Racking Gantry, Y Axis Paired Stepper Motors Stalling

[BoneMamman]

Hi folks!

I have recently "finished" building my CNC Router. I have done everything in terms of Acorn setup and testing that I am aware of, and the initial motion tests were successful. In fact, after doing my homework, which at first seemed daunting, getting Acorn to run was quite a breeze.

The problem I'm having is that when the Y axis has to move "fast" over a distance greater than a few revolutions of the stepper motors, one stepper motor will stall, racking the gantry (sometimes more than a few millimeters out of alignment), and causing the other stepper to stall. The two Y axis are software paired, and drive two SFU1605 ball screws, through a 1:1 timing belt coupling.

Through bench testing, the steppers could rotate pretty fast, corresponding to 8000mm/min. I did not even test until stalling, as that seems plenty fast for me right now in these beginning stages.

On the actual machine, I can get the X axis to move 8000mm/min with no problems (I limited the max rate to 7000mm/min just to be safe). For the Z axis, since it has only a few inches of travel, I limited it to 4000mm/min.

The Y however would have this stalling and racking issue when it would rapid. So I initially limited it to 4000mm/min. Then after a few incidents it would stall at slower and slower speeds. I now can possibly reach 2000mm/min, but at this point I have the Y axis screws and motors disassembled while I try to troubleshoot the mechanics. Also after a few of these racking incidents, one of the Y ballscrews has this very light and "dry" dragging or scraping sound, so maybe i messed up the races and it looks like the screws are slightly bent (like ~0.004" runout over 750mm length). However the X axis screw isn't great either and that can go much faster (though the distance between rails is about 4", versus the distance between Y rails is ~36".(I've ordered new screws just in case)

I think this is most likely a mechanical problem, there probably is some binding, or misalignment between the two linear guide ways, 4 carriages, and two ballscrews and ballnuts. My machine is based on aluminum extrusions from 80/20 (which is why I spent very little on the bearing components, I see no point in using Hiwin/highquialitybrand on aluminum extrusions) so I know they aren't perfect reference surfaces, but with my initial tests, I was getting things pretty darn close to square and true.

So to get to the point, I am wondering if anyone has suggestions in the realm of software/electronics troubleshooting. (I also won't say no to mechanical suggestions) I wonder if it's possible that I'm driving my steppers improperly, or if there's a synchronization issue that causes even a slight bind in the mechanics to cause a stall. Would bumping up the pulse rate from 100,000Hz help?

Here is a google drive folder with all the documentation I have gathered about the machine at the moment:
https://drive.google.com/open?id=1A0of4 ... XMS-SKSGra

Best,
Max

[Richards]

I use the same type of stepper drivers with a 48VDC power supply. Your motors have about 425 oz-in holding torque and 3.4mH inductance, which should work very well with that driver. (The lower the inductance, the "snappier" the performance.) It looks to me like you've found the problem with the ball screw. Have you run a dial indicator on the Y-axis to show how parallel the two rails are? As a quick test, I would add some lightweight grease on that axis and then run the machine at moderate speed until the ballscrew was warmed up. I'm not at my test bench now, so I can't bring up a Centroid screen, but I believe that the default acceleration in Wizard is 0.5 seconds. You might try lengthening that time to give the motors more time to accelerate.

[BoneMamman]

Thanks for the reply Mike!
I'm currently in the process of realigning the Y axis rails, I have a decent if not janky method of measuring parallelism with a dial indicator and a long strip of thick plywood, but I'm having trouble figuring out a way to measure straightness of each rail other than using the extrusion as the reference. I'll try that warm-up test when I have the machine back together. Also I had the stalling issue even at Accel settings of 1.5 with a max rate of 4000mm/min.

Another question, maybe if you or someone has a good answer. How critical is it to use the circlip on the "floating" bearing end of the screw?