Z Axis - precision and accuracy

Is it something to do with the firmware. When i received my printer it was working fine. Then I install the latest firmware update and that is when i think it went down hill from there. Mike guided me to the thread where Jerry uploaded the firmware for the auto bed leveling, and that is when i noticed the ribbing issue.

 

We seem to have a problem with a microstep position snapping to the wrong microstep. If there was a problem with a tooth on the rotor it could make a full step a bit weaker (less torque) with some periodicity related to the number and width of the stators. Does this make sense? I'm open to these suggestions but I don't see them corresponding to Ziggy's measurements yet.

Also - thanks Ziggy for your persistence in this debug.

 

Are your threaded rods dead center in the holes at the top of the housing? For me it seems no matter what adjustments I make mine rides that side wall. The opposite side is towards the outside as well, not as bad as as the left though.
They maintain the position there But now that I am watching the left side I would bet money that is part of my problem.

 

If that nut is resting inside at an angle for some of us, that could be it right there!

 

I'm seriously considering these as an upgrade... http://www.omc-stepperonline.com/th...-body-300mm-lead-screw-t8mm-8mmrev-p-218.html

 

Problem is this. Just one revolution of the stepper will lift 8mm.
Thats not exactly going to give precision for z axis.
These are better for fast axes like x and y.

 

I'm wondering if a neat resolution and stability improvement might be to get an ultra fine M9 or M10 Threaded rod, with 0.75 or 1mm pitch, turn down one end so it locks into the 8mm couplers properly. Then machine some nuts for it to fit in the m8 nut holes. Lots of work to machine, but could be an easy upgrade to install if someone else does the machining for us. Obviously the firmware needs changing for the increased steps per turn that would result.

 

They do have these as well. http://www.omc-stepperonline.com/th...ody-300mm-lead-screw-t635mm-2mmrev-p-225.html Earlier in this thread though, Jerry mentioned that the coarse thread was preferred for the z-axis as it lowered the coefficient of friction.

 

My understanding is that stepper motors are only accurate when on a full step. It has been mentioned before, but wouldnt it be worthwhile based on that to disable microstepping just to see if the issue goes away? It would either eliminate the microstepping thought completely or bring it to full light.

 

Hmm. I dont believe for a minute that too much friction is the problem. The torque of the nemas x2 only has to over come the force to lift the carriage. A tiny distance at a time.
Well lubed rods 9f any pitch wont be the problem. What magical thread would be ideal? Maybe I can machine the perfect non standard rods? Yes, I have the tech at my disposal.

 

I thought it was a little strange too but apparently some research went into it...

It looks like the stock imperial rods are relatively close to 2mm per revolution.

 

Research? Interesting. Maybe so but

"The fine thread means the threads are at a smaller angle from horizontal. This translates to the force from the weight of the x axis assembly being amplified in the vertical direction"

Doesnt sound like the full story. "Hey my engine isnt powerful enough to drive up the hill. I'll find a steeper one, that will be easier!"

Taken by itself, The logic isnt good.

I may have better results since I threw away the hopelessly weak z rod holders and replaced them with some cnc machined ones. My z rods are "solid" with nothing touching the case.

Additionally if I do try it. I wont be using off the shelf fasteners (which arent intended to travel all day anyway).

 

My z-rod holders were cracked on delivery. They travel about as well as I do. I've got brackets at the tops of the rods for now. Since I'm going to have to redesign the two pieces that travel up and down the threaded rod anyway, I may do dual smooth rods on each side. Something significantly more substantial to hold the bottoms of the rods as well. When I do it, it'll be detailed here... http://forums.robo3dprinter.com/index.php?threads/donhuevos-e3d-bowden-fed-x-carriage.2555/

 

*sigh*

Ok so, with translational forces you have what are called point of contact. Points on a surface that contact points on an opposing surface to make movement possible. With a spur gear you have 1 point of contact laterally, with something like acme thread you have many many points of contact. These points of contact make more precise movement by restricting points of non-contact between the two surfaces.

With a coarse thread you have less points of contact because the number of threads per distance are less. Thus a finer thread means more points of contacts, meaning more friction. Thus the amount of torque required is greater.

The logic is indeed sound.

 

I understand that more engaged threads can equal more contact points, but im impressed if its causing the level of increased friction. It sounds like you have tried. That doesn't mean it cant be made to work. Its not really like the z axis is doing much especially with two motors...

So anyway, just recess the thread in the nut or make the nut shorter, less threads means less contact area but it also means higher contact pressure. I would bet there is a decent oil or grease that would make this a whole lot better, especially if the rod threads were precision made. I just want better prints and im sorry to say im not 100% convinced by the engineering statements coming out.

Either way if I attempt it its my time wasted.

I wish there were affordable precision ball threads that had lower pitch. Like 2-3mm.

 

I gues I just assumed that the two nema 17s would make light work of the z axis esp since they have such a low 'duty'.