Anti-backlash Nut worn out

[DanOrgan]

Hi there. I had worked with my LittleRP for a few month, not actually printing, but experimenting with different resins, pigments, flex vat constructions.
And recently print quality has decreased dramatically along Z-axis. Take a look at this print (Fun to Do black, 0,025mm layer, diy FEP flex vat, 800/2500):

Then, I disassembled Z-axis and found this:

That cause the issue, as I think.
I had ordered new parts from RobotDigg. Is there any way to prevent fast worn out of the Anti-backlash Nut?

[goopyplastic]

I haven't heard or really noticed this being an issue.

The two halves are spring loaded against each other so wear on the threads shouldn't really become an issue until the threads give way. You can increase the tension between the two halves, perhaps they weren't installed all that tight to begin with?

Also the arm is mechanically attached to the larger portion of the nut, the top of the nut pictured is just to provide tension so I would be much more curious to see that the threaded on the AB nut housing looking like, they are more likely to affect print quality.

I am more than happy to swap you a new AB nut for your old one to do a little post mortem.

[James]

Yeah, an anti backlash nut that has springs for taking up backlash can only be considered worn out when the threads are completely gone. In addition to the previous suggestions I would check the leadscrew bearing holders and bearing holder brackets for any amounts of play. Also, I have not seen the details of the system you are referring to, but another problem that can occur with anti backlash nuts is they can jam up when it comes to their linear motion away from each other as the spring tension tries to move them apart. If that happens the spring cannot perform the compensation.

[DanOrgan]

Another, larger part of the nut is not in good condition too, but I was too lazy to take a photo of it.
Bearings are probably OK, they are inside NEMA stepmotor, I can't access them, but they "feels" normal.

Maybe, I was wrong with threaded rod oiling. I had oiled the rod only few times. Need to do it regular.

Goopyplastic, thank you for the offer, but shipping cost will be higher than Anti-backlash Nut. Anyway, I had ordered some parts from robotdigg, want to try FEP vat + tilt modification to solve any possible "stick to vat" issues with any resin.

[James]

I see I see.

I'd just wanted to add that I once had a ballscrew with five ten thousandths (12.7 microns) play in it and I couldn't feel it with my hands. With my hands it felt like zero backlash, but when measured it showed up clearly.

[DanOrgan]

The Anti-backlash Nut was not one to blame.
I had changed it and see only slight improvement on Z-axis surface quality.

Most likely, some problem with the Stepper Motor Driver. It seems, the stepper motor was not able to execute G-code command in time. Every time I look at CW "Estimate Build Time", it is much shorter then the actual build time (20-30% difference).

[James]

Could be, but if you change it out and you still get the inconsistency I would check some of these things.

What about the linear way? Could it be that the linear way is too cumbersome for the stepper motor? And then there is the stepper motor itself. Is it an old surplus motor from yesteryear? If so, older designs of stepper motors can be quite a bit less than the rated power for some brands. I got some old old OLDDD Slo-Syn stepper motors once that somehow lost most of their permanent magnetic field strength and consequently were VERY weak. Also along this line, I was loosing steps once on a CNC lathe I had built. Turned out there was some kind of goo (best description would be chewing gum that was dark in color like molybdenum disulfide grease and had got formed into thousands of tiny spheres) that got between the rotor and the stater that was significantly reducing the power of the motor. I cleaned it out and everything started working perfectly after that. It was surplus, so who knows what that goo was or where it came from. It was from a batch of premium quality Pacific Scientific NEMA 23 Powermax II so I was baffled as to why it was loosing steps and couldn't even think of a way that could happen until I took it apart.

[DanOrgan]

Sorry, I forgot to answer to this thread.
It was not a problems with mechanic or electronics. My PC was the cause of that. CPU was to weak to transfer G-code to Arduino, and it is results in loosing steps.
Now, my prints are not bad:

[James]

Great print! Glad you tracked it down and thanks for telling us what it was.

[James]

Oh! I didn't notice that! And I'm a jeweler too! HAHA!

Yes! I too would love to see how this comes out!