G93 requirement on Haas

[So not a Guru]

We have 3 Haas UMC750s. For the past 6 months one of our machines has been stopping with These errors:

9971     EXCESSIVE AXIS SPEED OR ACCELERATION

1.103    X AXIS SERVO ERROR TOO LARGE

It requires resetting and restarting the program, and as often as not the machine will then breeze past the problem point, only to alarm out again at a later point. This causes a lot of lost time. The errors come on both 3+2 dynamic paths and simultaneous 5X paths.

We've had Haas out here several times, but they have been unable to fix the issues. Now they are telling us that the problem is that we are not using G93! We have never needed to use G93 on any of these machines before, and the other machines do not get these errors. 

They want us to test using G93 on the machine, That will require that we get our post modified, a post that we've gotten tweaked to work exactly as we want it to.

I think Haas is blowing smoke up our shorts, but I'll be the 1st to admit I'm not sure.

Has anyone else had similar errors?

Are any other UMC750 users forced to use G93 in order to get good paths?

[Colin Gilchrist]

1 hour ago, So not a Guru said:

We have 3 Haas UMC750s. For the past 6 months one of our machines has been stopping with These errors:

9971     EXCESSIVE AXIS SPEED OR ACCELERATION

1.103    X AXIS SERVO ERROR TOO LARGE

It requires resetting and restarting the program, and as often as not the machine will then breeze past the problem point, only to alarm out again at a later point. This causes a lot of lost time. The errors come on both 3+2 dynamic paths and simultaneous 5X paths.

We've had Haas out here several times, but they have been unable to fix the issues. Now they are telling us that the problem is that we are not using G93! We have never needed to use G93 on any of these machines before, and the other machines do not get these errors. 

They want us to test using G93 on the machine, That will require that we get our post modified, a post that we've gotten tweaked to work exactly as we want it to.

I think Haas is blowing smoke up our shorts, but I'll be the 1st to admit I'm not sure.

Has anyone else had similar errors?

Are any other UMC750 users forced to use G93 in order to get good paths?

I underlined the important part. Are you saying that the same program will run without fault on either of the other two machines?

Are you running with TCP and DWO? (G243 and G254)

I've seen some very funky things on these machines, depending on the vintage (build date) of the particular machine. Keep in mind that the UMC-750's are Haas' first attempts at using more advanced software functions in their controls.

From what I've personally witnessed, I believe that whoever is telling you that you need G93 is simply trying to pass the buck, since they themselves are incapable of solving your issue (and likely don't want to admit that they don't know how to properly solve the issue).

I would highly recommend that you get someone (capable) at Haas to come in and tune the drives on all 5 axes of this machine. (Even though it seems to be the X-Axis only that is causing your issues...) My hunch is purely based on anecdotal evidence, but allow me to describe the situation I personally encountered:

Back in 2008, or 2009 (can't remember the exact dates), I was helping the local Reseller in WA with a project on a brand new Haas VF-4 SS, with a TR-160 Trunnion table. I programmed and setup the job, and attempted to run it on the machine. This was a demo part, and was an attempt to show off some of the new High Efficiency Machining paths to some programmers at Boeing Auburn. When the program got to the section of "live 5-Axis" cutting, the drives on the Trunnion started stuttering, and shaking during the cut. Not only was the finish really bad, but I could tell that something was waaaay off at the machine level. I was unable to get the machine to run above about 20 IPM, without it trying to shake itself to pieces. Even when cutting very slowly, the path was faceting and leaving small gouges in the wall I was cutting.

I got on the phone to Haas, and they sent out a AE to tune the rotary drives. Well, the first guy they sent wasn't very experienced, and while some improvements were made, the results were still very poor. I made a couple more calls, and felt like I was getting the runaround. After finally getting on the phone with the sales guy, and telling him in no uncertain terms that we would be sending the machine back if this didn't get fixed, they finally flew in someone from the headquarters in California to work on the machine.

