Accuracy upgrades and stepper motor lifespan

Okay, so my print quality has been steadily getting worse and worse over the last while, and I've traced the problem to the stepper motors that came with my kit. It seems that a lot of rapid reversals have taken their toll on the parts inside the motor, and the result is an amount of backlash that is no longer acceptable to me. It probably doesn't help when I get the occasional failure of the mechanical end stop switches (The little levers flip too far open, a wire prevents the switch from closing, or something gets knocked out of alignment) and then the motor tries to push things beyond the physical limitations of the printer. That can't be doing the motor much good either.

I have new motors which I plan to swap in, but I'd like to improve the lifespan of these, and maybe get more accuracy to boot.

I bought a set of optical end switches, which I hope will improve the reliability and consistency of the printer's margins. I still need to work out a mounting scheme which will allow proper travel and accuracy, but I don't think that'll be too hard, and there are plenty of examples to follow.

I'd also like to do away with the belts for X/Y travel, and replace them with threaded rod. I got a set of 8mm threaded rod which has a 2mm pitch and 4 starts (So, a total of 8mm travel per revolution) which combined with a 200 step/revolution motor would give 25 steps per mm of travel (Or 400 pulses at 16x microstepping). Compare this to the current belt system which is close to only 5 steps per mm (Or about 78.xxx at 16x microstepping) and you can see that a bit of slop in the motor will have about 1/5 the effect on print accuracy with the threaded rod than with the belt. I'm pretty sure that the printer doesn't have to be able to position the print head to 0.0025mm resolution, so I was going to change the RAMPS microstepping from 16x to either 2x or 4x.

My concern with the screws is overall wear and tear on the motors, as well as print speed. With the increase in motor RPM could come increased wear on bearings and motor rotating assemblies. the real issue though is a potential loss in print speed. I'm sure that at 4x microstepping, the change from 78.xxx to 100 steps per mm isn't going to be an issue, and that the Arduino and RAMPS boards can handle it at the print speeds that I'm using, but can the motor? I have a feeling that I should probably change the acceleration settings as well, since the threaded rod will have more mass than the belts (Though both are minimal compared to the X gantry and print head and the Y movement of the print bed.) I am reasonably sure that the motors can handle the increased RPM and maintain accuracy, but there's always that doubt until I actually try it.

I'd also have liked to get good anti-backlash nuts. The threaded rods come with brass nuts that have a tiny bit of slop to them. I'm thinking that I could probably rig something up with another set of nuts and some springs that would do. I'll need to do a bit of research on that.

So in order to change from belt driven to threaded rod driven, I'll need to change a few parts, the motor mounts need to be changed by 90 degrees, and the belt clamps need to be changed to hold the nuts for the threaded rods. Designing the parts isn't hard, but getting decent print quality to make them... Well, hopefully replacing the motors and giving everything a good tune-up will result in adequate print quality to make usable parts.

I think you're overly concerned about motor wear. I would expect belts to wear out much faster than motors. Even cheesy steppers ought to give many years of service in a 3D printer. Look at how they are constructed. Ball bearings and steel shaft. There isn't much to wear unless the bearings are really poor quality (possible, I suppose, in really cheap motors). If you are really concerned about motor life span, use quality motors- don't just replace bad motors with more bad motors unless you don't mind replacing them periodically. Larger motors have larger bearings and are built for larger loads, so a NEMA-23 motor replacing a NEMA_17 motor will probably have longer life span.

There's only one limit switch that has any effect on print quality/accuracy- the Z axis switch. It determines how thick the first layer is actually going to be in tandem with whatever you set when you sliced the STL file. X and Y limit switches just tell the machine where the XY home is and doesn't really matter much. Slicing software will generally try to arrange prints around the center of the bed.

Changing to screw drive will require a recalculation of motor torque. Massive rods require more torque than belts. Antibacklash nuts will increase friction, also requiring more torque. Check out the motor sizing tools at Oriental Motor's web site for an easy to use calculator. I have a ball screw driving the Y axis in my printer.

Edited 1 time(s). Last edit at 09/11/2015 01:22PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Thanks.

I am aware of needing to recalculate torque. the rods that I have aren't too heavy, so I think I'll be fine. My concern with the limit switches is because I do prints where I change the filament, and sometimes I bump things and move them, so it's nice to be able to just re-home the X and Y axis and get back to where I was, but I've found the occasional problem where the X axis isn't quite where it was, because the switch holder has moved slightly. For single colour prints, things are fine, I don't get missing steps often enough to matter. I've got the Z axis switch pretty firmly in place, but it sometimes needs to be adjusted slightly it seems. I will probably change the maximum acceleration settings in the firmware to accommodate some extra mass. Friction losses are actually pretty bad with the belts being tight enough to avoid backlash as it is, even with decent bearings, and I'm really not a fan of how the belts need periodic tightening, sometimes even needing a recalculation of steps per mm where large prints will be off by a bit after re-tightening. (A better clamping system for the belts might help, but again, if I'm making new parts...) Hopefully motor wear will not continue to be a problem.

I'm pretty sure that it's my stepper motors which are causing the backlash that I'm experiencing. The belts are in good condition, I replaced them as a precaution, and things aren't better. I'm confident that I have the tension set appropriately as well. I'm pretty sure that the kit I bought came with the cheapest stuff that the seller could possibly find, and I've replaced quite a bit of it with better since, every time I do so, things get better, which is good, but now things are getting bad again. Well, I'll replace the motors and start with making new parts, but it's not like I'm going to just toss the old parts, so if things don't work out, it shouldn't be too bad.

Another issue with the end stops is just the overall design of this printer, the mechanical switches pretty much force things to be in such a position as to reduce my printable area by about 12mm in each direction, and I'd like to be able to use as much of the bed as possible. If I have to re-design switch holders anyway, I'll move them over to the optical ones at the same time.

It will also be nice to have a good round number for steps per mm, which the ball screws allow, and the belts don't.

What type of belt are you using? Some aren't designed for motion in both directions and as such might be more prone to backlash even when new. Make sure that the pulleys are tight on the shaft as well, as loose pulleys are often the cause of additional slop in the axis movements.

I removed the metal arm from my z endstop switch. seems to be more accurate.