I hope im not posting this in error.
Does this sound strange. My x axis when using MDI and homing will not place the head in the same spot twice from home. Y is on but x is off.
So g20 g90 f500
G1 y9.834 x15.772
G91
And its diffrent every time. Also i have cuts inside of projects that are off in the x axis
Check the nut on the end of the ball screw, it should be tightened to where the washer under it won’t spin.
Pat
Hey Thomas,
looks like you loose steps on this axis.
Onefinity provides a help file which deals with many kinds of erratic movements of an axis:
I would check all points one by one to see why your carriage seems to loose steps.
There are three causes that are not mentioned in the document above:
Should your machine not be properly rectangular (“squared”), the Y movement can block.
Have you ensured accurately that your machine is rectangular (“squared”) (bar gauge) and coplanar (“not twisted”) (fishing line method)?
And the second possible cause not mentioned in the document above is: This mysterious issue some time ago.
Electromagnetic interference (EMI), as e.g. mentioned here.
The issues you have with Onefinity machines, both buildbotics-derived Original Series and Masso-G3-powered Elite Series, are usually not issues with the controller but
Lack of a serious cabling solution
1a. Connectors that are made for internal electronics are used on the outside of the machine/controller and subject to permanent movement (possible failures with unpredictalbe axes movements)
1b. Total lack of stepper cabling shielding, so overall susceptibility to EMI, especially when upgrading to a spindle (VFD = heavy source of EMI, but also the Makita hand trim router’s carbon-brush commutators emit a lot of EMI due to the usual sparks)
Thank you for the idea’s. I had a battle making sure I had a perfect parallelogram. So that’s not it lol. EMI is like the perfect scape goat to clear any problem isn’t it 
I willlook at it for sure.
Hey Thomas,
the total lack of any cable shielding is unusual in the industry (for this reason), the same with not enclosing a VFD in a grounded VFD Control Cabinet. Also a seller of a machine that has to be assembled and made functional by the operator and is a class I device should provide instructions for Proper Grounding; How have you grounded your frame? and for What is needed to ground your Onefinity CNC machine.
I would never put a Onefinity CNC into service in its delivery state. Even if it’s assembled according to the manual. I would at least retrofit the lacking strain relief on the connectors that are subject to permanent movement, and ground and bond all moving parts of the machine. Many lost steps and thus, lost workpieces, as well as static electricity induced by the dust collection, reportedly disappeared after grounding the machine properly.
Soooo Because I’m a doof I didnt look where you were telling me to look LOL. Till just now! The washer was loose enough to rattle. So I tightened it up enough to learn why you don’t over tighten it. …and then loosed it up a smidge. Too tired to keep going I will check it tomorrow . It was enough play to cover how much the axis has been off. THANK YOU!! this has been a problem since I bought this machine.
If this isn’t the problem ill be back:)
Hey Thomas, hey all,
I think everybody when having received the machine should think of metals changing their length with temperature. So on first setup, but also a few weeks later, I would have a look at how tight the bolts and grub screws on the couplers etc. are. A 120 cm ball screw gets longer by 0.283 mm if warmed up by 20 °C.
But also when put into operation, vibrations can loosen bolts.
I think to keep costs down, Onefinity and other hobby CNC router manufacturers, use a very simple connection system for the ball screw ends. The next step up - greatly in cost as well, would be to use machined supports at the fixed and moving ends of the screw (i.e.; BK12/BF12). They usually come with a pair of angular contact bearings and locking nuts with grub screws, so provide no, or at least specified, axial play. Their accuracy class can also be specified.
I am looking at such components for my Z axis assembly build, but these alone can add hundreds of dollars per axis. I think one needs to consider/measure their value relative to the rest of the machine components used.
I assumed that 1F did use angular contact bearings but without a more sophisticated method of setting them, does anyone know what bearing/ bearings they use?
The notion of tightening until the washer is not loose is pretty subjective in my opinion, as you mention it is related to economy of production and yet it does work pretty well.
Thanks, pat
I agree they work. I was thinking more of the matched pairs of angular contact bearings.
There are always consequences to such design decisions. One would be the extra space needed to house double bearings means a longer gantry block = more $. Also, given the tolerances expected from all those big linear bearings, does it make sense to add expensive bearings?
I always enjoy discussing such design and build topics, but always remind myself that I am not the one who has invested a great deal of my time and money, and probably risked a lot as well, to start and maintain a business. If I really was not happy with the direction Onefinity took, I never would have purchased one in the first place.
Hard for me to think a single row ball bearing in this application would provide a near zero back lash, in addition there is a one piece double row angular contact that I have used on CNC machines in the past that do provide near zero backlash. A single angular contact would not work in my opinion.
I would like to know what bearings they use just out of curiosity.
Pat
Myself as well…but I am too nervous about touching my machine as it seems to be set up and working so well for me. They sell it as a replacement item for $2.50 for customers only. At that price I can’t imagine we are talking NSK or SKF 