UMC-500 flooding, not ready for production

[cameraman]

Hello Alex ,

I did almost the same.
I removed the sensor from the motherboard and remeasured a tool. I saw almost 1mm deviation.
At that point I realized what's going on. I measured the resistance of the thermal sensor and I added 20 M ohm resistor at it's place ( this gives a value of the I/O P45.1 of about 620 ( whatever this is ) which corresponds to about 24 deg.
I am running the machine with this hack since Friday , and it seems to hold about 0,01mm when properly warmed up.
My local HFO said that this week HAAS should came up with the new software to fix it but we will see how good it is.

Regarding the nod of the head in Y. On my machine I see about 0,03mm sag through the stroke of the Y axis.
Mostly after passing the center.
I don't know why HAAS put the tool probe in the worst possible place , as if you mainly work in the center of Y axis your tools will be shorter.

So far the rotaries are still surprisingly good. I put a micron indicator on B and I jog in 0.001 increments and holy moly there is
no backlash.

Another anomaly I see is the C axis. Especially the brake on it. When the brake is engaged I see 0.02mm of pull down of the indicator. On the B there is a movement of the needle when brake is applied but it's in the range of 2-3 microns.

I am doing the Ball calibration for a 4th time. I am getting +/- 0,01mm in X and Y offset variance.
Z is surprisingly about 0,05mm out ( no matter what I do ) I need to manually readjust it while machining ( after the calibration If I have to face a block on the opposite sides I get 40,1mm actual size instead of 40.00mm.
I am using a 2mm ball stylus on the probe , but don't think this is the problem.
I learned a lot in the past 2 weeks since calibrating and adjusting this machine....

After 3 days of big material removal - surprise surprise the chip conveyor channels are now clogged and 50% of the coolant is staying inside the machine while running.

So far the rotaries are still surprisingly good. I put a micron indicator on B and I jog in 0.001 increments and holy moly there is
no backlash.

^^^ That's really helpful / useful to know - a lot of Machine Tool Builders don't state practical sensitivity of movement for rotary axes (and associated backlash through a mechanism, compensated or not.). Important for making final finer corrections either referenced to the spindle or to a third (more accurate and precise) reference system.

The other values and measurements useful too. Interesting about the pull down of the table when braked but I assume dynamically the cycloidal drives have a lot of torque so the brake does not have to be applied a lot of the time. And when you do then you know the additional offset value / repeatability.

It's easy to forget that these rotary axes on 5 axis machines can have reasonably substantial runout / broader concentricity than something like the concentricity of a spindle on a cnc lathe. + the center bored hole on a table as reference surface is not necessarily going to be concentric to the slightly uncertain axis of rotation of the C axis rotary table. [Quick and dirty finding of centers.].

Repeatability over accuracy will be super helpful to close in on stuff (in terms of "Process" - blah blah blah );

HAAS advise to iteratively tweak a separate (intermediate ) working CAD model to converge on what you want as a final part. [In some cases.] or edit the cnc program directly.



MRZP and DWO

^^^ This is what HAAS themselves say about DWO + MZRP and realistic expectations etc.

HAAS troubleshooting section (as usual) has a lot of good stuff and details about calibrating the touch probe itself.

 

[GeorgiMirov]

Can somebody else also measure the sag of the head in Y on their machine please ?

I would like to know if those 40 microns are common or my machine is not levelled correct.

 

[Orange Vise]

Can somebody else also measure the sag of the head in Y on their machine please ?

I would like to know if those 40 microns are common or my machine is not levelled correct.

Common. A major weakness of this machine.

I don't know if it's 40 microns, that seems high, but it's definitely far from zero.

I need to pick up a granite square and check it out.

