[JerryBurks]

I built my first CNC machine (see the Big Bamboo thread in the build log section) and overall I am quite happy with the outcome. One thing that bugs me is the pretty high friction of the lead screws. I can turn them by hand, no problem when gripping the (fat) motor couplers but according to my uneducated hand feel it is harder to turn then it should be. No much problem for my over-sized NEMA34 steppers but I suppose it is partially responsible for them stalling at higher speeds.

Anyway, I was wondering if I could improve this if I replace the pairs of 3/4" tapered roller bearings that I use for lead screw thrust bearings with regular angular contact ball bearings. While they do the job well with proper pre-load, I am wondering if they just have too much friction (being trailer wheel bearings, after all). For that matter, I am not aware that anybody else here used roller bearings....

Any opinion? Thanks!

JB

[RomanLini]

Hi JerryBurks, I don't have an opinion on the tapered bearings but I would suggest quantifying the drag.

If you put a small lever on the coupling (even an allen key in the grubscrew) then measure the force needed to rotate the screw at that distance and calc oz inches.
Then you could do the test again with the bearings removed and the oz inches would be for the motor detent torque and the leadnut drag.
The difference between the two measured torques would be the drag solely from the tapered bearings.

[crane550]

Hi JerryBurks, I don't have an opinion on the tapered bearings but I would suggest quantifying the drag.

If you put a small lever on the coupling (even an allen key in the grubscrew) then measure the force needed to rotate the screw at that distance and calc oz inches.
Then you could do the test again with the bearings removed and the oz inches would be for the motor detent torque and the leadnut drag.
The difference between the two measured torques would be the drag solely from the tapered bearings.

Agreed. This will give you a good idea of exactly how much your dealing with. Something else I would recommend would be disconnecting all your leadscrews and seeing how hard the gantry pushes by hand. Run it back and forth a few times with no influence from the motors or screws and see if it gets harder or easier anywhere on the table. If so, you have a better idea of where to start looking. Then install the screws, but loosen your nuts (I assume your talking X axis? This will work for both actually) and spin the motors via the belt and see how it feels. If there is ANY misalignment from your screws and leadnuts you will get binding.

To prove my point, consider this:

https://lh4.googleusercontent.com/-P...-45-34_732.jpg

When my Z reaches the top, it starts to bind. Not to the point of it stalling, but there is a noticeable bit of resistance. This is because the lead nut is ever so slightly off. Less then what a digital caliper will read. When loosening the top screw mount it gets easier. I need a shim of about .003 to correct this. My point is, even if your dead nuts close (my machine work was done by top notch professionals, my bill will prove it) if it's not exact you will have binding. The ACME nuts on my Solsylva were just the same. Try loosening your leadnuts and got it a bit, see how that affects your top speed.

[arizonavideo]

If you replace the grease in the TRB with way lube you will much less friction.

For your use they can run almost dry.