UMC500 UMC 750 calibration

waikatowilly · Sep 14, 2020

so we have two of these onsite, a brand new UMC500 and a year old UMC750, in total we have about 15 Haas onsite

does anyone have a step by step calibration sequence for these that does not involve the useless MZRP?

Im just a fitter turner with limited CNC experience but HAAS cant get it right, neither can Renishaw

The UMC750 got close when the only good Haas tech we had in this country modified the parameters, he has since left and the machine was crashed, now its out, I have no idea what he changed originally

The UMC500 is brand new and is about 0.2mm out on the flip

I have used really old fanuc controlled Taiwanese machines for years, even a rip off Mazak 5 axis labeled as a Hyundai that held better tolerances than these two machines, surely there is a simple fix, I think that Renishaw system was worth about $10,000! what a waste of money

mhajicek · Sep 14, 2020

What exactly do you mean by "calibrate"? Do you mean finding the center of rotation?

waikatowilly · Sep 14, 2020

yep I guess thats what I mean, I follow all the exact procedures through the manual, the Haas tech has done this twice, the machine has been checked for level twice,Renishaw even came in and could not fix this blaming the program, no one seems to run a test piece? thats bewildering to me and what is the point of the MZRP if i cant get closer than 0.2mm

markp · Sep 14, 2020

Ive got a 2000 DMG DMU50V thats does not have 3d quickset. I use a simple manual method that gets me within .0005. Clamp an aluminum block on centerline. Mill top and perimeter down 1.5" to create a square part. Mill a pocket on all 5 sides. Mark the faces with a sharpie. Measure the side to side distance from edge of pocket to edge of part on all 5 side with a caliper.

Adjust y parameters to move y center to the middle within .001. Same with x parameters. Then adjust Z parameter for depth of the holes. This takes a few passes on the pockets to get close, 001-002". Then remove the part and measure with a cmm. Adjust again to center to last <001. Then adjust to get matching Z depth. My machine has temp compensation and sag compensation, not certain but I dont think Haas UMC have temp comp, so I would warm the Haas machines up pretty well before starting.

bryan_machine · Sep 14, 2020

The first question is whether the renishaw probe probe shank is tight (and not cracked), and the tool length for the probe is set correctly. (Do you have a tool probe on these machines? Has it been calibrated against a reference tool?)

Then run MZRP against the tooling ball (one comes with 3d quickset cannot say for haas) - say on the top and all 4 sides. If everything good, then the location should be constant. If not (appears to be the case) then how the locations different from 0 will characterize the error. [That is, setup MZRP, then go back and probe to see what issues you get.]

(This is not unlike what markp suggested, but it starts with measuring a reference object instead of machining a reference.)

No amount of calbration will change the tolerances of the base machine, or correct for things like drop of long drills, the machine being out of tolerance, bad fixtures, etc.

bryan_machine · Sep 14, 2020

Also, a Haas is not a DMG. In particular, the DMU class machines (at least from the era of my machine) are a single large casting with 3 feet - it's not possible for 3 points of contact like to not be in a plane. (And their motions in 5 axis are still not perfect...)

The UMC750 appears to have 5 feet - which CAN be out-of-plane with each other.

Haas's own writings talk about the limitations of MZRP and DWO, and point out that good results are very sensitive to the leveling of the machine.
(Has somebody run a ballbar test on these machines?)

MRZP and DWO

waikatowilly · Sep 14, 2020

perfect thanks heaps, all great options, Im no expert but this will all point me to getting this machine accurate, a lot more help than anyone the company has been paying for to come and try and get it right, really frustrating, especially because there is more Haas machines on order

No ballbar test done-Im going to have to google that(not ever seen or heard of this test), they just use machine levels

coyoinu · Sep 15, 2020

if you wanna pay for a round trip flight from california to NZ i'll come get those machines straightened out for ya.

waikatowilly · Sep 15, 2020

coyoinu said:
if you wanna pay for a round trip flight from california to NZ i'll come get those machines straightened out for ya.

ha ha awesome, if it wasn't for this bloody covid19 Im sure the company would seriously look at this!

Ill be cracking into this today and post back the results

waikatowilly · Sep 17, 2020

so after about 4 hrs and fiddling with the wrong axis I realized I could just adjust the centre of rotation, the UMC750 was 0.025 away from 0.00 to get 0.02 on the flip and the UMC500 was 0.0457 from 0.00 to get within 0.02 on the flip

so super happy its all machining good now

added picture of my test piece

5AXIS505 · Feb 20, 2024

I run a UMC-750 and have been struggling with its accuracy for some time. After spending 3 days with a Haas tech and messing with it myself this is what I've found.
The tools vary from cold to warm quite a bit, up to .004, this will effect the relationship between the tool tip and the center of rotation as it moves. When I say cold I mean the spindle warm up has been ran and another program the jobs all the axis for 10 minutes or so. After running parts for an hour is when i see consistent numbers. I have also noticed that my y axis will very about .0025 throughout the day. Sucks but I work with it. I leave setting 255 and 257 alone, but I have 2 numbers for 256 that Ill swap as the day progresses to correct the position. To figure out which one to use I pick up on my part in 4 views until I like what I see, which ends of being no more then .0008. Not ideal but its what i have to do for my parts to come out right.

Hebrewhammer8 · Feb 20, 2024

.004 tool growth? .0025 Y axis growth? Wow these machines are POS. My 10 year old VF4 the z/spindle grows less than .0005 and once warm X/Y will be vary .0005ish. Looks like haas tech peaked decades ago!

rjwalker1973 · Feb 20, 2024

waikatowilly said:
so after about 4 hrs and fiddling with the wrong axis I realized I could just adjust the centre of rotation, the UMC750 was 0.025 away from 0.00 to get 0.02 on the flip and the UMC500 was 0.0457 from 0.00 to get within 0.02 on the flip

so super happy its all machining good now

added picture of my test piece
View attachment 299659

Please tell me that is MM and not inches.

5AXIS505 · Feb 20, 2024

Hebrewhammer8 said:
.004 tool growth? .0025 Y axis growth? Wow these machines are POS. My 10 year old VF4 the z/spindle grows less than .0005 and once warm X/Y will be vary .0005ish. Looks like haas tech peaked decades ago!

Yeah not great, what weird is my x direction stays the same. I believe they are both based on the y axis just flipped 90 degrees. I've been informed that the z float is a software issue so I imagine the y comp issue is software as well.

Search

Largest Manufacturing Technology Community on the Web

Search

UMC500 UMC 750 calibration

waikatowilly

Plastic

mhajicek

Diamond

waikatowilly

Plastic

markp

Hot Rolled

bryan_machine

Diamond

bryan_machine

Diamond

waikatowilly

Plastic

coyoinu

Aluminum

waikatowilly

Plastic

waikatowilly

Plastic

5AXIS505

Plastic

Hebrewhammer8

Cast Iron

rjwalker1973

Stainless

5AXIS505

Plastic

Similar threads