over your new "identity"... I agree with others... avoid set screw holders... including the ones that are "pre-balanced". They work if you need them but you can get better results with other types of tooling. We have a FH480 with a 25k spindle. Using shrinkers, Nikken mill chucks, BIG Kaiser collet holders and chucks, ... etc. Watch out for hyrdraulics. Many aren't designed for a bunch of side-loading. Some hydraulics can though like a Sandvik CoroGrip and I believe the Schunk models are capable of it as well. 3 locks won't benefit much unless you have a Big+ spindle. Effects to the spindle are usually "long term" unless you have a really whacked out tool. But you'll see immediate differences with surface finishes, accuracy and tool wear.

quote: .0002-.0005 tolerance is very common for our parts. When they change inserts they need to be able to adjust offsets. We suspect that this where the problem is, they make a mistake adjusting an offset, then crash, DPRINT as others suggest will work well. Just need a DNC system to capture the output. Another thought... You could just keep a copy of all the offsets (work and tools) as a baseline. Then write a macro that will check for a variance in the offsets. Since the adjustments are only tenths, you could just set up a tolerance in the macro. Even with an insert change, the tool should still only change by tenths (or whatever the value equates to if you're constantly adjusting for wear). Then in the main program, have a macro alarm trigger for the bad offset input when it falls out of the adjustment window.

This sounds similar to the way the old A series cell controllers used to work. Yes, I would try and put it all under 001. Since you need to run them consecutive anyway, it should be listed under 001. Are all ops on the same pallet? What you may have to watch for is the the Process itself. It may think you have 4 stations and 4 parts once the pallet is in loop with a full load. The other alternative to try is this: 001 = wss / mc / wss program A 002 = wss / mc / wss program B 003 = wss / mc / wss Program C 004 = wss / mc / wss Program D I wish I knew the MAS system for you. The B2 is similar.... but different... Your system is based on the model A3 system and I don't know the lay out on the software. There are some Makino users here so maybe someone has an A3 sytem? But it really sounds like there's a button that needs to be on/off and it should work the way you originially had it laid out...

quote: the stylus wasn't snug on the bread-away Sorry... "break-away" On another note, had another instance where a loose pull-stud caused the same grief....

Lee... all of the Renishaw docs are online... everything from set up, parts manuals, usage, software/programming, etc. Here's a quick link to the OMP 40 . As long as someone understands that the "road map" can also cause some grief is where I was getting at. I've seen people try to point the finger at bad calibration or keep buying parts for the probe system only to find out that they had a slight twist in the machine.... Another good one... had a guy who kept calibrating and calibrating the probe and it wouldn't repeat. Re-installed software, changed about $100 in batteries, attempted to re-write the macro, adjusting probe speeds, wrote a macro from hell to comp .... nothing made it repeat. As it turn out... the stylus wasn't snug on the bread-away.. All he had to do was probably indicate the stylus and he probably would've saw it turn. Still can't figure out how he changed all those batteries without noticing that....

quote: He explained that the calibration routine should compensate for all these variations unless they are excessively large IMO... sort of. Probe calibrations only compensate errors to itself and not to the entire machine. Think of this... calibrations happen only in a relatively small area of the machining envelope. If you have a small twist in the Z axis (HMC) for example, the probe will pick up error in the XY elswhere. The check for this would be to set up the calibration gage in the same area as the feature you're probing. If you're still picking up error, then it might be something mechanical like the probe itself as someone mentioned.... or the toolchange motion, taper seating, changing probe speeds, etc. Or, simply measure the gage location (no need to calibrate for this test), then recheck the location with the probe again. Then tool change out/in and re-measure the position again. Even take the probe out/in by hand at the spindle and measure. If its still floating excessively, I'd have to guess mechanical....

So, your steps should be: op1 op2 op3 wss op4 wss Start cycle again.... Check your part process page. I haven't used or seen a MAS5 (we have B2 and A2). I think somewhere along there, the info for complete parts is not right and so its looking for another part to load before step 4. You did only create one process for this part number right? and not 3 or 4 different parts?

