Hello Guest it is March 28, 2024, 05:12:12 AM

Author Topic: Y AXIS with 2x Ball Screw Alignment  (Read 4433 times)

0 Members and 1 Guest are viewing this topic.

Y AXIS with 2x Ball Screw Alignment
« on: April 29, 2014, 11:01:40 AM »
Hi
I'm looking for some advise about how would be the best way to bring the 2 ball screws into alignment with each other, I have had some problems with Homing errors of approx 0.1mm so maybe with this can be a little better for what I'm looking at doing anything under <0.05mm would be ok.

As can be seen in the pictures the timing belt has adjustment on both the Top and Lower sides and equal tension on both sides would be required, My thinking here is that each of the Ball Screws must be sharing the load equally in both directions.

I'm also having thoughts about loosening the Taper lock on the Driven Ball Screw and somehow measuring how much backlash there is, perhaps this would allow me to find the sweet spot between both Ball Screws.

I'd appreciate any thoughts about how to go about this  ;D
 
Breastfed

Offline rdean

*
  •  94 94
    • View Profile
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #1 on: April 29, 2014, 07:16:00 PM »
If I understand correctly you have a deviance of 0.1mm or about 0.004" between the movement of Y axis sides.  

What is the pitch of the ball screws?
What is the diameter of the belt pulleys?

Ray

Re: Y AXIS with 2x Ball Screw Alignment
« Reply #2 on: April 29, 2014, 07:51:55 PM »
Hi Ray,
Yes correct 0.1mm or 0.004"
The pitch is 15mm or 0.590"
Pulley diameter is approx 25mm or 1"

Breastfed

Offline rdean

*
  •  94 94
    • View Profile
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #3 on: April 29, 2014, 08:43:01 PM »
OK the most likely reason for the in balance between the screws is from belt stretching.  I commend you on having built a very excellent adjusting system but there is still a stretch on the drive side of the belt.  

Everything else being the same if the ball screws had been of 7.5mm pitch or about 0.25" per revolution the error would be reduced to slightly over 0.05mm.  Since it is not practical to replace the ball screws then replace the belt pulleys.  

Increasing the belt pulley diameter to say twice that or 2" would make the belt travel farther to get the same result of screw movement.  Since the load or work done by the belt is the same then the belt stretch would remain the same but the error would be reduced significantly.  I will have to do some calculations to tell you actually how much but at first thought it would be 0.025mm.

Ray
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #4 on: April 29, 2014, 08:51:42 PM »
Place a shim (perhaps split) under the appropriate ballnut flange (or shoulder).

You might be expecting a bit much from that small of a belt over that length too.

Regards,
Russ
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #5 on: April 30, 2014, 01:57:09 AM »
 I have considered changing all the Ball screws to a 5 or 10mm pitch as there is some wear and tear but thats a later upgrade, I don't know what this Gantry was originally used for but I liked the look of it and the price was cheap so I cant take the credit for the adjustment. I really only wanted a machine to carve out HD Foam and wood for molds used in Casting Aluminium, so for this purpose I should not be worrying about 0.1mm.

I can see that bigger Pulleys & Belts and a better choice for the Pitch would help out with regards to Back Lash,

BTW the actual pitch of the ball screws seems to be 0.600" the shaft Dia is also 0.600" (ISSOKU E01 90 420188-2) are the markings on the Ball Nut and ISSOKU's website and catalog's don't help, perhaps its an old model that's been superseded, and I'm still google'n for this one. Does this numbering suggesting what the actual pitch could be ?

Breastfed

Offline RICH

*
  • *
  •  7,427 7,427
    • View Profile
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #6 on: April 30, 2014, 07:43:12 AM »
I think you should know the system before you make any changes. By that I meaning, check each component / part of the system by isolating each part to check for backlash / play / belt and bearing  adjustment, etc.

As Russ said you can shim the bearings to take out or remove play. It cannot be zero! You shim and then see how much torque it takes to rotate the shaft while also checking backlash. I use a torque gage which provides data as compared to how it feels. Make note of the backlash (it should be small).

After  the bearings are adjusted then you check the ball screw. You have what you have for the ball screw unless the nut is adjustable. Make note of of any backlash.

Do the above for both axis.

You are direct driven on the one axis. FWIW I have found problems with flex couplings in that they are not repeatable some times ( all depends on the amount of torque ). Direct drive that axis using  a solid coupling for testing the incremental and also the overall travel for checking the calculated steps per unit.
Make note of incremental variance in both the forward and reverse travel and any backlash on reversal.
Your checking the lead acurracy. Do the same for the other axis without the belts ( seems like you can easily mount the motor  on the other side the same way for testing.

Backlash will adds up so be meticulous. You now know what each axis can do. You may even tweak the
steps per unit for each axis as compared to calculated based on testing.

Now add the belting. Belt tension can affect accuracy. For short lengths i have found its in steps but can't  say how much for long belts. Two discs of know diameter ie; a  linear movement of belt equals rotation of disc can comparitively show  differenece between the movement of the two axes.
Easy to do ....place large ones on the idler pulleys and one axis. Align the marks and and see the differences between the marks on the pulley ( you can calculate the step differences) Play with tension to see the affect.

Summary:
Both screws move the gantry and you know they are refined / adjusted as best you can. You know
backlash for each axis and also linear movment accuracy for each axis.
You add the belting and now check the affect of belt tension and also check  linear movement of the belt as compared to linear movement of the gantry. See what is the worst component of the belting system. Note that a complete  rotation of the large belt should be repeatable as compared to the same linear movement of the each axis.

Hopefully the above reply gives you the gest of what to do....

RICH

Re: Y AXIS with 2x Ball Screw Alignment
« Reply #7 on: May 02, 2014, 06:25:57 AM »
Well after some thinking about all your suggestions and some messing around I have come up with what seems to be a repeatable +/-.004mm  when homing.
What I did was tighten the belt considerably using both of the belt adjustment pulley's then I loosened off the Taper lock in the Driven pulley to hopefully relieve any unbalanced tension in each side of the belt.

The Y axis of the Gantry has a natural spot where it likes to sit without forcing it in a + or - direction, gently locking up the Taper Lock and checking this a few more times until I was happy with the results.

Gents thank you for sharing your wisdom and helpful suggestions  ;D

Breastfed

Offline Sam

*
  • *
  •  987 987
    • View Profile
    • hillbillyhilton.com
Re: Y AXIS with 2x Ball Screw Alignment
« Reply #8 on: May 05, 2014, 03:07:11 AM »
Something to be aware of....place an indicator on the axis that is controlled by the long belt. Place your finger in the middle of the belt, and press downward (or pull upward). Note how much the indicator moves. Now, 0.004mm (or even 0.1mm) most likely seems like a small drop in a large pool. The belt will most likely spring up and down during use. What's really entertaining is when the programmed axis motion hits the resonance frequency of the belt, and it REALLY starts vibrating like a plucked guitar string. Springing can cause all sorts of unwanted issues. Cut quality being one of them. On long stretches of belting, it really helps to have tracks for the belt to lay in, to eliminate the springing. I'm not trying to be a "negative Nancy", just pointing out something to keep an eye on.
"CONFIDENCE: it's the feeling you experience before you fully understand the situation."