It took him about 30 seconds of watching it run to diagnose the issue as "improperly tuned rotary axis dive units", and he proceeded to change about 40 different parameters on the control. I didn't catch the entire description of what he had changed (I walked in on the middle of a conversation between him and the Reseller), but the gist of it was the feedback loop between the drives and the controller were set too tightly, along with the accel/decel parameters not being opened up enough. Once he finished tuning the drives, I was able to bump the feedrate up to about 150 IPM, and the machine ran the cut as smooth as silk.

For purposes of this anecdote, we were running the machine in G93 mode (Inverse Time), so my "20 Inches per minute" and "150 Inches per minute" were the G93 equivalents of those feeds.

Now, it is possible that running your machines in G93 mode would help, with the live-five portion of the program, but if the machine is also getting errors on the 3+2 motion, then G93 will not have any effect on fixing that error whatsoever!

So, between my past history with having tuning problems with Haas machine axes, and the fact that you are getting errors when the rotary axes aren't even active in the cut, that leads me to believe that someone is trying to CYA and blow smoke up your butt.

I would be on the phone, and making sure they pass you up the chain of command at Haas, until you can get to someone with enough authority to fix your problem. I believe the solution lies in having someone come and tune your machine properly, and/or install a newer version of the Control Software.

The hardest issue with solving this problem will be that it is intermittent. You can't just run the program until it hits "block 2205", for example, and say "see, there is the alarm, fix it!". So that makes it a difficult problem to solve for any technician who hasn't dealt with tuning the drive motors before.

Whenever you call for service on a machine (no matter the model or builder), the service department will always attempt to send the least costly technician (for them), to attempt to fix the problem. The issue you are facing won't be solved by a greenhorn. You'll need a Senior AE, or the equivalent to come fix this issue.

If pleading doesn't work, get on the phone with the Sales Guy who sold you these machines, and tell him you are going to return all of them for being defective, and are going to replace them with Mazaks or Okumas (whomever they compete with the most, in your particular area), if this issue isn't fixed. Tell him this issue has caused your boss to go nuclear and that you need this solved within 2 weeks, or the riggers will be dropping the machines off at the Haas showroom floor. I literally had to use a similar threat to get my issue solved a decade ago...

 

 

[WestRiver]

• 
• Members
•  3
• 6 posts

• Report post
•  
•  

Posted March 7

We too have a Haas umc 750 and a Postability Post. Ours comes up with error codes 2.103 y axis servo error too large , 9971 excessive axis speed or acceleration , 949 internal feed error detected. The machine just stops, right after gouging my part, started doing it when I first started doing 5 axis simultaneous machining now it is doing it on simple surfacing. It has ruined (5) , 3500.00 parts in the last 2 weeks. Haas finally fessed up to the fact that they have a software issue and are going to up grade the software  some day soon  I hope .I would double check the   WRZP numbers as I don't see any 12" numbers in the code you have posted It could be the software with yours as well. Good luck!

 

•  
• Quote

[So not a Guru]

1 hour ago, WestRiver said:

they have a software issue and are going to up grade the software

This is what appears to have done the trick. Haas updated the software today and, so far everything is working smoothly.

[Mark VIII]

Did they give you your $3500.00 back?

[Colin Gilchrist]

9 minutes ago, Mark VIII said:

Did they give you your $3500.00 back?

I've never heard of a Machine Tool Builder compensating a user for scrapped parts. Most of the contracts that go with a machine purchase limit the liability of the MTB... I believe that is what business insurance is for.

And, it was 5 parts, so the total is $17,500 I think...

[Mark VIII]

12 minutes ago, Colin Gilchrist said:

I've never heard of a Machine Tool Builder compensating a user for scrapped parts. Most of the contracts that go with a machine purchase limit the liability of the MTB... I believe that is what business insurance is for.

And, it was 5 parts, so the total is $17,500 I think...

 

It's never too late to do the right thing.

[So not a Guru]

51 minutes ago, Mark VIII said:

Did they give you your $3500.00 back?

That was WestRiver not us.

[WestRiver]

The date on my previous post was March the 7th and up till a couple of weeks ago everything was running well. Then one morning the machine wouldn't boot up

it ended up being the software again it must have ate itself up or something anyhow they installed the newest version again. We will see how it goes again.