Add to that... are you using a spin probe with a real long stylus? Is your calibration speed different than your part probing speed? And as Mark stated, you may have some twist in the machine. Ball-bar will not always show twist. You need to check with a test bar and a machine square for the spindle/axis. Another thing I've seen this happen is a fretted spindle or taper. The probe doesn't repeat its seating position in the hole causing the calibration to shift. As Chris mentioned, recalibrating as part of the machine cycle may help the situation. But I've seen some bad ones too.....

So, your OP 3 has two stages in the cycle? Haven't tried that one in conjuntion with previous ops. What version cell control are you using?

Calling Crazy_Millman.... !!! As James said, he's the resident expert... they're a couple others that knows this stuff pretty well too... In the future though, might at least label these types of questions as "Off Topic" or "OT". Or post them in the Off Topic forum. Unless of course, you have specific questions on maybe modifying a post to work for the machine....

Toby... you've got mail.... Tool calls break down like this for 640T (which is a Quick Turn machine right? If its a 200MSY) T0101.1 First 2 digits is tool number, next two digits are offset number. The value after the decimal would be the Tool ID code from the tool data. If you're using 6 digit tool system, it would be: Txxxyyy.1 "T0101.1.1" would be something you might see on an Integrex. The middle digit would be tool direction. With all of the Mazaks you have, you should get a hold of the books. Theres alot of info in there on EIA programming. Although for these machines, learning the Mazatrol would really be ideal. Not as difficult as it looks and you'll be surprised by it.

Yep.... that works too. Then you just need a couple macro statements towards the end of the program to jump for pallet change, offset change, etc. And for this example, you wouldn't even need a M97 to repeat. A M99 will do the same thing (I think on a Haas.. it works on other controls).

I agree with Ron.... Just use one Main program and call it with a sub from a "Pallet Program" with a M98. Macro and variables can identify pallet numbers and G10 lines can be used to for workoffsets for adjustments or if the set up is slightly different on each pallet. Some machines also have a pallet page to identify the main program to run (this eliminates the use of a "pallet change program"). Use a machine variable to identify the pallet and pick up G10s, then run. Makes restarting easier as well. I've never run a Haas with a pallet changer so I don't know about them. And your program will be much shorter with only one Main program do have to deal with...

quote: I have look ahead but if I try to go faster than 75 ipm the machines seem to try to start cheating corners. I should have asked this to Dragracer as well... but what machine are you on? Was the Look ahead part of the original install or optioned later (Connarmac)? Some machines just plain have a hard time going fast. Even if the lookahead is set properly, there some that just "can't get out of their own way". This has to do with other issues like axis drives, motor types, guideway types, screw pitch (which brings up pitch comps, accel/decel capabilities), etc. You can have some adjustment by means of parameters for servo drive, vector comps (for angle of motion) to try and boost the acceleration and accuracy in the corners. It's a game of trade-offs. But these things can screw up your machine in a hurry so I do not suggest someone with the lack of this type of adjustment experience to start messing with these parameters. It could take some time (even for the experienced) or it could be as simple as tweaking a few things in with a Ball-Bar. Code-wise, some machines may force other criteria. As with James' machines with AICC/AI-NANO, he has to use certain codes to meet certain conditions. G8 is also not the same as G5 for criteria. G5 has a couple other "requirements" for use. For the most part, on the dozens and dozens of machine brands/models I've used G8 on, the "factory" settings were adequate. Some I've had to tweak in with a metrology guy or factory support. A couple machines were just simply "hopeless" and only used it to force finish accuracy and constant speed (which was at the least, "some" improvement over the original state).