 

[GeorgiMirov]

Hello guys ,

I've been working a lot on some very complex small 5 sided parts and I gathered some info.
After the fix I made on the Z sensor - now the Z axis is the most repeatable linear axis.
I don't see more than 0,02mm difference ( typically 0,01mm when the machine is warmed ).
Problem now comes with the X and Y axis. They are drifting around 0,05mm during the day run.
Some may say use probing to fix it - it will work if it was a 3 axis machine.
Once you have such a shift in the X Y direction your rotary offsets are messed up. I did the manual adjusting on the parameters 255,256,257 several times , and the machine doesn't repeat the next day. When you reprobe your fixture and something shifts the error is doubled when at C180. So practically this machine is useless for complex 5 axis parts as the offsets around the rotary axes are constantly changing. I though it will be make it once set it and forget it ..... NO you can do it every day and still not get the parts right. My fixtures were remilled after every offset adjustment hoping that it will be good this time ... next day " oh here we go again " ...

I was hoping that they will fix the Z axis growth issue with the sensor , and they might do so , but the other axes will cause more problems and I don't know how they will fix those. There is some sensor on the Y axis but there is no sensor on the X axis.
If one of the linear axes don't repeat within 5 microns ( not to say 50 ) the rotaries are just floating in the space.

Also haas is playing with parameter 254 offset for unalignment of the rotary axes. If this is not set correctly the errors get even worst.

I said to the HFO if they don't came up with a solution by the end of this week , I will have to start a return procedure.
I had and still having great micron repeatability with my 2 years old 1st gen DT 2 but this UMC 500SS is a huge mistake, and will be staying away from 5 axis machines without scales from now on.

 

[thesidetalker]

Hello guys ,

I've been working a lot on some very complex small 5 sided parts and I gathered some info.
After the fix I made on the Z sensor - now the Z axis is the most repeatable linear axis.
I don't see more than 0,02mm difference ( typically 0,01mm when the machine is warmed ).
Problem now comes with the X and Y axis. They are drifting around 0,05mm during the day run.
Some may say use probing to fix it - it will work if it was a 3 axis machine.
Once you have such a shift in the X Y direction your rotary offsets are messed up. I did the manual adjusting on the parameters 255,256,257 several times , and the machine doesn't repeat the next day. When you reprobe your fixture and something shifts the error is doubled when at C180. So practically this machine is useless for complex 5 axis parts as the offsets around the rotary axes are constantly changing. I though it will be make it once set it and forget it ..... NO you can do it every day and still not get the parts right. My fixtures were remilled after every offset adjustment hoping that it will be good this time ... next day " oh here we go again " ...

I was hoping that they will fix the Z axis growth issue with the sensor , and they might do so , but the other axes will cause more problems and I don't know how they will fix those. There is some sensor on the Y axis but there is no sensor on the X axis.
If one of the linear axes don't repeat within 5 microns ( not to say 50 ) the rotaries are just floating in the space.

Also haas is playing with parameter 254 offset for unalignment of the rotary axes. If this is not set correctly the errors get even worst.

I said to the HFO if they don't came up with a solution by the end of this week , I will have to start a return procedure.
I had and still having great micron repeatability with my 2 years old 1st gen DT 2 but this UMC 500SS is a huge mistake, and will be staying away from 5 axis machines without scales from now on.

Have you tried playing with setting 158 & 159?

 

[coffeetek]

So after nearly a year of running this in unattended production I can say UMC-500 is utter shit for that. If you are going to run it with operator it will do decent job and can be used like that. If you plan to deploy it in unattended operation with robot etc. forget about it. Its crap.

1. Tool setter routine for tool break check intermittently errors out with Alarm 1092 PROBE OPEN/UNEXPECTED SURFACE FOUND. Service tech even saw it happening. Haas will not do anything about it.
2. Spindle probe sometimes would just hang mid operation without any errors.
3. Through spindle coolant pump is not working correctly from time to time and errors out. Haas also provided no help.
4. Uncontrolled Z-axis growth so far unsolved.

All in all, for unattended operation, run away hard from this machine. If you plan to run it with operator it will do OK job if you can deal with point #4

 

[kremin inc]

Found this thread a year too late - same machine, 8 pallet pool - same issues. Installed March 2022.....

 

[empower]

Found this thread a year too late - same machine, 8 pallet pool - same issues. Installed March 2022.....

haas strikes again!