Here's an example of code: N1G0G17G40G80G90 T1M6( 1.0 CARB EM ) G55X-3.1Y3.S12000M3T1 G43H1Z.1M8 G8P1 (Turn on lookahead) G1Z-.5F200. G41D9X-2.5F480. X2.5 G2X3.Y2.5J-.5 G1Y-2.5 G2X2.5Y-3.I-.5 G1X-2.5 G2X-3.Y-2.5J.5 G1Y2.5 G2X-2.5Y3.I.5 G3X-1.734Y3.3572J1. G1G40X-1.6697Y3.4338 Z.1 G8P0 (Turn off lookahead: optional) M9 G0G91G28Z0M5 G91G30X0Y0 M30 G8 can be specified nearly everywhere. The main condition is that you specify it in a block by itself. Most machines utilize "G8P1" to turn it on, then "G8P0" to shut it off. "Reset", M30 and such will also default the machine back to off. A few notes.... Most machines will need to have the G8 mode shut off for several things like canned cyceles (tapping, drilling, etc), High speed skip, polar coordinate interpolation, etc. It should continue to work for things like rotaion (G68), mirror image, etc. I normally don't ever have G8 shut off in a program unless I'm drilling and tapping. It's also good to have it called up in the main for every tool you use it on after toolchange. That way, if you restart a program, you're in G8 if you need it. Then also add a G8P0 to your canned cycles in the MC post file.

John, something to keep in mind on gundrilling too.... I will always spin the drill backwards into the hole before turning it on. The drill you're using isn't that long on the DIA/LOC ratio so you're probably ok. But I had times with really long gundrills where the cutting edge catches the side of the pilot hole going in and crapping the drill out (and/or the part). Dropping the drill in spining backwards eliminated that.

Is this a large chamfer on the diameter? If so, how big?

Maybe Gcode's wording is throwing you off. It's referring to Spiral Out motion. In MC, under Pocket cycles, look at "True Spiral". Its that effect. You can use this or draw your own spirals to further control the start/finish points of the arcs as well as how the program takes the last sweep.

Whoah Nick, that's hefty... Even with 50HP its going to be a doozy.... I thing I know the part you're talking about too... I'd try as Gcode suggested. A spiral helix... blowing it out some with smaller undercutters would help too. To try and take it all with one tool would be a lot of cycle time. You doing this on the Kuraki or 1080? Watch your side load (depending on your hangout). Something I like doing for this type of stuff is to helix out radially (pick a number... an inch lets say), then helix ramp in Z (if the cutter can cut in that direction). If you can't ramp, then just stepping down in Z before mover out more radially. Something to judge this on is to watch the amount of time and linear distance a tooth spends "in the cut" and under the shelf. Your load meters will let you know....

You can do as Ron said.... Or, for G code... Get rid of the M29. Mazaks don't use a synch code unless you have an oldie with a Fanuc board. The G84 will work with your feed in IPR as long as you have an H1 or the machine parameter is set for IPR. Otherwise, you use IPM. If you use G84.2, then the tapping is in IPR regardless. quote: The H value needs to be there, but I cannot remember what the values meen. The H value doesn't need to be there but in your case, it does because your machine parameter is probably set to asynchronous.

I'm with James. I wrote (eerily similar) custom cycles to have MC handle all (99% as stated) of probing. The big difference on mine is that I did spend the time to condense some of the stuff. For example: Bore/Boss, Slot/Web, etc. Then using the drill parameter info combined with misc integer/value page, identified the detail and outputs of each cycle. Also included skew (find corner for G68), B-axis alignment (HMC), A-axis alignment (VMC), B-axis alignment for 360k indexer (still working on the macro some),etc. I did include a check position and position tolerancing variables as well. The only thing I haven't done is for DPRINT output and such. Even changed the names of the tab pages for the drill cycles to show the cycle you're "cutting". It really wasn't that tough to get going. I will say though, I still need to tweak it some, but its "usable". Just got to find more time....

quote: Psycho, 6 spindles in a Makino? Yep... Every machine builders puts out a "lemon" now and then... including the "World Class" builders. Just some builders put out more than others... The Makino thing was interesting. I believed the machine just simply didn't like spindles. Went through all kinds of stuff with Monktons. When the factory came out though, he noticed a few "odd things" about the previous installs. I'll leave it at that. IMO though, if factory has been out and you're still having problems, then they should be replacing the machine period. Or, they can just cart the damn thing away, refund the money spent, and move on to a different builder. Then, they (Haas and Monktons) can hope that you don't go after lost revenue as well....

I agree with LT... I'd be standing on top of the guys from Monkton. Doubt Haas is giving you the run around, it's Monkton. With barely over 300hrs and 4 spindles, they'd be picking up the pieces from the front door.... Went through the same thing with them on a Makino A55. 1/2 dozen spindles in 14 months... until factory came out...